Seven Questions for Seagate CEO Steve Luczo About the Effects of the Thailand Floods
The country is now beginning the arduous job of cleaning up from the floods that killed upwards of 600 people and dealt a body blow to its industrial and manufacturing base.
One industry hit especially hard is the computer business. The world relies on factories in Thailand to turn out critical components used to build hard drives, and factories there are out of commission for now. This is not a trivial problem — the factories in question are not easy to replace, retool and restart once they dry out. Nor is the answer simply for the hard drive manufacturers to build new factories somewhere outside the flood zone.
This is the kind of supply chain disruption that the computer industry hasn’t seen in many years. I had a chance to talk with Steve Luczo, the CEO of Seagate Technology, for his view of the situation. Seagate has been relatively lucky in that its factories haven’t been directly impacted like those of Western Digital and Toshiba. But many companies that supply Seagate with necessary components have been hit, and it will be some time before they’re back on their feet.
Luczo told me that the computer industry as a whole — including companies who make PCs, servers, workstations and any other device that contains a hard drive, whether a set-top box or an enterprise storage device — can expect acute supply-chain disruptions to last well into 2012, and that it will take until the end of 2013 for the industry to return to its pre-flood operating posture. You read that right: It will be two years before the supply of hard drives is anywhere near “back to normal,” and there are simply no easy solutions for getting it fixed.
An estimate by the market research firm IHS iSuppli pegs the available supply at 125 million units, which is about 29 percent short of demand of 175 million units. By its reckoning, more than one-quarter of the world’s hard drive manufacturing capacity has been disrupted in one way or another, including 45 percent of the capacity devoted to making hard drives for personal computers. I spoke with Luczo by phone yesterday, and tossed in an extra eighth question because of the importance of the subject.
AllThingsD: Steve, at a high level, I think everyone understands the problem. There’s been a terrible flood in Thailand, and a lot of factories that make crucial parts for hard drives are out of commission. To that end, I think people expect this to be a temporary problem that works itself out in a couple of months. But you say it’s a much more complex problem than most people realize. You’re tracking this situation day to day, and probably hour by hour. So, how bad is it, really? And what’s likely to happen?
Luczo: What’s surprising to us is that even with all the data out there — we’re six weeks into it — there are a lot of fairly sophisticated companies that haven’t fully come to grips with the depth of the problem and the duration that is likely to occur. What is going to happen in the next couple of weeks is that the real shortage begins to show up right about now. There was already a lot of built inventory and a lot of finished goods moving through the system. And now all that is gone, and I think customers are starting to see shelves of parts go empty, and realizing that they’re not going to be filled for anywhere from one to two months. So the concern is heightened.
We heard Meg Whitman talk about this on HP’s earnings call Monday. She said HP stepped in and started doing some strategic buying. She says HP is going to see effects at least through the first half of next year. Apple talked about it on its earnings conference call, too. Are you hearing from them?
Tim Cook at Apple was way in front of this. I saw Tim the first week it happened, and took him through the situation, and in 15 minutes he understood the magnitude of it. Meg was on the second week of her job as CEO when I went to see her, and she got it right away. HP’s procurement VP, Tony Prophet, was also early to understand this. Companies like that reached out to us early on, because they understood that this is going to be an extended problem. They started asking for longer supply agreements. Deals that would typically last about a year, they’re now asking for two years.
How bad is it really going to be? What’s your outlier worst-case scenario, and then what do you think is a little more realistic?
If you think pre-flood, a mix [of products] that the customers need, the industry had the capacity to ship about 190 million units a quarter. Pre-flood, we expected the demand to be pretty consistent at about 180 million a quarter, with a bump in September 2012 for Windows 8. We now believe the March quarter is going to much more difficult than the December quarter, and December is going to be about 120 million or so. We think the March quarter will be about 120 million, in the best-case scenario. And that’s with customers mixing down pretty aggressively; and by that, I mean companies like Western Digital, who don’t have access to the sliders [a critical component in a drive], are shipping one- and two-headed devices so they can ship more units. So instead of shipping a drive that contains two disks and four heads, which is what the market needs right now, they’ll be shipping a one-disk, one-head or one-desk, two-head product. They’ll be maximizing the units they can sell, rather than shipping the product the customer actually needs. … So we see something like 130 million for March on the optimistic side, and then 150 million for June, 170 for September and then 190 million for December. And so by the end of 2012 you’re back to being close to industry demand. But even then, you’ve not included the impact of that missed 100 million units. And that will take another year to absorb, because it’s not like the industry is building new factories to chase that demand. We can’t over-invest to meet some bubble and then get stuck with excess capacity.
I think, intuitively, people expected companies like Seagate to just build more factories outside of the flood zone, but it’s not that simple, is it? Would this not be a moment to add capacity?
There are some in the investment community who think that’s what is going to happen, and that there will end up being a supply glut after all this is over, but it’s not the case. For us, it’s more a function of how to recover the supply chain and then work with the customer to get a good read on what their needs are for the next several quarters. If we see a multiquarter shortage that goes beyond what I described before, then we would think about maybe putting some capital in place. But we’re not going to do that to solve a temporary problem, because we end up being stuck with the excess capacity. Now if it turns out there is no recovery, and then the industry is more constrained than I first described — and that, by June, the industry is still 30-40 million units short and looks like it will be for the next six quarters — we might revisit. But then we’d want longer-term commitments to make sure we’re not overinvesting. But we’re not to that point yet.
What is this doing to prices? And what does that mean to the person who wants to buy a computer or server this year or next year?
If you look at a 10-year moving average trend, the industry has in general seen prices come down about 2 to 3 percent a quarter, and that is for a particular product. In 2009, there was a little price erosion, and that was because the storage industry recovered quickly from the recession. And there had been massive capital cutbacks, so there were big shortfalls through all of 2009 and into 2010. Then, when the Greece crisis happened, that put a big flatline on a lot of growth, and the industry had put in a lot of capital because everyone expected there would be growth. So, since spring of 2010, the price erosion has been higher than normal, which would show that supply is greater than demand. And what this flood has done is drive the supply curve down, while the demand curve has stayed constant. For OEMs [original equipment manufacturers, or the PC and server manufacturers like Apple, HP and Dell, who buy directly from Seagate], you’re seeing an average increase of about 20 percent, and in the channel [resellers who sell parts to smaller PC and server vendors], probably much higher. So all the sensational quotes you see about pricing are about those that occur in the channel, where we have no control whatsoever.
The markups in the channel are much higher? Are the channel guys taking advantage of this?
Yes, they’re higher, but I don’t think they’re taking advantage. I’ve heard stories about drives that we sell to OEMs for $60 that show up in the channel at $105. Normally the channel price is within about 10 percent of the OEM price. It’s just the law of supply and demand. They can’t get supply. The channel is getting about a third, at most, of the supply they would typically get. The OEMs are the ones with the supply agreements, so everyone in the channel is way short. In some market segments, supply is about 70 percent below what the demand is. And so those shortages are very acute. The channel is selling the few drives that are out there to whoever needs them the most and is willing to pay for them.
So what does all this mean for Seagate, specifically?
For us it’s a different story, because we’re going to be driving more volume than our competitors, because we’re not as directly affected, and we’re going to be making some technology transitions. When we do that, it lets us take cost out of our product, so we can offer more capacity for the same or fewer parts. That helps us drive down pricing. Our goal is to recapture some of the more aggressive pricing of the last eight quarters, in order to sort of get our business back in balance. Our long-term business model calls for gross margins of 22 to 26 percent. And we use our manufacturing expertise to drive down our costs and then pass that on to our customers. This quarter, end users really won’t see it, because product has been built and has been on the shelves. As the shortages just started occurring, you’re starting to see prices increase in the channel. And then at the OEM there will be shortages in some high-value areas like enterprise storage or cloud computing. You’re going to have to see price increases, because there’s such big shortages.
One thing that occurred to me when I first wrote about this a month or so ago is that it represents an opportunity for the flash memory chip companies to make some inroads against hard-drive guys like you, mainly on notebooks. Is there a threat that flash could pick up some of the demand?
Some of it, but not very much. I think to the extent that there is a high value purchaser who can afford to pay $200 for 100 gigabytes, then that market will expand from 1-2 percent to 3-4 percent. Of the 35 to 40 percent shortage that exists, could you see a little of that get absorbed by silicon? The answer is yes. But there’s a cap. There’s just not enough of a raw supply of silicon to meet all the demand. Our industry will ship 400 exabytes this year. We would have shipped 450, were it not for the floods. Of that, 180 exabytes is notebooks. Reduce that by 30 percent, and you get about 55 or 60 exabytes. If you were to take all of the capacity from Samsung’s newest state-of-the-art flash factory, and dedicated it just to notebooks, it would only put out 7 exabytes a year. Plus, there are already other markets demanding flash, like tablets and cellphones and other things. So it’s not like you can steal from those other markets. You’re not going to take a $32 product and replace it with a $350 product. Can you do it at the edges of the market? Sure. But the threat is capped by the amount of silicon available and the price point for flash storage, which is still an order of magnitude higher.