Western Digital Corporation (NASDAQ:WDC) Morgan Stanley Technology, Media & Telecom Conference Call February 28, 2019 1:10 PM ET
Company Participants
Mike Cordano - President and Chief Operating Officer
Conference Call Participants
Joe Moore - Morgan Stanley
Joe Moore
Great. Thank you. I am Joe Moore from Morgan Stanley. Very happy to be here with Mike Cordano, President and COO of Western Digital. First, I do have to read this Safe Harbor real quickly. Please note that all important disclosures, including personal holdings closures and Morgan Stanley disclosures appear on the Morgan Stanley public website at www.morganstanley.com/researchdisclosures or at the registration desk.
Question-and-Answer Session
Q - Joe Moore
So with that out of the way, thank you very much for joining us. Maybe we just start with kind of big picture overview, it’s been 3 years now, I think, since you bought SanDisk. You have seen sort of the best times and worst of times and in over that full cycle. Just can you kind of put into perspective the reasons for doing that acquisition and the benefits that it gives you guys in terms of being kind of an agnostic storage supplier?
Mike Cordano
Yes, great. Well, I will do my mandatory disclosure. So obviously, we’ll be talking about some forward-looking statements. I encourage you to go to our site, our SEC filings and our risk factors and cover that mandatory part. So listen, relative to that, I mean, our premise for making the acquisition remains, in our opinion, valid and constant. So for us, we saw data growth and storage, technology in general as a high-growth area over time. We see ultimate bit demand in the market let’s call it mid to high-30s into the low-40s. That’s across both technology types, right. So as we think about this, an end market with that kind of growth gives a lot of opportunity. And what we chose is, let’s say, we want to participate across that growth with the full range and full portfolio technologies and products. So when we finished the acquisition, our starting point was really we had a product portfolio in hard drives, which we knew was evolving, right. We saw certain markets that were going to be in decline, but certainly, capacity enterprise is going to be a growth segment. We focused on that and continue to invest in that and continue to have a leading role, a leading participation in product portfolio and the capacity enterprise base. On the flash side of our business, as we came in, we had of course a very strong product line and market presence in retail, but we had more nascent participation across the commercial segment, so in enterprise SSD, client SSD, embedded and mobile. And we know we needed to do some product line expansion to be full participants. So our goal in that was to increase breadth and allow ourselves to diversify our customers but maybe most importantly, it was to get to the high-quality revenue, right. So that’s where we spent the last couple of years. We were running our OpEx a little higher.
What we are trying to do is make those accelerated R&D investments to put ourselves in a position to do just that. And so on the last earnings call, we talked about 2019 and where we are coming into this year. And we are in the best position from a product portfolio standpoint than we’ve ever been. And if you sort of go down that just very quickly by category, in the hard drive segment, we continue to lead at high-capacity, capacity enterprise. We remedied what was a product line kind of deficit in the mid-cap. So we have now solved that problem and we’re competing more vigorously in a better way there. And then, across our flash portfolio, really, we’ve done very well at getting compliant SSD products that cover the range, both value and performance. You saw some of the results of that, even in our last quarter reported. We had nice revenue growth and market share growth in that particular segment. And in mobile embedded, we’ve invested in UFS as a key technology enabler. That’s coming to market this year, so we can now participate both in value and performance in the mobile handset market, also allows us to expand our customer portfolio. And then, finally, in enterprise SSD, that’s been an area of disappointment in the rearview mirror in 2018. We are about a year behind our initial plans in entering that market. But as you might have seen a couple of days ago, we announced our NVMe offering. That is in qualification. That will be ramping throughout the year. So we see our participation in enterprise SSD again on the uptrend through the year. So that’s optimism about kind of where we are, the premise behind the acquisition and really our ability to have a positive relative performance as the year progresses.
Joe Moore
Great. Maybe before we go deeper in some of those things, just thinking big picture, I think you guys, as this has become a more challenging environment, have navigated it pretty well. You’ve talked about the digestion in NAND kind of earlier than some, some of the softness in cloud. What’s your visibility now? What kind of visibility are cloud customers giving you as far as when the order patterns come back? I know you’ve already seen some better trends on the drive side, I believe. And how much inventory do you think they have ultimately?
Mike Cordano
Yes. So I think our view is this, I mean, obviously, we talked about year-over-year compare, first half of ‘18 and ‘19, that we’d be flat to slightly down on bit consumption, all right? We maintain that view as we see it today. And a lot of that is really burn off of inventory. We see in our discussions with our major customers here, service demand trends remain pretty solid. So it’s really getting the inventories underneath in alignment. So we think that will be done in the first half of the year. We actually think we will see some positive trend maybe even late in Q2 going into the back half of the year relative to bit consumption rates from the hyperscale community.
Joe Moore
Okay, great. So in hard disk drives, you’ve made some strategic decisions on mission-critical versus nearline, prioritizing nearline, and I think you’ve sort of talked about walking away from a little bit of the mission-critical side. Can you talk about the rationale for that decision?
Mike Cordano
Yes, I think, for us, the strategic decision was to say where do we want to allocate capital? We have really been doing that on the basis of long-term growth. So we actually made a decision to de-prioritize performance enterprise in favor of enterprise SSD. So really, where we would say we judge ourselves sort of critically on performance was execution in enterprise SSD. And that’s been as we wanted. I think this would have all been the right ultimate strategic outcome. Obviously, on a longer horizon, we still believe we made the right strategic decision. But ultimately, as we looked at it, it was just a choice of getting capital between enterprise SSD which we saw as a superior opportunity versus performance enterprise.
Joe Moore
Okay, great. And then, last call, you talked about accelerating the closure of your factory, which will give you a lot of cost improvement over the year. It’s also caused you to build some inventory as you kind of do that transition. So maybe just talk us through any new developments there.
Mike Cordano
Yes. No, I think what you heard us say is we accelerated that. If you looked at our margin for our hard drive business in the quarter just reported, two factors in that. One is this build ahead of inventory as well as underutilization in our hard drive business at large. The other factor was the sort of decelerating growth in hyperscale. But as we go through this calendar year, we think, with the timing of that decision, we will have two things happen. One is as the year goes on you’ll see how our hard drive inventory come down significantly. So working capital will reduce and that’s sort of per plan, as well as our hard drive margin will continue to trend up across the year as we get our fixed cost in line, so really just saw the sort of demand profile for hard drives at large drove us to bring that forward a couple of quarters.
Joe Moore
Yes. Okay, great. And I guess, on the nearline side, it was a little soft in December, as you talked about, but the year-on-year exabyte growth are sort of stable in the first half would actually imply that the sequentials are pretty good there, but you sort of talked about not seeing a pickup until kind of later in the year, but I would think flat to up sequentially in March is pretty good, albeit off of a down December.
Mike Cordano
Yes, that’s right. I think, you are right, you’ve got to look – we’re doing sort of half year or full year compares. We do see sort of the sequentials improving modestly. But if you look at the year, we would say that exabyte growth will be in the 20% to 30% range, with the first half that’s either zero or slightly negative. So the back-end growth, we are still optimistic about.
Joe Moore
Yes, okay. And then, the gross margins, you alluded to 27% in drives last quarter, which I know is kind of well below the long-term run rate. How quickly do you get back to low 30s?
Mike Cordano
Yes, I think, certainly, we’ll be back in that range by the back half of the year, and we’ll be trending up between now and then.
Joe Moore
Okay, great. And then on the NAND side, maybe starting from a core technology perspective, you have talked about getting 96-layer, getting the cost crossover this quarter, big crossover towards the end of the calendar year. Can you give us an update on those transitions, and how would you benchmark your own cost per bit versus competition?
Mike Cordano
Yes. So I think, on an absolute basis, we maintained what we believe is lowest cost bit in the industry. I think, relative to our progress, we’re right on plan. So we already reached cost crossover basically in the quarter we’re in, and then we’re on track to get bit crossover in the fourth quarter, so that all is going per plan. And then, if you translate that down to products, we’re already in market in retail. We’re in market in client SSD. We’ll be in market with enterprise SSD as well as embedded products as we get into the middle part of the year and the back half of the year.
Joe Moore
Okay. And it doesn’t seem like the changes that you’ve made, the JV CapEx has come down a bit. You’ve [indiscernible] talk about. It doesn’t seem like that’s really set you back in terms of those 96-layer...
Mike Cordano
Yes, not at all.
Joe Moore
Okay, pricing has been tough. You guys were early, as I said, but to highlight that over a year ago, and there’s been some rumblings in the press that we could sort of see stability in Q2, but it also seems clear that there’s quite a bit of inventory on the balance sheets of producers, probably less so for you but from your competitors. So can you just talk about what has to happen for NAND margins to stabilize? And what are the lead indicators to look for there?
Mike Cordano
Well, yes, so I think look on both sides of the equation. So first, start with the supply side. I think what we see, and we are exhibiting ourselves is continuing to adjust supply to try to bring it more in line with our view of end market demand. Those are indicators to us of kind of more broadly, it appears the industry is moving in that direction, right? Continuously as we see what’s talked about publicly, people are adjusting their supply expectations, their production plans and their corresponding CapEx consistently with the way the market’s evolving. On the demand side, obviously, we’ve seen a [indiscernible] on the first half of this year. So what we would – obviously, our view is, very clearly, we’re going to see seasonal demand pick up in the back half of the year. So just bit consumption, both on hard drives and flash. So as we move into the back half of the year, we think we’re going to see total year-on-year growth let’s call it in the mid-30s, with that being more disproportionate in the back half of the year. So a combination of supply actions and what are people doing relative to that, which we see consistent activities trying to sort of deal with operating environment we’re in as well as just back half demand. So what we really needed to see is that demand materialized as we expected. That will be the key thing.
Joe Moore
Yes, okay. And I guess, is there a scenario where – I mean, Christmas will happen, there will be seasonal strength. Is there a scenario where there’s still too much inventory and prices – because we went through this last year where seasonality, we thought, would be more helpful than it ended up being because of some of the demand issues we saw.
Mike Cordano
Yes, I think it’s a matter of degree, right? I think that’s something that we continue to evaluate, and so we certainly think there’s some scenarios that are certainly trending to positive, but we’re going to continue to kind of put ourselves in a dynamic comparative posture. But I will say this, I think, in the current environment, in the quarter we just completed it was really a sort of no good news, right? We were just managing a lot of headwinds across all of our lines of business. Certainly, we see headwinds in the current climate today, ASPs we talked about in the flash, but there are some indications of some things that are going to the positive side of the ledger. So that is something we want to – that for us, is an indication of when we’re reaching sort of a better sort of position and a more stabilizing of our business. We are seeing a little bit better sort of end market demand for us, and PC is a little bit better with certain mobile providers than we would have expected. So those are sort of some positive signals in our business, and those are the things we look for as we begin to stabilize, and I think there’s cautious optimism around what we’re seeing now as a setup for the second half of the year.
Joe Moore
Yes. Okay, great. I know some of those positives came out from the call. You talked about a 60% increase in content for client SSD. It seems like units, nobody wants to hold inventory when prices are falling, so the units haven’t been that good. But it seems like when you get unit stability and you add that to the growth rate you’ve seen, that’s pretty good. Are you seeing that in other parts of the market? Are you seeing elasticity in phones? And are you seeing I don’t know if it’s elasticity in cloud but either elasticity or better availability sort of leading to more demand in those other markets?
Mike Cordano
Yes. And I think definitely, we are seeing elasticity kick in. Let me describe it because what’s different now than perhaps in the past for flash, flash used to be dominated by consumer markets, USB sticks, SD cards, et cetera. The amount of time you could translate at price point to end market demand was very short. When you think about these broader scale markets, mobile handsets or PCs, handset cycles once a year, right? So those plans are laid in concrete. They execute those plans meeting a mix of capacity points. And by the way, their price points are also fixed. So we don’t see the benefit of price reductions and stimulating demand, so that takes a little longer as they look at what their expectations are for ASP per bit. That will drive the next cycle, what their plans are for the next cycle, and you’ll see this delayed elasticity, right? Because once they understand they can get a lower-cost storage module, they’ll look at their portfolio announcement and we have seen a couple of announcements recently that sort of indicate where people are going. We see a couple of large providers talking about terabyte, right?
Joe Moore
Pretty good markup on that terabyte though.
Mike Cordano
Yes, pretty good markup. [indiscernible] product, but when you look underneath it, all the other mid – sort of midrange lines are moving up as well.
Joe Moore
Yes. Okay, great. And then on the inventory side, you guys made the decision, as you alluded to, to bring down factory utilization, which doesn’t happen a lot. It happens a little bit in prior cycles when margins were still pretty healthy. But it’s interesting that you guys are sort of relatively early to do that. Can you talk through the economic decision because I assume there’s still positive contribution on those wafers? So what...
Mike Cordano
Yes, I think, for us, at the level we saw, we were sort of comfortable carrying a certain level of inventory. I think, in a regime where flash cost declines around 15% per year, there’s a little less risk than say in the past when things were depreciating at a much greater rate in terms of cost declines. But at a certain point for us, there was a decision to make what we saw as economically positive and cash flow positive, which drove us to make this decision. So obviously, in this period, we’re looking to optimize certainly our P&L performance, but we’re also looking at cash, and we want to make a balanced decision that optimizes both. And when we looked at the cut, we actually saw substantial cash benefit for us of making reduction.
Joe Moore
And what drives that utilization going forward, like when you decide to bring those wafers back on? Just how do you go through that process?
Mike Cordano
Well, if you just listened to where we talked about the way that the cost of that decision roll through, we roll through it about 3 quarters because what that would indicate to you is the back half of the year, we expect to be fully utilized. If the market is where we think it is, we’ll be fully utilized the back half of the year. Obviously, we’ll continue to look at the market and make whatever appropriate adjustments, but at this point, that’s where we sit.
Joe Moore
Yes, and any indication that others are following suit in terms of that?
Mike Cordano
Yes. I think we see public indications of CapEx. I think we would view others considering let’s call it modest underutilization as well, they’ve been less public about it, but we would believe that people are making some of those decisions now.
Joe Moore
Okay. And how does that I mean, for you guys, you’re sort of non-GAAP in out some of those cost, but I would assume that there’s a negative cost ramification of that lower utilization. Just how do you is that a trade-off for you to think about? Or?
Mike Cordano
Sure. No, we thought about that, and obviously, the reason we treated it the way we did as a GAAP-only was this is something we don’t see as necessarily a regular operating decision. So, we chose to treat it that way. But certainly, if we were to do this more consistently, it would be treated differently.
Joe Moore
Yes. Okay, makes a lot of sense. And I think, when we think about NAND, I think there’s a tendency to just look at like the raw NAND prices that we see at retail or components. There’s obviously a lot of value-added parts of the stack. Can you talk about some of those and maybe we start with enterprise NV and ME that you talked about, it seems like I mean, the bad news is you haven’t been there good news is that you’re migrating towards that, so it’s become a rate of change perspective that’s good for you. But like what took this long to get into that space? And to what degree is that going to have to be custom controllers versus can merchant controllers, kind of help?
Mike Cordano
Yes. Well, in our case, frankly, across our flash business, we think there’s a strategic advantage of having our own controller and for more platform. So, we’ve done that in enterprise. So, what took us so long was we didn’t execute as well on that first generation.
Joe Moore
I didn’t mean to say it if I said it like that.
Mike Cordano
No, that’s okay. We’ll hit it straight up. But what we now have is a very solid platform. So, we have a controller architecture combined with a firmware architecture that we think will give us leading products in the market. So, for us, it’s about the breadth of coverage. There’s a lot of segments even within enterprise. It’s not a one-size-fits-all. So, we have to create a modular architecture that can effectively cover those markets from performance all the way down through value. So ultimately, we think the right way to do this both in terms of a value-add and getting again, trying to pursue that high-value revenue is really about delivering these custom capabilities that we can deliver unique differentiated value. So, we made a decision 2, 3 years ago that we’re going to have a platform for our embedded mobile, and that’s again a controller and firmware combination. Same thing for our client SSD and then same thing for enterprise SSDs. So, we are largely internally designed in our platforms and our controllers.
Joe Moore
Okay. And on the enterprise side, you mentioned the product is out. That market seems to be largely been Samsung and Intel today. How long does it take to get qualified with enterprise customers with those new products?
Mike Cordano
Yes, those qualifications depending on the player, is 3 to 6 months long. I think we feel good about our ability to do that. We’re a trusted partner. We have big roles with these same customers and capacity enterprise relationships. So, this is where some of the leverage of bringing the 2 companies together. So, we’re kind of we’re a known partner. We’re a trusted partner. They value the way we behave through all market cycles. So, they want us participating in the business. Now it’s up to us to execute. So, we have good pull. Now it’s on us to execute and deliver to our promises.
Joe Moore
Great. And you mentioned client SSD. I think of that as being a premium margin product over time where there’s a lot of value-added. I also look at retail prices for a terabyte at $125, subtract out the chassis and controller and the retail margin, and it’s pretty low pricing. Obviously, OEM pricing is probably different. But over time, is that going to be a high-margin higher margin versus raw components?
Mike Cordano
Yes. I think, over time, relative to raw components, certainly. And then, within that, there’s a spread at margin, right? So, I talked about earlier on, initially, our participation when we first acquired SanDisk was in the value side. Frankly speaking, there was one dominant player in the performance side through great product execution. So, we chose to pursue a product strategy that allowed us to participate both in performance and in value with optimized products. So certainly, participating in the high-end is part of how you blend that margin up as well.
Joe Moore
Okay, great. And then, on the phone side, there’s some portion of the phone market, you talked about having custom controllers there, but there’s also some portion that likes integrated DRAM plus NAND. How do you navigate around that and what are the interesting sort of interface opportunities?
Mike Cordano
Sure. So, first of all, the controller, whether it’s on MCP package with DRAM or discrete is the same. So, we would use our technology in the same way. Reality is we are not a DRAM player, so we skew towards trying to participate at the high end. At the high end, it’s more discrete business, so you don’t have to do the MCP packaging. We are an ongoing participant in MCP, but we do that more strategically, right? We don’t see that as an overweight segment of our business. We’re going to be there with our strategic partners at a level that makes some sense. But where we really want to differentiate is in discrete, sort of high-value offerings within mobile.
Joe Moore
Okay. And the high-end smartphone business has had some ups and downs as well. What’s your sense there now for customer inventory and sort of the demand trajectory like how?
Mike Cordano
Yes, that’s actually one of the areas that certainly we talked about and we saw that as we exited the year and our expectation coming in the year was that there was going to be some softness there. Relative to that reduced expectation, we’ve seen some areas of, actually, I guess, our plans of slight outperformance. So, a little more positive bias than we originally expected, even in high-end smartphone.
Joe Moore
Okay. So, I have a couple more questions, but let’s see if there’s any questions from the audience first.
Unidentified Analyst
Thanks. So, from a technology perspective, Seagate and WD have been diverging with nearline on the HAMR and MAMR. Can you talk a little bit about MAMR for WD, the advantages it brings to clients and why you think that’s a better roadmap?
Mike Cordano
Yes. So, I think, for us, we see in the next, let’s call it 5, 8, 10 years, this midrange, MAMR is, for us, the right option, and the reason for that is we think there’s a better risk-reward trade-off. And really, there’s a couple of things. One is the cost to bring it to manufacturing is lower, so we can reuse all our existing CapEx with no modifications. The other thing we can do is within our MAMR technology development, broadly all energy-assist, we’ve had some technology learnings that allow multiple positive aerial density effect. So, we don’t have to we can manage technology risk in terms of implementation. So, we will bring these to market sort of with the right combination of these positive effects at the right time to support our roadmap. And we are very confident, and we will be shipping that in volume production this calendar year towards the second half of the year. So really, for us, it’s about which of these energy-assist technologies are right. We chose MAMR, from our standpoint, it’s the lower risk-reward, particularly in this mid-range. Now we continue to do some investment in HAMR just as a sort of a prudent matter for the longer term, but we’re very comfortable with how we’re tracking relative to our milestones and enabling, let’s call it, the next 3 to 5 years of products.
Joe Moore
Questions from the audience? I wanted to get a chance a little bit to talk about the balance sheet because I think there’s been some misconception around the covenants and the debt. I mean, it seems like your debt maturity structure is very healthy. You don’t have to worry about paying it back for quite some time, covenants, like they’re in pretty good shape. Can you just maybe just address some of those?
Mike Cordano
Yes, I think, in general, the balance sheet is in a good position. We have a lot of flexibility within that and our liquidity is pretty good. So, from this standpoint, we think, given even these more challenging kinds of conditions we’re in now, we think it’s a manageable situation given the balance sheet structure and where we expect this calendar year to play out. So, I think we feel confident about it. We reiterated our commitment to the dividend. That should give you some sense for how we feel about it.
Joe Moore
Yes, okay. And then just a little bit on cash flow, it’s obviously going to be a little bit challenging cash flow environment this year. Can you just talk about what adjustments can you make as the year goes, on sort of on the expense side, I mean, you’ve already taken a lot out on the expense side. Just how comfortable are you that if this isn’t getting better in the second half, what that’s going to look like?
Mike Cordano
Yes. So, you’re right. We’ve already made a big move on the cost and expense. Obviously, you would expect we would do contingency planning, what if sort of planning if we need to do something beyond that, and we got those plans in hand. At this point, we’ve been able to do this in a way that has protected all of our technology and product investment. Obviously, as time moves on, we have to evaluate what we might need to do. But we’re pretty confident that we’ve got the expense line, not only the tranche we’ve announced but, in the event, we would need to do more, we understand kind of where those levers are and what we might need to do. So, I think that’s a big part of how we think about the year. The other thing, we’re going to see working capital come down as the year goes on. So, we’ve got this inventory position, both on hard drives and flash. That’s going to be positive to cash flow as the year progresses. So that’s another big lever for us as we go through the year.
Joe Moore
Great. And can I actually ask about that inventory? It seems like the build that you saw in second half was more drives than NAND. I think the level of both is a little high. Are you going to draw down both of those as you move through the year?
Mike Cordano
Yes, as we move through the year, I see both come down. And certainly, hard drives in particular will come down as we move through the period of deploying enclosure.
Joe Moore
But NAND will come down as well.
Mike Cordano
Yes.
Joe Moore
Some that’s like some people have set that as an objective, but it more recently, it’s been challenging just because there hasn’t been.
Mike Cordano
Well, again, we made a choice to cut production, so I think that’s why our NAND inventory profile is a little different than others.
Joe Moore
Yes. Okay. Good. Anything else from the audience? If not, we can wrap up there. Great. Good.
Mike Cordano
Thanks very much. Appreciate it.
Joe Moore
Thanks, Joe. Appreciate it.