Western Digital Corporation (WDC) Earnings Call Transcript & Summary

Good morning. Thank you very much. I'm Tim Arcuri. I'm the semiconductor and semi-equipment analyst at UBS. And we're very pleased with this session. We have Western Digital. Very, very pleased to have David Goeckeler, the CEO; and Bob Eulau, the CFO. And before I start, I'll turn it over to you, Bob, to read the safe harbor.

All right. Thanks, Tim. We will be making forward-looking statements, and I ask you to 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 thanks. I'll turn it back to you.

Perfect. Thank you. Thank you, Bob.

So I have questions sort of in 3 different categories. First on the HDD business, second on the NAND business, and then I have some general strategic and stock questions and things like that. So first, David, let me ask you just about the HDD business. You've really done a great job on the AT&T side until just this past quarter. You're kind of under-shipping your competitor on the HDD side in the calendar fourth quarter. Some of this is due to challenging comps, but there's other factors, too. I think there's some 16 [ key ] inventory in China. You saw a pushout that was fairly unique to you. Can you just kind of run through again what the current state of the market is? And sort of where you are in Q4 and heading into Q1?

Sure. Happy to, Tim. So first of all, thanks for being here. We appreciate the time and to talk about the business. We love to talk about our business. So we appreciate you hosting us. So HDD business, I mean, let's start from the big picture, and we'll zoom down into our particular portfolio and then into maybe quarter-by-quarter. So first of all, the business, as we all know, very different than what people think it is. I mean the -- what they have thought of it as, I mean, the industry has been going through a transition over a decade now, which we're through, which is this is HDDs are the foundation of the cloud. It's just as simple as that. Over 70% of our capacity is shipped into the cloud. A lot of the manufacturer -- almost all the manufacturing capability now is going into the cloud. So this has been a big transition over a decade. And now we're seeing the fundamental dynamics of the business start to change, which we think is going to continue. So on a demand perspective, we look into '22 just as we look into '21, and we were hearing very good things for our customers on cloud demand. We continue to hear good things in '22, especially as we go throughout the year. We expect to see north of 30% exabyte growth in the cloud. So anytime you have a technology portfolio that's proxy to cloud growth, I think that's a good place to be, and that's where we are in the HDD business. Now you take a step-down into our portfolio, and we made a lot of progress over the last year. I mean I appreciate your comments on '18. We feel really good about where it's at. I mean, a year ago, we were sitting here and having these conversations with the business where we still had to prove the AT&T technology. We had a lot of confidence in it, but we had to prove it to the market. We were commercializing energy assist. We got through that. We ramped it. It's now our main capacity point. A year ago, we were also talking about a business with gross margins in the mid-20s. Now we're in the low 30s. So just a lot of progress. Share has renormalized in the industry around '18. So we feel very good about getting back on our front foot with the product. We're also set up very well into '20, which we recently announced, and into '22 and beyond all the way, all the way up to '30, where we'll see HAMR take off. Now on quarter by quarter, we have, coming into the current quarter, just a tremendous amount of progress in the portfolio. This quarter, we have a few idiosyncratic things going on. We have a big customer that pushed out some demand. We've got a little bit of excess inventory in China that the industry is working through. We have our own supply chain issues. All this stuff will normalize as we go forward over the next several quarters. And we'll get back to that big picture of this is a secular growth story. We have a great road map. We just announced OptiNAND, which is the next step on that. So we feel really good about where the portfolio is, the technology, the value proposition we're offering our customers and just the setup of the whole market.

Great. Can I just -- as we move into margin, question that I always get is sort of how to normalize for some of these factors as you look into March. I know, obviously, the push out, that's not perishable business, so that's going to come back. And some of your supply chain issues, those will be resolved, maybe not in March totally, but at some point, they will be resolved. So I'm just wondering if you can kind of maybe handicap us on the moving parts as you think about what is perishable and what's not. And I guess in that vein, it seems like a normal seasonal march is sort of just basically down its niche is the sort of normal seasonal march. So I'm just kind of wondering whether you can go through each of the pieces and say, well, that's perishable, and that's not and sort of help us handicap what the right baseline is for December into margin and as we try to model out 2022.

Yes. I'll make a few comments, and Bob, I'm sure, will have a few comments. So first of all, we don't see a whole lot in the quarter that's perishable per se. I mean, I think, as we look into next year, Q1, we're still working our way through the pandemic. So what is seasonal is a little bit tricky given all of that. However, I think we're going to get back to close to what seasonal looks like traditionally our business. And Q1 is a little bit of a down quarter for us, especially coming off -- and we have a big consumer business as part of the portfolio and coming off of a strong holiday season. That's one area where we see that business down a little bit off a strong calendar Q4. Supply chain is a big issue. It's something we continue to face. We work through week-to-week, month-to-month, making sure we have enough supply. Whatever supply we get, we remix the portfolio, so we get the best result. And then on the customer-by-customer basis, you're right, that will start to come back. That will normalize over the course of a quarter or two. So you add it all up, I would say we'll probably be, I think, get back to that seasonality trend as we go into Q1 is a decent way to think about it. And we've got our eye on -- we still have our eye on COVID in the supply chain. So those things can change things quickly as well.

Yes, I don't have a lot to add. I mean we're very positive on 2022 for the year. I think it's right. We're probably headed back towards more normal seasonality. We still have supply chain challenges. We're pretty tight in terms of supply this quarter. I expect that will be the case in March. The controller chips for both the hard drives and on the flash side are going to be an ongoing challenge. But I think as we move through the year, we'll definitely get stronger.

Great. I wanted to talk next about the cost road map in HDD. And there sort of continues to be this perception when I talk to people that the business is a fairly rapidly melting ice cube. But I think in reality, for the mass capacity applications where the growth is, the price delta is still like 8 to 10x. And even if you could get to parity, the cost declines you can get on the HDD side are pretty similar, it seems like, to the NAND side. And so can you just given how fast the storage requirements are growing in the cloud, it would take so many decades to spend enough money in NAND to possibly replicate all of that bit capacity. So can you just talk about that? There seems sort of a perception of the drive business versus the reality and maybe what you see as the cost road map in HDD versus in NAND.

Yes. So I think you got it right. I mean drive industry is very different than it was several years ago. I mean it's an industry that's been going through a major transition. I mean a lot of people think of the drive business as a client business. And that's all been replaced by NAND. There's no doubt that was a replacement situation. You used to have a hard drive in your laptop. Now you have a client SSD. That's been fantastic. We played that transition well as a company. But we've had the rise of the cloud. And now the drive business is simply the foundation of the cloud. It's storage in the cloud, huge percent, 90% of the storage in the cloud is all done on hard drives. The cloud is going to continue to grow. I think that's something probably most people would agree to that if you have a technology franchise, having it correlated to cloud growth is a good place to be, and that's where we are in the drive business. So we feel really good about it. That's why I started out talking about the structure of the market and drives. It's about making sure that we can deliver high capacity drives to the cloud and continue this fuel data storage, the mass data storage and things like energy assist, OptiNAND, these are all things that give us a very, very robust road map to continue to bring a strong value proposition to our customers. Now the substitution question in the cloud is a marginal one. It's not like it was in the client. Two things can be true. Two things can grow. And so the enterprise SSD business is going to be a great business. I mean we've worked very hard on that. We can talk about that a little later, maybe. That is a very large and growing -- quickly growing TAM. Again, storage in the data center is a very good business. However, the HDD business is also growing. Maybe another ice analogy I'll give you is the iceberg, not necessarily the ice cube, in that there is enormous amounts of data that is not stored today because it's not economically viable to store it. Now our customers talk to us all the time about if we can store more data, that's a good thing. I think people realize in a modern world with AI/ML, the more data you have, you can do something with that data. You can monetize it. You can figure out how to deliver to your customer better service, whatever it happens to be. So there's kind of an insatiable demand from our customers to store more data. At the same time, the new data we're creating all the time just continues to go up. So although enterprise SSDs are going to take more of the data, there's going to be more hot and warm data to store in the data center. That's fantastic. We're going to play that trend, but there's also an enormous amount more of cold data to store in the cloud as well. And that's an ever-increasing pie. So we have 2 markets. We're able to play in both of them. And from an ACD perspective, it's a fantastic business. It's great to be the foundation of the cloud. It's going to continue to grow. And finally, on your point on cost, that's why innovation is so important. This is -- we're a technology business. It's driven by innovation. It's one of the things I feel really good about the changes we've made in the last 1.5 years, drive more innovation. And we do continue to drive cost downs in HDD. And so that just means we'll have more opportunity to store more data. So we've got 2 great markets in the data center for storage, one in enterprise SSD, one in HDD. They're both going to grow. They're both great businesses to be in, and we have the opportunity to participate in both.

And you would agree roughly that the cost road map in HDD is roughly comparable to what it is in NAND roughly mid-teens?

Yes. It's in that teens level, right? So it's at a point where you're not going to see crossover for any foreseeable planning horizon. I mean you're talking of a decade or more, we've done the analysis out for at least a decade. There's no crossover in costs. And so there's a long and robust road map for HDD storage in the cloud. Like I said, I'm pretty sure the cloud is going to continue to grow. That's what the business is today. HDD is simply a cloud business.

Got it. And then, David, you didn't talk about OptiNAND. Maybe you can talk a little more about that? And that's a pretty important new technology. So can you just sort of help investors place that?

Yes. So I think one way to think about it is what we're trying to do is get the most density we can on each platter we put into the hard drive. So we're constantly looking at technologies to increase aerial density. And OptiNAND is another step in that direction. It has a number of other benefits around reliability that we could talk about as well, but let's just center on the amount of data we can store on 1 platter and helps us get to 2.2 terabits per platter. That was a big step forward. Again, we're just coming off of commercializing energy assist, where like we talked about at the beginning a year ago, we were sitting here having to prove that in the market. We've got that proven. Now we have the next step of OptiNAND, which takes us to 2.2 terabytes per platter, which means on a 9-platter drive, we can get 18 terabytes, which we just -- which we have in the market. And in fact, in 20 -- more importantly, in a 20-terabyte drive, we can get it on a 9-platter. So we still have the ability to add the 10th and go even higher. So it's another technology. It's another step in the path that gets us to this gap between '20 and '30. And '30 is where we expect HAMR to take over and then yet another cost curve will inflect and take us higher. So OptiNAND is a very, very important technology. It drives another step in the road map on aerial density. It gives our customers just a better value proposition so they can store more data at a lower cost.

Got it. Got it. Maybe we can shift over to the NAND business for just a couple of minutes. And I guess BiCS4 96-layer was a super-efficient process for you, very, very good yields. BiCS5, you're now crossing over this quarter. But as you shift to CMOS under the array for BiCS6 and the capital costs go up a bit, how do you kind of handicap the cost curve as you're getting the BiCS5 crossover this quarter, how will BiCS6 be different than BiCS5? And sort of what are the milestones that we should be looking for on BiCS6?

I think it's really important to understand what part of our R&D philosophy in the entire BiCS road map. And of course, this is in collaboration with Kioxia, our JV partner. We've always had great technology. Very focused on the productivity per node per dollar of capital. So more layers equals more capital. So how do we get the density we need with the least amount of capital possible. And I think if you look at our road map over the years, you'll see that we've consistently delivered the most productivity with the least amount of capital. Now BiCS5 was designed as a very capital-efficient node. In fact, we didn't need to transition much of the tooling in the fab, which allows us to get through the transition quite quickly. BiCS6 is a more fundamental design change, but it will be a more -- it will also deliver more productivity per wafer. So we still have that focus there of delivering the most amount of bits per capital dollar invested. That's a very explicit goal of the road map. And that's why you see our layer count's a little different for other people. We're focused on other things as well, the die size. We have a lot of very good technology on memory hold density. So we're able to do -- we're able to get the premium product with the densities and the performance we need in a more efficient package. And I think that's just a reflection of how strong the JV is with Kioxia. And between the two of us, we're the largest provider of merchant bits in the world, so we have a lot to invest in our road map, and we think that pays off.

One thing that I keep getting from investors and I just wanted you to address this, is people keep saying, well, Kioxia and WD are generally more focused on brownfield versus on greenfield, say, for some of the Koreans are focused more on greenfield and they have much higher capital cost. If you look at their WFE per installed wafer, it's much, much higher than yours. So how do you sort of handicap the capital efficiency against longer-term share goals? Can you really -- do you care about market share? Is that the priority? Or if you began to lose market share, would you shift a bit more to -- from brownfield over to greenfield...

When you say brownfield and greenfield, you mean on new fabs? Is that the way you're thinking of brownfield and greenfield?

Yes. Yes, I just mean you're focused on basically layer migration and tool reuse, whereas they don't focus as much on that, which are why their capital costs are higher and their layer counts are higher.

Yes. So look, our goal is very simple. It's to maintain share. We're going to invest to maintain share. We don't necessarily want to lose our gain share, but we're going to invest to maintain share. But we're going to do that in the most capital-efficient way. I think that's just a smart thing to do. We build our fair share of fabs as well. We're building [ Light 7 ] right now. We just don't want to always use that. We want to be as capital efficient as possible. Again, this is a very specific goal. If you talk to the R&D team, they're working on many, many versions of BiCS well into the future, and they have a very specific goal of how do we make this as capital efficient as possible. We think that's the foundation of the business. I mean the foundation of the NAND business is the technology. And having the best technology and having a really good cost profile is extremely important. It's an explicit goal of ours. And I think, as I said, we've been very successful there with our partner, Kioxia. But we'll build new fabs as we need them to make sure that we can get the output we need to maintain share. We're not going to get it right every single quarter. And I think if you look at all of the players in the industry, they go -- even on a year-to-year basis, they go up and down, just because of the nodal transitions and the capital investments. But over time, we've been successful with that strategy, and we expect to continue it.

Got it. I wanted to ask you about consolidation in the NAND industry. What is your view on that? And I guess, specifically, do you see yourself as a potential consolidator? And maybe in that vein, can you talk about the situation with your JV partner?

Well, I think any kind of consolidation in the industry is always welcome. I mean I think we see sign -- Hynix and Intel coming together now, which is, I think, a step of consolidation, which I think is welcomed by everyone. I think we've seen pretty rational CapEx investment over the last year, 1.5 years in the industry as well. So fewer people making CapEx decisions. Now us and Kioxia, we're -- from a production point of view, we are consolidated in the sense that we have the same fabs. We co-invest in those. We have the same road map. It puts us in a very, very, very strong position. Additional consolidation would always be welcomed. It's a -- the nodal transitions are getting more expensive, more capital intensive. There's additional synergies to be found through consolidation. I think WD, in particular, has been a consolidator in the HDD business, and that's worked out well for us. In the NAND business, we're not in a position right now, the best position to do that. I mean we're kind of focused on rebuilding big parts of the company and getting our execution right. We're getting our balance sheet in shape. We've had some good progress there just within the last week. We'll always entertain listening and understanding where the industry is going. But right now, we're focused on making sure that we get our company in the best shape possible. And I guess the final thing I'll say is it's a complicated chessboard in the NAND industry. There's a lot of geopolitical issues. It's not something that's super straightforward. And so we're going to stay close to that, but we're also going to be very, very focused on what's the right thing for our shareholders. If it's not the right thing for our shareholders, then we're not going to participate in it, whether we drive it or not.

Great. Maybe we can move on to some bigger picture sort of strategic questions, if you will. And I know, David, you're very execution-driven given your history at Cisco. I think you've replaced over half of the Board and most of the exec team since you've been there. So my question is, how do you sort of weigh your key priorities and the benchmark that will determine your success?

Well, I mean, the final benchmark is shareholder value. I mean that's what we're driving for. But to get there, we've got -- we had to make a lot of changes in the foundation of the company. And brick by brick, we're making those foundational changes. Lot of change in the management team, change in the management structure. We went through a BU -- cycled BU structure with 2 very capable general managers. That's led to more innovation. We've seen that over the last year. We just talked about it in HDD. We see the same thing in the NAND business, whether enterprise SSDs being qualified in more places. We brought more diverse set of skills on to the Board of Directors as well. So we're making fundamental changes in the company. We have made fundamental changes in the company. It's working as far as getting the results. Since I came in Q3 of '20 to last quarter, revenues up 21%; gross margin, we've added 6 points of gross margin from 27.9 to 33.9; EPS has nearly tripled from $0.85 to $2.49 in the most recent quarter. And we made a commitment to our balance sheet. We paid down $2.1 billion worth of debt. Our debt-to-EBITDA ratio has gone from 5 to under 2.5. Our gross debt is down to $7.7 billion. And then just this last week, we achieved investment-grade for the company. So we believe we made a lot of progress. We need to continue to execute, continue to stay focused on that and get that to translate into shareholder returns.

I guess just on that front, Bob, maybe can you just talk about the leverage ratio targets and talk about how you think about shoring up the balance sheet? And maybe kind of walk us through what your priorities are in the near term and in the longer term?

Yes. So as Dave said, I mean, we made a big commitment to delever the company and improve our balance sheet. We made a difficult decision 2 Aprils ago to suspend our dividend. And since that point in time, we paid down $2.1 billion in debt. We derived our goals really by looking at both peak EBITDA and trough EBITDA. And in the trough, we determined that for a company like ours, we really did not want our gross leverage going above 3.5. And our trough EBITDA was $1.7 billion. So that's how we derived our goal of getting to $6 billion in gross debt. In terms of capital allocation, we continue to have the #1 priority being to invest in the business. We invest in gross CapEx, around $3 billion a year. We invest over $2 billion a year in R&D. It's absolutely fundamental that we invest in technology to have the road map right, and that's always going to be the first priority. Second priority as of -- when we suspended a dividend was to -- is to delever the company. And once we get to our goals, then we'll look at how we return capital to shareholders. And at that point in time, everything will be on the table, looking at dividends, looking at share repurchase. And I think all of this is enabled by the cash flow generation. And we're really pleased by the cash flow generation in the last 2 quarters, over $1 billion in the last 2 quarters. So I think we're on a good track in terms of improving the capital structure of the company, and that will help enable us as we move forward.

And Bob, how do you -- do you have any specific targets in terms of what the threshold is from a leverage point of view or from a debt point of view that we would then start to talk about more capital return?

Yes. Again, the goal is to get around $6 billion in gross debt. And at that point, we'll look at what the alternatives are. And we're making good progress from where we started.

Great. I think maybe just on that front, David, I think on paper, people would have treated it makes some sense to have the NAND business with the HDD business, given the demand from the cloud. But you haven't really been able to, I think, drive enough profitability beyond the NAND side. I think that you would like to have better profitability through the cycle on the NAND side. Hopefully, that's a fair statement. And I guess when you look at the NAND business, what are the benchmarks you're using to determine success in NAND? Is it sort of cross-cycle metrics? Because obviously, you can kind of tune things up and down. You can tune things up to get better profitability in the peak, but you might sacrifice margins over the cycle. So can you just talk about that? And maybe is it a cross-cycle number? Is it market share? What are the key metrics?

So let's start with what we talked a little bit about before, which is the technology. So having the best technology road map, working with our partner in Kioxia, making sure we drive BiCS road map forward and that we have the fundamental cost reductions we need in the portfolio. And we're very comfortable with the 15% year-over-year cost reductions. We've overperformed that for a few quarters, but we're very comfortable with 15% going forward. So that's where it starts. You got to have a great cost position, and you got to have a great technology base. You got to have the right product that can be qualified at all of the premier vendors, and we are qualified at all the premier vendors. And then it's about running a portfolio that is balanced across all of our routes to market. I thought when I came to Western Digital, one of our biggest advantages is we -- is our customer base. I mean, we have a tremendous customer base, everything from a multibillion-dollar consumer business, where we sell all around the world in every country in the world and in every e-commerce platform in the world, to selling to the largest hyperscalers that are out there. And then everything in between. We have an at-scale channel business. We work with every OEM out there. And so we just have tremendous routes to market and tremendous customer relationships. It's about them having a portfolio that can play all across that. So we have a very strong position in client SSD. I think that's a -- something that's happened post-SanDisk and is a manifestation of the synergies between the portfolios coming along. Mobile, the consumer businesses I talked about, in IoT, a very ascendant gaming portfolio. And then one of the biggest prizes is enterprise SSD. And I think if I look back to your question, we just got over the hurdle in this calendar year. It was just walking into the calendar year this time last year where we got qualified at our first cloud titan with our enterprise SSD product. We've seen that ramp throughout the year, and we've been focused on additional qualifications throughout the year. We're now qualified at 3 cloud titans and enterprise SSD, making good progress with the big OEM. So that's a big piece of synergy of the portfolio that needs to come through. So when I look at the -- what are we looking to do in the NAND business, back to your question. One is make sure we've got the right technology foundation under us. That's why the JV with Kioxia is just so, so important because the two of us together invest as much or more as anybody in the industry to make sure we have the right foundational technology. On top of that, because we have broad access to customers, make sure we run a portfolio where we can get the best out of it and we have all of those routes to market covered. And we've just got -- as I said, client has been a big pillar of the portfolio. Mobile has been a big pillar of the portfolio. Consumer has been a big pillar of the portfolio. And now enterprise SSD is an emerging, going to a major pillar of the portfolio. So I think the idea is to generate good through-cycle margin across all of those routes to market and -- as opposed to riding one or the other just up and down all the time, trying to figure out how we can mix our portfolio across all those route to markets to get the best through-cycle profitability and the ability to generate cash to do all the things we need to do to, right now, pay down debt and then get into a shareholder return policy.

So it sounds like when you're looking at the metrics and your directive to those running that business, it's really more of a through-cycle margin metric or more of a through-cycle return on invested capital metric. Is that right?

Yes. We're looking at through -- we're looking at -- I mean, it's a very complicated business, as you know, because there's lots of stuff you could do with the wafer when it comes out of the fab from selling the wafers to building systems. Each of those requires a different amount of investment and each of them has a different return and each of them has a different level of volatility. And each of them has a different TAM and TAM growth rate. And so it's, what Rob's job is as a General Manager of this business. This is why I believe going to the BU structure was so important, because you got to get somebody thinking about the profitability of that business every single day. And if we're going to invest OpEx in something, am I going to get enough return from that OpEx to justify that investment? And so getting that portfolio mix right on a consistent basis is extremely difficult, and it's something -- but it's also a huge, huge opportunity, and of course, in getting all the execution behind that. And I think over the last year, as we've gone to this new structure inside the company, we're getting much more predictable results out of the business.

Got it. To that point, when you bought SanDisk, they didn't really have much of a client SSD presence. And as you said, now you're really now #2 in that market. And can you talk about what drove the success there? And sort of what tasks have you given to the sales team as end of BUs, which you just talked about to sort of increase your presence in profitability in other product categories?

I think this is an example -- I mean I wasn't here at the time, but this is an example of we had tremendous customer relationships, which leads to really understanding what it takes to build a client product. We're essentially playing the substitution game. We had the hard drive that was in the client. We're substituting that with a client SSD. You have tremendous customer relationships. You know exactly what the requirements are. You understand the demand signals very well of how that transition is going to play out, in which parts of the market, on which SKUs. That gets reflected back into your engineering teams. So I think that the fact that we have both of these portfolios together, that's what it leads to when you have the right product on the client side, and we've got the right client -- the right product on the client side, and we see our share in our position in the market increase over time. I think we have that same opportunity in the enterprise SSD market. We just haven't had the product. Now we have the product. And we're talking about the cloud, where we have relationships with those customers every day because of HDDs. Again, going back to 90% of the storage in the cloud is on HDDs. So we have tremendous relationships with those customers. We understand their architectures. We understand what they're trying to achieve and the ability to have that conversation now across enterprise SSD and HDD. Again, the big, I think, insight here is these are now complementary technologies. They're not substitutes. Whereas in the client, it was a substitute. These are complementary technologies. They're two very important TAMs. They're both going to grow. And so our ability to participate in both of them is quite exciting.

Great. I just wanted to ask you sort of -- this is a big picture question really, but the HDD business reminds me a little bit of the DRAM business from 4 to 5 years ago and how we've seen higher highs and higher lows there in terms of the gross margin profile. And your -- you've -- now we've replaced most of [ 4T ] capacity with AT&T, and you're now on the cusp of a CapEx cycle, not just you, but both of you are, the whole industry is. And it seems to me like that can start to drive LTAs that are both longer term and then have a little more teeth. So can you just talk about the changing dynamics around the agreement that you have with your customers? And maybe what the discussions are with your customers?

Yes. And it is a big change. Again, I think the way you handicap the market is right on the money. It's a big change just since I've been here in the last 6 quarters. Remember when I came, it was very much a transactional business. Every quarter, there was an RFP from every customer. It was bid on by the industry participants, and the customer would make a share decision and then that's how much we would ship. That was a world where -- we came from a world where hard drives were plentiful, and we're going to a world where we're going to have to invest to have hard drives. So hard drives are going to be scarcer. So it's not surprising to see the things happening that are happening. One is gross margins are moving positive. It's an industry where we do invest a lot. We have a very robust technology road map. We think as we continue to bring a good TCO, total cost of ownership proposal to our customers with each successive generation of technology, we can have a more robust conversation about profitability of that. It's a win for both of us. They're able to store data cheaper. We can continue to drive innovation to make sure that happens for the next decade going forward. But also, you're seeing this dynamic, which, again, is, I think, somewhat predictable, but we had to force it to happen, which was -- because you have to change the whole dynamic with the customer, which is, "Hey, we both need to look a little further down the runway here as far as how much -- what is demand, because we now have to invest." I mean, Bob and I are making decisions about how much we're going to invest in head wafer fab expansion over the next couple of years. And that's going to be how many -- what percent exabyte growth do we have -- do we expect? When you're only looking at that market with your customer 1 quarter at a time, that's a more difficult conversation. So we've moved that now into a multi-quarter discussion. And if I look back, let's say, 9 to 12 months ago, it was just maybe with 1 or 2 customers. Now that's expanding to a little bit larger pool than that. And the length of the conversation, it's going a little longer, whereas it was 2 quarters, now we're talking 3, 4 quarters. So the dynamics of the industry are changing right before our eyes, and it's going to continue to go in that direction because we're now -- hard drives are secular growth. The cloud is going to grow. This is what's going to provide all the storage in the cloud, and it's not surprising we're having those conversations, and we'll continue to have those conversations.

Can you just talk about those agreements? And maybe just give us an example of sort of how they work. Are there any examples where you have true LTAs that have teeth where you have a customer that says, "I'm going to take x exabytes and I'm going to pay y for those exabytes. And if I take anything above that, I'll pay z." I mean is that how these LTAs are working? And is it such that if they didn't take what they said they would, that there's teeth to the agreement where you could say, well, Mr. Customer, you didn't take what you said you would, and so you have to pay us something.

Yes. So I think your x and y were pretty much accurate there. I think one thing to remember is these are customers where we have very, very long and deep relationships. They go back many, many -- more than a decade, way more than a decade on capacity enterprise drives. I mean people in the industry still talk about what happened during the floods in Thailand and how the industry responded. And a lot of the same people are still there having a conversation about the relationship. So nobody has an interest in doing something where they make a commitment to their partner and then not fulfill that commitment. So I think that's less of an issue. I think the agreements are around, hey, for the next 3 quarters, this is the number of drives we're going to consume. We're starting to put pricing agreements into those as well to get more visibility into what that demand is going to look like. And we don't really worry too much that somebody is going to come along and not meet their demand. Upsides are upsides, right? I mean, upsides are then discussed at market pricing. So that's just a little bit of thumbnail the way the process is working.

That's great. That's interesting to hear sort of how that's working because that's super evolving thing. So we'll kind of see where that goes in the next 6 months. I guess one final topic that I wanted to talk about was your stock price. And I know that as the CEO of the company, I know that this is front and center to your mind. And the market, obviously, is clearly not valuating the synergies of these businesses and the value of the NAND business, if you just take Seagate's sales multiple, you -- I mean, it's basic math. But it's pretty clear that the NAND business is getting very, very little value in the marketplace today. So I guess I have 2 questions. One, what is the market missing? And two, I mean, you were -- you've been levered and you've been focusing on delevering. But as you kind of come out of that, how does that change sort of how you think about your stock price? Because obviously, you must be salivating with the potential to basically buy your stock at this price? So I'm just kind of curious how you think about that.

Yes. So it's -- again, it's the ultimate measure that we're focused on, which is driving shareholder value. That's what we're here to do. As I said earlier, we've been making very significant changes in this company over the last 1.5 years. You talked about the significant changes in the management team. We've added new members to the Board of Directors. We've restructured the company into a BU structure. That's -- and I think those things are working. We're giving -- we're driving more innovation. We're getting better results. I talked earlier about revenue, up 21%; gross margin, up 6 points; EPS has nearly tripled since I got here. I think with many -- look, what we're getting people to realize is those results are because of structural changes we have made. These are not like one time, we did something just to get a 1 or 2 quarter result. We have been doing the hard work of making fundamental structural changes in the company to get a better result over time. And we think as we continue to execute and that continues to show up, any valuation gap will get closed. And you're right, we are very anxious to get to a shareholder return policy. There's no doubt about it. Focusing on the balance sheet was something we made a commitment to do. We need to follow through on our commitment, and we're doing that. I mean getting to investment grade is a big milestone for us. It just happened late last week, and that's another step in the process of getting to where we need to go. So you're right. We're very focused on, one, make sure we execute the best we can execute and get the most out of what we have. We feel really good about the franchises we have. We talked a lot about the HDD business and the way that -- where that's at. It's the foundation of the cloud. That's a cloud franchise at this point. NAND, we've talked about the industry is undergoing some consolidation. We think we got a great position with our partner in Kioxia. It makes us the #1 producer of merchant NAND in the world, gives us a lot to invest in our technology road map. We've got a long heritage of really, really good technology, and we expect that to continue. And that's an evergreen market where we continue to see more and more applications being invented every day. So it was get the most out of those businesses we can, execute better, show the synergies and really show that they're real, and we're on our way to doing that and are executing better as we continue to do that and get back to a shareholder return policy, we're very anxious to get to that point.

Yes. I guess just to wrap up, I mean, my view on the stock is really informed mostly by the NAND business. I just don't like the current structure of the NAND industry. But I guess my question there is, do you think that you can achieve the metrics that you're driving for in NAND, given the current structure of the industry? Or do you think that we need to see consolidation in NAND for you to get the metrics that you think you do -- that you need to achieve for the market to then recognize that?

No. We think we can get there in the current structure. I mean the current structure is changing. Again, I think with -- between us and Kioxia, we are consolidated from a production point of view, which is, I think, how many people are making decisions about how much to invest in NAND and now with Hynix and Intel coming together is another step in that direction. So we feel good about where the industry is at. Now clearly, more consolidation would be welcome, more investment so we can continue to fuel the innovation that's needed in this market. It's a -- as you know, it's a very big TAM. It's one of the -- it's becoming more and more proxy to the data center, which is fantastic. And so more consolidation would be welcomed, but we don't think that, that has to happen for us to achieve the goals we want to achieve. And we think we're uniquely positioned as a company that can deal with the complete storage needs of our customers from HDDs to NAND from the foundation of the cloud to the device and everywhere across that entire architecture, we feel like we're very well positioned.

Got it. Well, I think we've run out of time. So I really super appreciate the time from both of you. It's been a great conversation. So thank you. Thank you again.

All right. Thank you very much. We appreciate it.

Yes. Thanks, Tim. Appreciate your time.