Micron Technology, Inc. (MU) Earnings Call Transcript & Summary

Perfect. Thank you, everybody, for joining us this afternoon. I'm Aaron Rakers. I'm the semiconductor and hardware analyst here at Wells Fargo. And pleased to have with us Mark Murphy, the CFO of Micron.

Mark, there's a few things going on in your market right now, and you recently had an 8-K filing. So maybe why don't we just start there, just level set us, the audience, what's going on at the company, the context of that 8-K? Then we'll roll from there.

Sure. Is this working? Yes. Thanks, Aaron, and appreciate being here, and thank you all for joining us today. I'll start with the safe harbor. I'll make forward-looking statements. These statements have risks and uncertainties associated with them. Refer you to our risk factors disclosed in our public filings, the most recent of which has been our 10-K in October. We did announce -- had a press release on the 16th of November, where we announced a reduction in wafer starts, 20% off of our fourth quarter '22 levels, and that's across both DRAM and NAND. It is a really tough environment. We still see customers doing inventory adjustments. That continues. And then in a number of markets, we see weak end-market demand. So we're working through that, but it's been very difficult. And again, that's why we chose to take the supply action. I'll talk a bit more about that later. I'm not going to provide a full update on the guidance today. But what I would say is that pricing has trended well below what we thought it would be when we had our earnings call. So the pricing environment has been difficult. And that's despite what we think have been very good and disciplined actions on our part around pricing. We have been walking away from deals where we think the market -- or the price doesn't reflect market. So it's been difficult. If we look out to calendar '23 and based on our discussions with customers, our own industry analysis, our own market view, it's become clear to us that customers are entering calendar '23 with high levels of inventory still. And so that's going to suppress their demand in calendar '23, and that we're incorporating that into our views. And again, that informed the supply decision that we announced on the 16th. In addition to our view on market, our customer discussions, industry analysis, there is, as you know, a poor macro backdrop. And we're seeing -- and you see that manifest most strongly in the consumer market still, and we are dealing with that. So based on that calendar '23 view, as we mentioned, we've taken aggressive supply actions, one of which is the 20% reduction relative to fourth quarter levels on wafer starts. That's in addition to the 5 -- that's -- we had announced 5%. So this is bringing that up to 20%, an incremental 15% reduction. And in addition to the wafer starts reduction, we're also continuing to look at CapEx and seeing where we can reduce CapEx further. We had announced in our earnings call that we would -- that we were targeting approximately 8%, so we're targeting below 8% at this point. The net result of those supply actions, which we think are important to try and bring supply-demand imbalance. The net result of that is that in calendar '23, we believe that our DRAM supply or production will contract in calendar '23. And as we'll talk about here today, we've got ample inventories to meet the demand gap. And then on NAND, we expect calendar '23 supply or production to just increase slightly. Now in calendar '23, we do expect there to be bit growth, so a bit demand growth. And -- but we do expect that to be lower than the long-term DRAM and NAND CAGRs that we've communicated before. And again, that's because of the demand outlook that I talked about earlier. As it relates to our fiscal year on the demand side, we do still believe those sort of near-term conditions are a bit worse than we expected on price. We do still believe that the second half of our fiscal year will still be higher volumes than the first half of the fiscal year. So we'll see we began as customers deplete their inventories and begin to recover, demand will begin to recover and pick up, particularly in the May quarter and beyond. So I think I'd just like to end opening comments with, we're in a cycle downturn, a cyclical downturn, we're managing it aggressively working to get supply/demand in balance. And -- but as we look out longer term, and we think through cycle, we're really excited about the market. The world is going to need more memory and storage. There's very strong secular growth drivers that we all know, data center, automotive, 5G, and so there's -- artificial intelligence, that's going to drive products and technology, the space will grow. And then Micron is in a great position. We've got a strong balance sheet to weather this downturn. We're leaders in technology. We've got the best set of products in this space, and we're manufacturing just outstanding manufacturing excellence at this time.

So Mark, you've left me with a tremendous amount to unpack. And I appreciate that kind of overview of what's kind of been going on. When we think about the production side of the equation, and you think about particularly focusing on DRAM, can you help us unpack maybe a little bit of how much you can pull down supply side via CapEx reductions? Or you've already outlined a 50%-plus cut to WFE into next year versus maybe slowing the progression of technology nodes. How do we think about those 2 when we think about that setup, what your plans are for '23 supply?

Yes, they're related. And of course, we will modulate it based on the market conditions. It's going to require all the levers that we can pull. And candidly, we can't do it alone. It's got to be an industry level. There's a gross oversupply in the space. So -- and our share is such that we do our part. The -- we've got -- it's a bitter sweet here in the sense that we've got outstanding technology in 1-beta and 232-layer on DRAM and NAND, respectively. And we're just not going to be able to do high-volume manufacturing and invest in that at this time. Now we've invested in enough capacity in order to keep the engineering efforts going and product quals going so that when the market picks back up, we can move those to high volume, but the node transitions are going to be slower or slowed down. And candidly, there's an opportunity cost and that we don't get the cost savings, but the more important item at this point is to get the market back into balance, which is what we're focused on. And then we talked about the reduction in wafer starts, which we've thought very carefully about where and how we do that. And so the org is mobilized and moving and has moved on that. And then we'll continue to look at other CapEx reduction opportunities that make sense to minimize the long-term damage and yield a good near-term result, like to use the expression - avoid, reduce, delay. I mean we're going through and seeing what we can avoid. If we have to spend something, can we spend less? And if we just have to spend that amount, can we spend it later? And we're going through that exercise because at this point, the fundamental financial performance of the business is something we need to improve.

So I guess with that, you mentioned at the beginning of the conversation, pricing has been more -- off more than what you thought this quarter. With the actions you're taking, do I think about '23 as what you would have previously thought from a cost down perspective is kind of modulated lower as well, given these actions you are taking, that's fair?

That's fair. I mean, I think our -- and we'll go into this in more detail during the earnings call in December, but they're based on the actions that we're taking, we would expect there to be less cost downs than we had originally.

Yes. So the question I often get is Micron's executed extremely well from a technology perspective. The balance sheet is in a much different place than what it was even 3, 4, 5 years ago. But the question I often get is your own owned inventory and how we think about the mechanics of -- you've talked about 150-plus days of inventory, where do you think that could potentially kind of go? When do you think we start to see that kind of begin to kind of compress? And the question that goes along with that is just remind us again of how -- I often get the -- is there a risk of inventory write-down? How do we think about that in the context of the story with Micron with their inventory and the balance sheet in general?

Yes. It's a good question. We, of course, manage inventory very carefully and spend a lot of time on it. I think it may be helpful just to set the stage as to what is our inventory strategy, and Manish talked about this a bit at the Investor Day. But we build our inventories and tend to build them principally into whitform, and that's to reduce the risk of obsolescence on that inventory and to reduce the overall cost at various points in the build, right? So we keep in wafer form as long as we can. And then when we need to singulate the die and begin to move it into packages and we can build -- add the other components. And then it's -- the most costly inventory is held for the shortest amount of time and the obsolescence risk is lower. And so that's our general strategy. And you'll see -- you can see it from our results that most of the inventory sits in whit. So there's always some scrap and obsolescence that occurs normally in a given quarter. But the risk that you talk about is a net realizable value risk, which we watch very carefully. We have -- we do forecast on pricing. We look at customer trends. We look at a number of other factors. And as disclosed in our K, and we've done this for years, we look at inventory in a single pool. So where we will face any potential for write-down in inventories is when the profitability of the business goes negative and -- a gross margin. And as of the last reported results and our outlook at that time, there was no need to write down. So we will continue to look at that and of course account for things correctly.

And I guess, given it's a pooling mechanism, it's -- like you could have a situation where DRAM turns -- or NAND turns negative, and in aggregate, you wouldn't have to take [ currently ] write-downs. That's the pooling mechanics of it.

Correct.

Okay. And when I think about just the balance sheet in general, maybe help us in the down cycle that we're going through, how you're modulating your views on capital return versus capital preservation on the balance sheet? Just any kind of thoughts and obviously, that coincides with free cash flow context of that as we look through this.

Yes. So certainly, free cash flow focused and doing what we can do. We said we're going to have negative free cash flow of $1.5 billion or more in this first quarter. We also said that the second quarter would be a challenge. And then as I just mentioned, the actual conditions have gotten worse. So it's a tough period, but the reason that the company has a lot of liquidity is for these cyclical periods that we can make the right decisions during these downturns. And you mentioned it earlier that the balance sheet is in the best shape it's ever been, and that's true. And that is allowing us to make the right technology and other decisions that we need to make that on the other side of the cycle, we'll maintain our technology leadership. We'll maintain a good product portfolio we have, and that we'll maintain our manufacturing expertise. The company really struggled, and a lot of companies did in the space a decade or more ago because you'd enter these downturns and you'd make a lot of bad decisions in order to salvage liquidity. And -- but we're fortunate and -- but that comfort that we have good liquidity position, don't mistake that for any sort of not working aggressively to turn the situation around, being very mindful of it. So you saw that we went out to the debt markets and raised some additional liquidity to kind of ride this out and continue to make the right decisions. And as it relates to capital return, through cycle, there's certainly no change. We expect this business to outpace broader semi growth. We expect EBITDA margins over 50%; operating margin, 30%. We've got free cash flow over 10% is what we expect to do through cycle. And that will then generate free cash flow to have the capital return that we want in the form of sustaining and raising the dividend and then share repurchase. As mentioned on the call, we did do some repurchase actually in this quarter. And -- but certainly, as the conditions are what they are, we're focused on liquidity and making sure the balance sheet maintains at an investment-grade level.

So a lot of what you're doing, in my opinion, is very aggressive, unprecedented moves and the right moves to make -- to get to eventually the other side of this. The context of that also comes back to the industry as a whole. As you guys kind of assess what's going on from that perspective, and obviously, there's a big competitor of yours that was vocal in saying like, we're not going to arbitrarily cut production, but how do you think about the discipline of the industry relative to what Micron is doing?

Well, I mean we can only control our actions, which we're doing. And yes, there have been some -- yes, there's been some indication that you see through the toolmakers that they've had folks who have been cutting CapEx. And we'll just have to see. All of us will want to improve the fundamentals of the business because it's not -- in the current state, it's just not sustainable. So -- and we all have dividends. We all -- yes, there's an expectation that we all deploy capital wisely. And so we'll just -- we'll see. We're focused on controlling what we can control and we've taken aggressive action on the supply side and are taking other actions just operationally around cost and other things.

Yes. At the Analyst Day, I think it was back in May that you guys held the Analyst Day and you mentioned earlier that this through-cycle earnings power, it sounds like you're comfortable in that through-cycle model framework. At the event, it was also an interesting topic of the discussion around these long-term supply agreements. And so I'm curious of where we stand today on those? They were never kind of volume contractually committed agreements. But how have those maybe -- are those still helping you shape a little bit of this kind of idea of getting inventories down at the customers and getting back to a point where you got supply-demand equilibrium, just any context around the LTA?

Yes. The -- and I'm just going to talk about the conventional LTAs because it's the bulk of our contractual structures. But I think it's 70%, 75% of our business now is under LTA. And they're not take or pays, but they are very helpful. And they're especially helpful, even though they're not take-or-pay, they're especially helpful in periods like this because you get a dialogue with the customers that you wouldn't otherwise get. And we get a clear demand signal from customers when they know there's a commitment. We -- when they're unable to meet their commitment, there's an escalation that occurs, that escalation occurs at the CEO level. So we have CEO-to-CEO talking. That then enables an exchange of information around their inventory levels and what they see in the market and a richer discussion than we would have otherwise. So I think there's a great benefit there. And there's also in these, I think, important to note, technology relationship building and just deeper intimacy on making sure that we're working together on future products and technology. And so I think that they're an important part of our business for ourselves and our customers.

Yes. A couple of other quick questions. I just want to maybe at the perspective of the production cuts and the changing dynamics in the market, maybe if we could talk a little bit about end-market, end-verticals, right? I mean is there -- it seems like it's broad-based. PCs are weak, smartphones are weak. But any kind of things that stand out from an end-vertical perspective as we go through what we're going through, be it data center and automotive as well? .

I'm looking forward to which one turns around. There's a lot of questions today on that. I'd say maybe start with the positive. Automotive, which we thought we saw in August, maybe some signs of automotive weakening has actually stayed pretty strong, so -- which is good. And I think they had probably the most severe supply chain issues. So I'm sure that -- in a way, that's [indiscernible] them from some of the other issues that we're seeing. On the PC and smartphone side, still very weak. I think we're expecting PC to contract again this year and smartphone to be single -- basically flattish in '23. So -- and they were both off 15% versus earlier in '22 expectations. So they -- I think which was unprecedented in those spaces. So it's been very severe, and they remain weak. And I think that's probably also the macro backdrop and consumer getting more cautious in spending. Data center, as the end demand, we believe, is still strong, and you see that I was sort of just reading a report on a company today where it said the growth in data center was 35% plus as I could see going out. So I think the growth there is still good, but they're working inventory levels down and they -- eventually, they'll need to replenish those inventories. So we do believe that kind of mid-next year, that be it data center or maybe some of these other consumer markets that eventually, they need to replenish inventories. And the world -- in the end, what's reassuring about our business is the world will need more memory and storage. There is no doubt about that. And so it's just a case of, again, getting supply-demand imbalance and then the fundamentals of the business will improve.

2 And I think it goes without saying like we are truly in unprecedented times because we've never seen a bit declines year-over-year like we're seeing right now. So eventually, that does come back. I don't think anybody here believes that data doesn't grow over time. With that concept???

To your point, yes, there's -- between the August quarter and the November quarter, we've never seen a sequential decrease in volume of that magnitude in the industry ever. So I mean -- and then on the revenue side, the decline we've seen really hasn't been seen since the dot-com period. So it is a very severe and sharp downturn that, again, we're taking the actions we think we need.

And I guess where I was going to go to is that for those of us that try to look through these cycles and subscribe to the through-cycle earnings power of the company, has there been anything that's changed your mind of, maybe remind us of the underpinnings of the bit demand longer-term model expectations you have DRAM and NAND Flash, mid-teens, what high 20s?

Yes. We see mid-teens on DRAM, and we've not changed any view on NAND, which we said high 20s, [indiscernible].

The final -- looks like I've got a couple of minutes left. I'd love to kind of talk about just the -- some of the technology stuff. I know that you're the CFO, but maybe just remind us, and again, appreciating that things are kind of moving out to the right a little bit as far as you're modulating the bit demand side of the equation. But the progression of 1-beta DRAM, the transition to 232-layer 3D NAND, how do we think about the cadence of now those transitions looking forward?

They're both great technologies. That's a shame here is they're both very good nodes, and so there's a lot of disappointment that we're not in a position to ramp those into high volume because we would have really been in a great spot. But the fact of the matter is that the market is such that we can't do that. So again, we're slowing the node penetration, keeping enough volume to be able to do quals, and things like that so we are on the balls of our feet, ready to go when the market recovers. And then decisions on next nodes, just we're still working that, right, as to, okay, does this mean these are shorter nodes or longer or do we delay the next nodes? We just -- we're still working through that, and that's going to be a function of the market conditions, supply-demand balance and other factors.

In the perspective of NAND, I think Micron at least has been one of the more vocal about the role that QLC plays, quad-level cell NAND. And the potential for demand elasticity to move more deeply maybe into competing against hard disk drives or displacing hard drives. So where does that stand as far as the road map right now and the expectations you guys have for QLC?

I mean we've got great expectations. We're in a leadership position there. And as you pointed out, there's a cost of ownership benefit that customers get with QLC. So we see continued penetration there, and we're especially well positioned in it.

And that is -- have you guys ever said how much that is as a percent?

We have not -- for competitive reasons, we've not [indiscernible] if that is fair, but we believe we're the leader.

So with the minute I got left, Mark, I'm just going to leave it open ended. Is there anything that maybe I didn't ask, maybe that we didn't talk about? I know oftentimes this comes up is book value support, and that's back to the balance sheet discussion. But is there anything that you'd leave us with in addition to what has been said as far as thinking about Micron right now?

No, I just think, I think maybe the takeaway is the market conditions are a bit -- well, they're softer than we expected at the -- as it relates to price from our last earnings call, and we are managing that aggressively, and you see that in the supply actions that we've taken. But I think the long term, we're in a great position on technology products manufacturing and we'll just continue to manage the business well. And then I think it's important to keep in mind if we invest -- if we manage the business the way that we are, that eventually that book value will continue to increase. And then it's important to maybe keep in mind the replacement value of the assets that we've got...

Replacement value...

$100 billion-ish number. Yes. And plus the patents and all that technology. So there's a lot of embedded value in the company that we just need to...

Okay. Mark, thank you so much for joining us. Appreciate it. Thanks for your time.

Thank you, Aaron.