Seagate Technology Holdings plc (STX) Earnings Call Transcript & Summary

Perfect. All right. Let's start that over. Why don't we? Erik Woodring; I lead the hardware coverage here at Morgan Stanley. Quick research disclosure, for important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. So I'm delighted to be joined this morning by Seagate's CFO, Gianluca Romano. He's been at Seagate since 2019. He brings decades of financial experience, multiple senior roles at other companies. But Gianluca, thank you for joining us this morning.

Thank you very much. Thank you for inviting us here.

So it's been about 5 weeks since you reported December quarter results. At that time, you're alluding to still some continued end market challenges across enterprise cloud, China, consumer, each one at kind of various stages of slowdown or recovery. And so kind of what are you seeing from a demand standpoint today if we touch on each one of those markets?

Okay. And before I answer your question, let me just say that I will be making forward-looking statements today and you can learn more about the risk associated with those statements on our website. I would say, in general, now the quarter is coming out as we were expecting with a fairly slow January, a little bit better February. We have a lot to do in March. But overall our estimate today is we're aligned in terms of revenue to what we discussed at the earnings release. In terms of the segment, the improvement quarter-over-quarter is coming from nearline. Now going into the March quarter, March is also a seasonally down quarter, probably the lowest quarter of the calendar year for our legacy business and also for our video and image application that is part of the mass capacity. Despite this seasonality in the March quarter, we have guided revenue a little bit higher compared to December. So the improvement is all coming from nearline. There are different trends in different parts of the business. I would say, for China, we start to get some good discussions with our customers. At the same time, I would say, this is still preliminary discussion on midterm, long-term, I still wait to see real orders coming in. I would say we were not counting on a lot of China business in the March quarter, and it's probably what is going to happen as we were expecting. But we are also now more confident of what we can see in the next few quarters with, we think, a sequential recovery coming from that part of the world. So that is a good news, not too much for the current quarter, that maybe is not so important, but I would say for the rest of the calendar year. In the nearline space, I would say the major problem was inventory. In general -- every customer is different, but in general, I would say enterprise OEM are going maybe a bit faster in depleting the extra inventory that they had compared to cloud. But more or less, now everyone is going in the same direction. So probably the cloud part will take a little bit longer, a few more months from now to fully deplete the inventory that they had. But as I said before, sequentially, they are actually buying more. So there it will be 2 different trends. They are buying more; they are depleting maybe a little bit less. So it's the 2 trends. They need to find -- we need to find as an industry the right balance between short-term and inventory depletion. So it's a little bit of both. Finally, for consumer legacy in general, of course, as I said before, March is the lowest quarter of the calendar. Still inflation, as you know, is still very high. So there is not a lot of recovery from that part of the business. Again, we think in the next few quarters also that part will start to improve.

Okay. That's a great way to start. You answered question 2 and 3 all in 1 question, so that's perfect. I just want to make sure -- one bit of clarification on the inventory buildup and work-down is, when we talk about the end markets that are furthest through that inventory correction, I just want to make sure it's enterprise OEM before cloud. Is there anyone else that you'd point to in terms of -- is that an issue in China? Is that an issue in consumer? Those are more just end market demand-related trends more so?

I would say maybe cloud is more general. It's mainly U.S., but it will be more general. For the rest of the business in China, we don't see a lot of inventory buildup. So we think it's only a slowdown of the economy that happened in the last, actually, 3 quarters. And now it's starting to reopen. As you know, there was a budget meeting last week. So the budget has been approved for certain segments, for -- and now that budget has to translate into business and orders to Seagate.

Right. Okay. Okay. I guess another way I kind of want to delve into the kind of the near-term dynamics is really what you're hearing from customers. I think throughout this week we're hearing a bit of a mixed message across a number of different customers, whether that's delays, cancellations, pushouts. Just broadly speaking, from your customers, is there fear? Is it just kind of short-term concern? Is there still confidence in spend longer term? Help us frame the customer conversations you're having today?

I will say for the medium term, long term, we don't see any differences. I think they are very confident their business is going to grow, as they were expecting. They need to manage the short-term down-cycle, and in some cases this inventory buildup, but we don't see any change in the long-term view. So they will grow. Especially the mass capacity part of the business will continue to grow. And now it's just a matter of time. I would say the only difference is more on China. We are now starting to have this discussion with the customers is giving us little bit more confidence for the recovery in the next 2 or 3 quarters.

Okay. All right. And then as it relates to production, when you had talked about the production ramp back at earnings, is everything kind of trending in line with your expectations, not only in terms of the ramp, but in terms of the cost or charges associated with that ramp?

Maybe this is a bit different from what we discussed at the earnings release. We kept January and February production fairly low, probably could be lower than what we were planning before. So we will have a little bit higher underutilization charges. Again, it's just a short-term impact. We are still -- we still give the same estimate in terms of revenue. So it's just how we manage production, the timing of the production, the timing of the ramp and how we manage our cash because, of course, there is a trade-off between underutilization and cash. And of course, right now, we focused more on cash. So I would say probably the underutilization charge will be at the same level of December. I think at the earnings release and subsequent meeting I said it was fairly close. I was expecting a little bit lower. I think our plan was more in the $60 million to $70 million. In the December quarter, we had $80 million. I think it will be probably around $80 million at this point. So a little bit higher, not enormously higher, but a little bit higher. But again, it's not impacting our view of the revenue or our expectation of the revenue in the quarter.

Right. Okay. And then just as we -- I don't want to get ahead of ourselves, but just as we think about the ramp from March into June, how do we think about production ramp, underutilization, costs associated again with that? Or should we think about that as being a relatively cleaner quarter, for example, from the cost perspective?

It will be. Probably not completely clean. My prior estimate was to have June with a very minimal underutilization charges. I think we will have some but way, way lower of what we have seen in December and March, maybe half of that cost. Of course, we will know better in few more weeks, and when we discuss at the earnings release in April. But also, we will have some costs. Again, a big improvement sequentially but not zero.

Okay. And then again, as we think about going from March into the rest of the year, whether it's fiscal or calendar doesn't necessarily matter, but where does the relative improvement come? Is it broader based? Where should we think about the relative, again, kind of like sequential strength coming from relative to [ the bucket ]?

I think all the segments will improve for different reasons. Nearline will improve because inventory will be depleted. China will improve because they are now in a different phase of their economy. And the legacy part, particularly the consumer part, hopefully, will improve because inflation will start to decline and also because of seasonality. Now that part of the business, if you look at the calendar year, as we know, sequentially improve normally, and we think 2023 should be a more normalized year compared to the 2022 that was not [ usual ].

Fair enough. Fair enough. Last quarter -- I just kind of want to touch on pricing, too. Last quarter, you talked about some pricing pressure and lower capacity drives; otherwise, relatively stable pricing. How would you frame the pricing environment today? Is that still the case? Or what has changed, if at all?

I would say it's still the case. The lower capacity drives pricing is impacting more from the NAND situation than the hard disk competition. As you know, right now, NAND is selling probably at negative gross margin. It's not going to remain like that for very long time, I hope. But right now, the situation is, of course, creating more competition with our 1-terabyte, 2-terabyte, 4-terabyte drives. So on that part, there is pricing pressure. Again, possibly, this will improve actually in the rest of the calendar year when the NAND cycle will start to moving on the upside. The high capacity, I would say no change. Wherever there is no overlap between NAND and hard disk, I would say the industry is actually behaving in a way to support as much as possible profitability even during a down-cycle instead of fighting for a few points of market shares.

Right. Okay. Okay. Very helpful. Let's touch on some of the announcements or kind of the key announcement that I thought you made at December quarter earnings in terms of starting to qualify 30-terabyte plus HAMR drives in the June quarter. Just as an update, is there -- are you still on track to start commercializing in June? Maybe just -- if there's anything you'd add relative to what you spoke about in December just in terms of an update on HAMR, and we'll kind of dig into a few questions.

We are going a little bit faster. So that is a good news. We are shipping samples. Actually, we started shipping samples already end of January, February. So it's a bit earlier than what we were planning. So we will start our qualification process in the next quarter, possibly fairly early in the quarter. That is little bit earlier than what we were expecting also. So I don't see any change for the commercialization part starting in the July quarter.

Okay. Okay. And then when you guys are having these conversations with customers and you are -- they are on 22 to 24-terabyte PMR drives, how do you communicate the decision to purchase something like 30-terabyte and higher? What are the key factors that get them to move from one capacity up to the next, especially with a new technology?

Yes. I would say probably technology is not impacting the decision. It's always the TCO that is impacting the decision. And of course, how much they want to invest in qualification. Of course, in the up-cycle of the demand, they want to go very fast because they are buying millions of units. In the down-cycle it can take a little bit more time. But the driver is always the same, is a better TCO. They are physical [indiscernible]. So they want to optimize the cost and the revenue for that physical [ slot ]. So moving from 20 terabytes to 22 or 24, and by the way we will have a 24 PMR drive coming out fairly soon also, so give them a good improvement in TCO and a good improvement in revenue for that single slot 20 to 22, they basically get 10% more revenue and they get a better cost. So there is a double advantage. The advantage is much bigger when you can jump from 22 to 30. That is a much bigger advantage for them and benefit for them. It will be also a benefit for us, of course, because we want to keep some of that improvement to increase our gross margin and our profitability in general and our EPS and our free cash flow. So it will -- we think it will be a win-win, for both for us and the customers. So far, we are progressing as we said, and we were expecting, so no change to our plan.

Good, good. So then just thinking about how that can have an impact on the model and the P&L, how do you initially frame the revenue contribution from this launch? How long does it generally take after launching a new capacity point for a customer to crossover? Are we multiple years away from this moving the needle? Just again, help frame the relative impact on the revenue side for us at least.

Okay, if you look at the past, crossover is usually 3 to 4 quarters. So I'll say, for the long term, this is what I expect also for HAMR. There is no reason why it should be different. For the first HAMR product, it's more difficult to say because it's a new technology. So we need to rent. We need to see what is the yield when we start renting hundreds of thousands of drives. And if the yield is good, we will do exactly what we have done with PMR. If the yield is not as good, we will go through our learning curve and then improve the yield, and then we ramp. We don't want to ramp a product that is not yielding financially a good result. So I would say, I would not -- I cannot be 100% sure on the first product, but we are very confident that after the normal learning curve, that by the way everyone will have to go through when they have HAMR, after that, it will be a normal transition.

Right. And then maybe just as a slight deviation on that, how do you think about, specifically when it comes to HAMR, the competitive landscape and the competitive advantage that you have on that end?

Yes. I'll say it's important that the industry is going in the same direction. I think at least one of our competitors is now developing HAMR. They are little bit late, mainly because they were looking at different technology before and then they decide to come back to HAMR. So that is -- our advantage is mainly, I would say, 2 things. One is we have a product today. So we don't need to develop HAMR anymore. We have the technology. We have a product. We just need to develop the next product. They still need to go through the development of the technology. So the benefit for us is the time and the fact that it's not sure that other companies will be able to develop HAMR even if now, we think, in longer term probably they can. But we think we have a fairly big time advantage. And again, HAMR is growing faster than PMR in terms of capacity. So the first product is a 30-terabyte. The following product will be a 36-terabyte. So having time advantage is extremely important because that 1 or 2 products advantage is a lot of capacity, a lot of benefit for customers. So in theory, they should be more interested in products having higher capacity. And I think we will have that higher capacity for a fairly long time.

Okay. Okay. So then maybe switching -- again last question on HAMR, but switching kind of to the cost side of things, longer term we'd expect it to -- or at least you guys have communicated it to drive an improvement in gross margins. How do we think about that ramp? And then is that a point of tailwind? And then again, in the near term, what are -- are there any other costs that we need to be mindful of just as you think about the ramp of HAMR and the impact that it would have on the cost side of the equation?

Yes. I would say that is the second major benefit is not only the capacity increase but also the cost decrease coming from HAMR. In a very simplified way, I would say, it should take 30 terabytes, 36 terabytes, 40 terabytes. I think at the earnings release, we discussed about 5 terabytes per disk, so a 50-terabyte drive. All those drives will have the same below material. They will have 10 disks and 20 heads, so an enormous increase in capacity, but with the same unit cost. So the cost per terabyte will decline a lot, much more than what you have seen recently with PMR, where the increase was just couple of terabytes and often was coming from an additional disk in it, [ 2 more x ]. So it's a great opportunity to reduce the cost and to keep a big part of the benefit into Seagate to improve our own EPS, to improve our own cash flow.

Okay. So I imagine the next question I ask, part of the contribution you'll see is from HAMR, but you're targeting 30% to 33% gross margins, 18% to 22% operating margins. We've really only seen you generate that type of gross margin when revenue is, call it, above $3 billion on a quarterly run rate. Is that what we need to see for you to get back to 30%? Can you do 30% with a smaller revenue base? And if so, how do you get there?

I will say when we go back to that level of revenue, and adding HAMR part of that revenue, we expect we have a higher profitability. We gave the range of 30% to 33% for gross margin and now 18% to 22% for operating margin based on a PMR road map. And we actually achieved that December last year, 2021, not 2022. So we were already at that point. Before the down-cycle, we were already at 32%. So when we go back to that level of revenue and with some contribution from HAMR, I expect to be higher than where we were before. Some is product, some is mix, some is our changing cost structure. Now as you know, we went through a fairly big restructuring effort in November. So this is also taking out cost that is not needed anymore for the company. And every day, every week we look to more opportunity to improve our cost structure and reduce our cost. So even the time that we will have will help us in also reducing our fixed cost and cost structure overall.

Okay. So I want to touch on kind of 2 debates or very common questions that I get from investors. One is just, again, HAMR plays a part of this, but thinking about the competitive landscape when it comes to flash and the threat of disruption longer term, how do we -- how should we think about that? How is HAMR and anything beyond that kind of insulating you? Is there -- does it keep you up at night? How do you think about that?

No, not really. It is an interesting question, but...

There's other things in there.

I'd say NAND is overlapping with hard disk only on the low capacity. And that part of the business is declining. Of course, until, say, mid of fiscal year '20, our legacy business or low capacity compared to high capacity was still in the 50%. So that decline was a major impact to the company. And despite that, company overall revenue was actually growing in fiscal '20, in fiscal '21, in fiscal '22. I think last year legacy was maybe 25%, so half of what it was just a couple of years before. So the impact is becoming smaller and smaller, and in the future will be even smaller. So at a certain point, that part of the business will be completely immaterial to Seagate. Now you also need to consider the impact of HAMR at lower capacity, because today everyone is focused on the high capacity revenue and profitability coming from HAMR. But if you think, in few years we will have a 40-terabyte HAMR drive with 10 disks and 20 heads. At that point, probably segments like video and image application that today use 8, 10, 12 terabyte, probably they will be using 20 terabytes. We can have an HAMR 20-terabyte with 5 disk and 10 heads. So that is a major improvement also for lower capacity applications where we can develop our HAMR and get an improvement in our profitability, so but -- and at the same time having a bigger gap with NAND. So we have a lot of benefits coming from HAMR that are little bit underestimated right now. Of course, we're in a down-cycle. So it's kind of normal to be less optimistic. But when we will meet again in a few quarters, I think we will have a different view.

Okay. Okay. The other debate is kind of around the cyclicality of the business. Again, you touched on it, but you're seeing this mix shift to mass capacity, more specifically nearline, if we think of those end markets as being more secular type growers, that becomes a more defensible model for Seagate relative to history. And someone made the comment to me, well, sure, that could be true but look what happened in fiscal '22. And so I guess my question for you is, should this mix shift actually make your business less cyclical? How do you think about that? Well, I imagine there will still be a degree of cyclicality, of course, but.

I would say, it's not only hard disk; I'd say, in almost all the technology industries, you have some kind of cyclicality. The important is to look at the business over a certain number of years and see what is the trend and what are the drivers for demand. And I would say even today there is no change in the drivers for data storage demand growth. All the important new applications that now are maybe progressing a little bit slower than what it was before, but they're still progressing, and they will restart generating a lot of data, artificial intelligence and machine learning and smart cities, smart factories, autonomous driving Internet of Things. All those are not disappearing. Have just slowed down a little bit. That will come back and it will generate a lot of data. 90% of that data is going to be stored on hard disk. The cycle can happen, of course. Even on the upside, if you look at the growth that we had in fiscal year '20 for mass capacity in terms of exabyte was almost 60%. And then the following year was 30%, and fiscal '22 was another 30%. It's not always the same. It's different. Now the other important point is the base is now much bigger. The cloud business was almost zero 10 years ago; now has millions and millions of units installed and exabyte. So you also need to consider that the percentage maybe can change a little bit, but the exabyte, that is what we produce, is actually probably increasing more, even if the percentage of growth is little bit lower.

So in and around this conference, the topic of AI has been pervasive. You mentioned it there. When we think about the big-picture drivers of data, is AI and machine learning kind of like #1 in your mind? Is there -- are we at an inflection point, at least from your perspective? Is it IoT? Help us think about the relative importance of AI for you.

It's a bit difficult for us to know which one will go faster. I would say they all generate data. And even AI there are, I would say, 2 parts of AI. One is a lot of computing, where we play less. And the second is a lot of data storage where we are the predominant industry. So difficult to know which of those applications will be generating more data in the next year or 2. I would say they will all generate more data and probably some are going to be faster and some of it could be slower. But for us, data is data. So we just store data and exabyte and...

We just want more data.

These are no different.

All right. Well, that brings us up the time. So Gianluca, thank you very much for [ talking with us here ]

Thank you, Erik.

Thank you.