Western Digital Corporation (WDC) Earnings Call Transcript & Summary

I'm Joe Moore, Morgan Stanley's semiconductor analyst. I want to introduce -- I guess, first, I do need to read a quick safe harbor from Morgan Stanley's side. For important disclosures, please see Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. And if you have any questions, please reach out to your Morgan Stanley sales representative. So I'm going to introduce Western Digital, we have, today, Bob Eulau, the Chief Financial Officer; and Siva Sivaram, President Technology and Strategy for Western Digital. So welcome, guys. I know you have a safe harbor agreement to read as well, and then we'll go through any questions.

All right. Well, thanks, Joe. We really appreciate you having us today. And as you said, we've got a quick safe harbor. We will be making forward-looking statements, and I ask that you refer to our SEC filings for the risks associated with these statements. We'll also be making references to non-GAAP financials and a reconciliation of our GAAP and non-GAAP results can be found on our website. So with that, I think we're ready to go.

Great. Thanks. So maybe I can start off with some of the bigger picture things that have been going on. And maybe, Siva, you can give some insight into this since you've seen now a couple of transitions from SanDisk to WD to a CEO transition. But coming up on a year now since the CEO change. And it feels to me like you've reweighted the priorities a little bit. A year ago, you were sort of a hard disk drive company that also had a NAND business. Now the 2 businesses are kind of on an equal strategic, putting an equal focus and a little bit of separation. Is that perception accurate? And then just maybe frame for me a little bit the way you guys see that.

Bob, you want to get started and then I'll continue?

Yes. Sure. I can start. And I've been here about 2 years now. So I saw about a year before the transition, about a year after the transition. I mean my take is both businesses were always extremely important. I mean there are 2 great markets. We have both the hard drive business and the flash business growing at this point. We had previously been organized around end markets. And Dave and the team just really felt we had too many execs worrying about both flash and hard drives. So what we've been able to do with this transition is really get more focus into the company. I think there's a lot more ownership and accountability. The flash and hard drive technologies are very different and require a lot of focus and so I think the product strategy is definitely getting better. The road maps are getting better. There's more intensity on each of those businesses. And so we've made good progress, I would say, in the last few months, in terms of establishing the 2 new business units. And we have great leaders in place on both fronts. They're focusing on getting the strategies right, spending a lot of time in terms of getting deep on the technology and the product development. They've been engaging with customers and so we're really excited about the level of traction we get in each of the areas. So I don't want to go too far. I guess I'll let Siva say if there's anything he wants to add. He's got a little bit longer perspective on all this.

Bob said it right. The whole thing is about focus. And what this has allowed, the 2 leaders we have to focus on their respective businesses. The technologies in the company have always been strong. The company has always been a very strong technology company. Converting it into the products the right way with the right portfolio focus, that's what the programs are able to do and get deliver those products to customers. And again, the synergy in the customers is very, very good, and that comes back together again.

That makes a lot of sense. And I certainly understand the rationale for the business segmentation that you guys talked about. I guess it's led to speculation about whether the 2 businesses, at some point, separate. And it seems like you guys are pretty clearly in the camp of the synergies that were the rationale for the acquisition in the first place, on the go-to-market side, are still the prevalent kind of way of thinking about it. But is that the right way of thinking about it. It's a storage company. You see a lot of benefit to the marriage of the 2 businesses, but you also see the need to focus a little bit differently on the 2 supply chains?

Yes. I mean there's still a lot of synergy on the go-to-market. And I think we've seen that in the past year, just like we saw it in prior years. If you think about it from a client perspective, we've been able to really manage the trend from client hard drives to client SSDs very effectively. And I think we're only able to do it that well because we had both the hard drive and flash businesses. From a data center perspective, we think the technologies are complementary. So on the client side, they're really substitutes. But on the data center side, they're complementary technologies. And so as we're having success with enterprise SSDS, we think that's going to lead to more success on the capacity enterprise hard drives and vice versa. And that's important when you're trying to get in and get qualified at various accounts, those qualification slots are really important. It's also -- we've done the work. I mean we know that of our top 20 customers, all 20 of the customers actually use both technologies. Now some are much bigger in one area versus the other. But there's a lot of synergy in terms of the go-to-market with those customers. And we have a big retail franchise as well, and we're seeing some of the large customers really getting into gaming, getting into VR headsets, wireless devices. I mean there's a there's a lot of synergy in terms of how we go to market. And so we did not want to lose that synergy, but we really want to get the focus on the product side.

Yes. That makes a lot of sense. I guess as the, sort of, most agnostic storage supplier between NAND and hard disk drives, so sort of anyone participates in these markets, can you talk a little bit about how you see the markets coexisting? Are they both going to grow? And what are the areas that you see one growing at the expense of the other?

Yes. Maybe I'll start and then Siva can add his thoughts. But I mean part of -- definitely part of why I came to the company as I see storage as a growth area, and I think it's going to be a growth area for many years to come. And I think we're just at an inflection point on the hard drives, given the capacity enterprise becoming such a high percentage of the total that we're going to see hard drives as the growth business as we go forward. On the flash side, most industry analysts expect 30-plus percent bit growth for a number of years as we go forward. So this is clearly a growth market. And I think we're going to be in -- we are in a really good position to benefit from that growth as we move forward. And I think the new focus we just talked about in terms of the way we've organized the company, I think is going to help us to capture some of those opportunities. I don't know if you want to add anything there, Siva.

No, but you said it correctly, Bob. We are the only approved spectrum storage company in the world. We are the largest approved spectrum storage company. Even we do participate in a wide array of markets on both of these, whether it is in retail, whether it is in client, whether it is mobile, whether it is enterprise, whether it's an edge devices, whether it's growing cloud. Everywhere, there is data that is exploding. And the need for storage is growing when it is past hot storage versus cold archival storage, in that all entire Spectrum data is growing. And we have to have that consistent philosophy of making sure that our big cloud titan customers are able to deploy the right storage technology at the right place, and we are there in every one of those market segments. So this is a good place to be, a rapidly growing market. The technology and the product portfolio are well situated, and we are serving the full breadth of the customer requirements.

Great. Thank you. I think maybe focusing a little bit on the NAND business. And I would start with the joint venture. I noticed you guys in sort of recent presentations have been sort of speaking positively about the JV. And I know this for 20 years, this JV has worked really well. And it gives you the scale that the biggest companies have between the 2 companies. And so it seems really important to me for the JV structure to work. Can you talk a little bit to that? And how smoothly is that working? And how -- is this going to continue to be a successful partnership?

Just as you said, this is a 20-year long marriage. It's been working very, very well. And between us, this is the largest flash wafer fab in the world. We produce close to 40% of the world's flash in these fabs. And the scale that each of us leverage off of each other, both in the manufacturing volume, of course, the fixed costs that can be deferred over the larger volumes, that helps. But more importantly, in the development side that we are able to pull our resources and develop the common die that leads to the technologies that develop, the die is designed and manufactured, allows us to fully exploit the entire market space that we are serving. So we are able to -- each of us contribute only half of the development needs to get the whole full benefit of R&D scaling. So that's been very, very good. Of course, as you know, we compete in the marketplace. We compete openly in the marketplace in the entire spectrum. But on the development side, we are brothers, we are comrades at arms, developing it together. And yes, we went through some difficult times earlier in the decade, but the relationship is very, very good and continuing to go strong.

Okay. Great. Can we talk a little bit about your competitive position in NAND? And maybe I want to talk about BICS6, which you just announced, we can talk a little bit about your BICS5 112 layer technology, how is your relative cost structure inspects kind of versus the competition there?

So let me set it up first, as we said, talked about the JV. The JV provides the scale, scale in manufacturing, scale in R&D, so that, given that in the merchant market, meaning in the non-captive market, we have 40% of the entire flash market. We set the BICS, right? What we do upfront sets the pace in where the technology needs to go. Having said that, back to the technology details. Our high-volume today is in the 96 layer, one of our most successful technologies. We peaked at over 80% of our volume going out of the 96-layer technology, BICS4. We have since introduced and now ramping BICS5, the 112 layer. When the market talks about 128 layers, we are able to do in 112 layers, all of the leads. That talks to the technology advantage that we have. When you add more layers, it adds more cost. You need more capital equipment. You need the latest capital equipment. We are able to accomplish this with a lower number of layers, both in terms of the big growth and in terms of the cost with the lowest possible capital needed. So the 112 layer BICS5 has been a tremendous success, the fastest yield ramp that we have seen in our history. We are ramping very rapidly from the 80-plus percent of bits and BICS4 to now by the second -- later half of this year, we will have the majority going into BICS5, the 112 layer. So that gives us that extremely rapid switch over from one technology to the other without the intensity of the capital needs. So we continue to be the market leaders in cost as we go forward into BICS5 ramp later in the year, where we'll have products across all segments. Currently, we are shipping retail and client products on BICS5. By the time this is done, it would have shipped in every product category that we'd want to switch over to BICS5. And then, of course, BICS6, which is coming up with 162 layers, again, you can see the same trend. When everybody is talking about even more higher layers, we are saying lower layers, still managed to get the cost reduction and competitiveness.

And maybe on BICS6, I mean, there's some interesting specs that I saw in the release, 162 layers, you mentioned, you're moving the CMOS under the array. So you get a pretty big reduction in die size. And I guess if I think about SanDisk as a planar company, it was always very clear that you guys had much smaller die sizes than everyone else. Like as you get into this kind of sixth generation of 3D, where are you going to stand versus Samsung versus Micron, Intel?

Yes. So die size, as you know, is an extraordinarily important metric. But you need to trade-off die size in the process complexity. In order to get to that die size, then you'll have to do unnatural things. When we now try to develop technology, that balance will always be there. The circuits under the array always use sort of one particular density, the smallest die. Only for one particular density, you'll get the smallest die. Otherwise, the array gets big, the CMOS gets big, right? We -- even though we had the technology in-house. I don't know if you remember, we shipped circuits under the array with the original matrix technology in 2005 in volume. So we know how to do circuits under the array. But we did not introduce circuits under the array till the BICS6 generation because that's when the array gets large enough to justify having circuits under the array. And so we want to make sure we are trading off constantly process complexity against die size. And when you come to the BICS6 generation, we see that matching. And this is the whole philosophy of delivering the right technology at the right cost point and performance point for the customer.

Yes. Okay. That's helpful. So you mentioned the lower capital intensity. And I've always been impressed by the amount of technology that you guys extract from relatively low CapEx, even at JV level, and then there's even less corporate CapEx because of the JV structure. How do you think your capital spending has been over the last 12 to 18 months? And does the number need to come higher as industry conditions improve? And obviously, this year, you had some supply growth from some of the negatives to supply last year. How much capital spending do you need going forward to stay on the curve that you want to be on?

Bob, do you want to take that?

Yes. So thanks for your comments, Joe. I mean we do believe that, as Siva has said, our designs have been very capital-efficient for a long period of time. And so -- and BICS5, in particular, is very capital-efficient node. We'll probably have to invest a bit more in BICS6, but that's a ways out. So I'd say to answer your question, last 12 to 18 months, I think we've been very efficient from a CapEx standpoint. As we've talked about, our goals in terms of CapEx start to be in total around 6% to 8%, our cash CapEx to be 6% to 8% on total sales. And that implies 8% to 10% on the flash side and 4% to 6% on the HDD side. We're a little higher right now on hard drives, but that will, I think, come back in line over time. And one thing that's really important to continue to keep in mind, when we're talking about capital, is we also have about $1 billion in depreciation in the JV that we're able to take advantage of. So within the JV, we fund $1 billion of depreciation every year, and that is then used to purchase capital equipment as well. And then we also do a fair amount of leasing in the JV for equipment. And then if there's any delta between the leasing and what the depreciation funded, then we'll actually make a loan to the JV. And if it's the other way around, we'll obviously get cash back from the JV. So I think we've got a very efficient CapEx structure from a flash standpoint. And I think we've done a nice job in terms of transitioning resources over time on the hard drive side. We're kind of to the end of that, where a client -- there's not -- there aren't a lot of assets on the client side, we're going to be able to shift to capacity enterprise as we move forward. So I think that will change as we go over the next couple of years.

Yes. And I think it's one of the elements that's maybe underappreciated is your cash flow and EPS are actually a lot closer to each other than they are for other memory suppliers. Maybe we talk a little bit about NAND economics. Your gross margins were up in December, which we didn't expect. You guided them to be up slightly again in March, and this is a NAND comment. And I know you've had some tailwinds from start-up expenses rolling off. But do you think we're past the bottom on NAND gross margin for this particular cycle, you're willing to kind of make that statement? And then generally, there have been some positive comments from your customers about pricing that sort of stabilizing, not going crazy, nothing untoward happening, but maybe where customers that we thought there would be price declines, it looks a little bit more stable things like that. Can you just talk a little bit to your general gross margin progression by quarter?

Sure. Yes. And I'd say, in general, I agree with your comments. We focus primarily one quarter at a time, as you know, and we did guide the gross margins up a bit if you looked at our overall comments after last earnings call. So I'd say in terms of the second quarter, flash pricing was basically in line with what we had expected. We continue to see very good cost improvements. The start-up costs came in better than we anticipated. I think they were around $41 million, and we had expected around $50 million. So there's a little bit of a benefit there. Fiscal Q3, which is the March quarter, usually a fairly soft quarter because it's a weaker quarter from a retail standpoint. But we're still seeing, as I mentioned, fairly good trends, good cost reductions in the third quarter. And it's -- we'll see how it plays out, as I mentioned before, on the OEM markets. And we've been seeing really since before the calendar year started, we expect calendar year to be -- calendar year '21 to be a pretty positive year. And I still think it's looking that way. So I always hate to call a bottom. But I guess if you look at the arithmetic, you would say that the December quarter was probably near the bottom.

Yes. Right. That sounds good. And then just one last NAND question. I mean if we go back over really a generation, we've seen an oscillation of gross margins, you sort of bottoming out in the 20s for -- and this goes back to the SanDisk days. NAND gross margins bottoming out in the 20s. Sort of generally peaking in the 50s when times are good. Do you think that's a general construct, I'm not asking for a prediction, but like is that still the right range to kind of think about, about where your gross margins can go over the course of a business cycle and maybe averaging somewhere in the middle?

Yes. I guess the way we've been thinking about it and talking about it, we need to prove this theory is we think in the 3D era because the flash is more capital-intensive that the cycles will probably be a bit more dampened so our theory is that the lows won't be quite as low as they were before and at the highs won't be quite as high as they were in the past. So we'll have to go a few more years to prove out that theory, but that's kind of the way we're thinking about it. So yes, I think we're -- obviously, we got slowed down a little bit by the pandemic this past year. We were starting to make progress forward prior to the pandemic. I'm hopeful that now we'll see progress as we move through this calendar year, as we talked about.

Great. All right. And then maybe shifting to drives a little bit. The transition in the nearline business to 18-terabyte has sort of gone well from certain standpoints that you hit your milestones on 18, but you've seen some pressure in the 16 that have maybe pressured margins a little bit. Can you just talk generally about what you've been seeing in hard disk drives, particularly on the margin side?

Yes, happy to do that. And just going back in time, the 14-terabyte dry was a fantastic product for us, it really allowed us to get in a market leadership position, sustain it for quite a while. Unfortunately, we ended up being laid on the 16-terabyte drive. So by the time we came to market, the gross margins were already under pressure on 16-terabyte drive. And in the first half of this year, we're shipping more 16 terabytes than we are 18 terabytes. We are starting to ship 18-terabyte drives in the March quarter. So as we go through the year, we think margins will improve because the structure is better on the 18-terabyte drives. But the year is still going to be a mix of 14s, 16s, and 18s, and different customers are going to adopt the 18-terabyte drives at different points in time. We do have the energy-assisted drives cloud at 3 of the top 4 cloud titans. We feel good about that. And we're seeing more and more adoption with other customers as well. So I think that we'll get back into a leadership position with the 18-terabyte drives. And as we do that, we're expecting to have a gross margin tailwind, and we should be in a much better place than we have been in the last couple of quarters.

Yes. I mean between the COVID cost issues and some of the things you just referenced, the -- our drive margins have been in the mid-20s, is still the right model there? You think about that as a 30%-plus gross margin business?

Yes. We think it needs to be north of 30%. Over time, to meet the demand requirements, there's going to have to be CapEx investment in this market as well. So I think it's going to be a -- pricing eventually will be a function of supply and demand in this market, not unlike it is in the flash market.

Yes. Okay. And then in terms of the longer-term road maps there, can you talk about the sort of HAMR versus MAMR debate? Can you talk a little bit about WD's positioning there?

Yes. So I will take this, Bob. Pardon the pun, but HAMR is just one more tool in our toolbox. I just want to make sure we keep that in perspective that we see a clear line of sight for us into the mid-20 terabytes without having to involve HAMR in our own energy-enhanced road map that is there. There's some art to go with it and adding a number of disks is still to go with it. We are still continuing to increase the aerial density with the energy-assisted work that we're already doing. Yes, we will get to use HAMR over time when it is the right time. And we expect that to be in the middle of the decade than in the early part of the decade. And -- but in the meantime, the product progression is very, very well set with the road map as to what's going to happen with the 18 to 20 to 22, et cetera. So that path, we have a lot of confidence in, we will continue to be the aerial density leaders because of that. So you need to think of it not just as one digital change from conventional PMR to HAMR, you got to think of it as many such transitions that are all in the past, whether it is going from 8 to 9 to 10 disks, whether it is going from conventional to [indiscernible] from CMR to SMR and actuators going from triple stage and better, more nuanced firmwave that is going to manage all that. And of course, HAMR eventually. So that's the path we see well into the decade. So the road map is very, very strong through the capacity enterprise technology growth.

Great. So there's a couple of questions from the webcast that I wanted to relay. One of them says I drink too much coffee, which is -- I didn't know my wife was listening, but, just kidding. The -- in terms of the BICS6 transition, you mentioned being a little bit more capital-intensive than BICS5, which was really sort of very, very capital-efficient technology. How much do you think that changes the capital spending profile?

So I'll start off and then Bob, you can continue on. The way you modulate it, along with the increased capital spending also comes higher bit growth. So on the same way, BICS6 has a higher bit growth than does BICS5 because of the circuits and the [indiscernible] small die size. So for a given bit growth rate, we modulate CapEx accordingly, CapEx spending accordingly, even though per lot or per unit, you need to spend more for getting the bit growth that we can moderate the capital spend.

Great. And then another question in terms of pricing being a little bit better, at least than I had anticipated. I'm not trying to put words in your mouth. But with NAND pricing being a little bit better than what industry pundits had sort of thought, do you think that's a demand factor? Do you think that there are supply issues. I mean it seems like capital spending has been high for the last few months. Do you think -- just do you see -- and you've been describing better pricing really going back a few months starting in the consumer business. But do you see that as a demand versus a supply issue?

Well, I mean, our view is that demand has been there. We've been expecting '21 to be a very good demand year for quite a while, and I think that, that's going to play out the way we anticipated. From a supply standpoint, we've said for quite a while, we think most of the industry participants behave pretty rationally and in the 3D era, it's very, very capital-intensive. And so I don't think there's a big surprise here. I think most of us are trying to preserve our market share, and it's a fast-growing market.

Yes. Okay. And then another question from the website, which I probably don't completely understand, but you guys will. CD outlined a new storage as a service product to enable and process all data created. Is WD looking at doing anything like that? And what's your perspective on that market?

We always had a platform business within the company, and that continues to do well. But we have a diverse product portfolio. We will service all market segments as needed. We are not coming in and making a separate big focus the way you heard from our competitor. We are not trying to do something that -- we are not planning on competing with our customer.

Okay. So maybe 2 minutes left. I'll finish with just a financial question. Bob, in terms of the cash flow and the dividend, I feel like you guys still get a lot of flak for having suspended the dividend, which seemed to me like the right thing to do. We actually -- had previewed in that quarter, we thought you should do that. So I feel a little bit responsible that people are negative on. But it seems like it was the right decision to make. Can you just talk a little bit about where you are with the deleveraging process? And where you might see those cash flows starting to get returned to shareholders? What are the metrics you're looking forward to get to that?

Yes. So thanks for your comments, Joe. It wasn't a decision we took lightly. We deliberated on it a long time. And I think now as we look back, it was clearly the right answer given everything that transpired in the pandemic. Our priority is to reinvest in the business. And we've got 2 growth businesses. They both require capital, and we've got to make sure that we fund them for leadership before we do anything else. And then our next priority is to delever the company. I think we've been really clear on our goals to get down to $6 billion in gross debt and $3 billion in net debt. That implies gross leverage that we think is very reasonable through a cycle. So we modeled the best EBITDA, the 4 best EBITDA quarters, which turned out to be FY '18, and then we modeled the 4 worst EBITDA quarters in the last trough. And if you look at that, you look at our debt goals, you'll end up with gross leverage anywhere between 1 and 3.5 as we move through the cycle. And we think that's very healthy. So we'll make as much progress as we can as fast as we can. Nobody is more motivated than us to get in that range.

Yes. Okay. Good. Well, we'll wrap it up there. Thank you very much, guys, for your time. And talk to you soon. Thank you, everyone.

Thanks. We appreciate it.

We appreciate it. Thank you all.

Thanks.