This last week, Eli Harrari presented at JPMorgan’s Technology, Media & Telecom Conference.
There was a bit of buzz leading into presentation day. The thinking was that Eli would have something exciting to announce- maybe a Samsung re-sign; maybe design wins, maybe progress with SSDs, whatever- but something with pizzaz.
No such luck.
There was no new news- but Eli was remarkably upbeat. Questions kept probing for specifics. Eli kept repeating that the outlook for the industry and SanDisk specifically was looking very positive- Boring generalities, on the face of it.
But boring can be good- especially if it leads to profitability. And that’s where SanDisk looks like its heading.
I’ve posted the JPMorgan transcript here.
In the live broadcast, the tech guys continued broadcasting after the official presentation was over: Lots of- “Hi how are you’s” and such, with one exception.
That exception was a Paul Coster (JPMorgan) monologue where he sounded like he was recording or calling in his summary thoughts, which went something like this:
“It feels like a positive one to three-year cycle. SSDs will mean that the industry’s lacking capacity in two years from now. Still aspires to a 15 to 20% operating margin, but it seems to be contingent upon the success of creating a consumer brand, which remains something that the company – a differentiated capability in the consumer space. Which is [inaudible]. The CEO accepted that, but believes that the many initiatives that have underway will ultimately yield to success. Here is cautiously optimistic around the near-term. So stated that the equity offering that was talked about on the fourth quarter conference call no longer is a pressing need.”
Pretty good summary, but there is a really big difference between one positive year and three positive years. One good year would be simply a brief respite and cause for a shrug. Three good years is a lifetime in this business and cause for breaking out the champagne.
A Positive One to Three Year Cycle
I suspect Eli is thinking three positive years. In fact, he said as much, before he went back on script:
“I think the current liquidity crunch and the current [pause] where the industry is, I think, is very, very good for the next three years. If you look at the – this is not just a classical cycle, industry cycle for flash. I think the fact that there is such a great liquidity crunch and, today, lack of return on investment on NAND manufacturing, that bodes very well for the next, I’d say, one to three years in terms of the industry [not] going back to that mode of exuberant CapEx investment.”
Time will tell how this plays out. It would be ironic though, if this current liquidity crunch turns out to be just what the doctor ordered for the NAND guys- SanDisk included.
Samsung is the gorilla in the room- the gorilla that sets the tone for NAND pricing and the gorilla that is currently in intense negotiations with SanDisk over re-licensing SanDisk’s IP.
NAND pricing has just about doubled from the start of the year. So all is OK there. New management at Samsung appears to have made a difference:
“We have new management at Samsung that has indicated a tremendous return back to focus on profitability. We – everybody is cutting back and I don’t think it’s just a temporary thing. But, I mean, the liquidity is one thing that is different in this cycle, and a liquidity crunch is one element that is different this cycle than in previous cycles. Nobody is out there that’s going to give you $5 billion to start a new fab unless the return on investment is there, and for that you need to have certainty about the technology, about your cost structure, about your competitors’ cost structures and, of course, the growth of the markets.”
SanDisk appears comfortable following Samsung’s pricing lead and doing its part on holding back on new capacity. The days of SanDisk talking up taking market share appear over- at least for the time being. Eli:
“We’re pretty confident about the growth of the future markets and we’re pretty confident about our competitiveness in that market. But clearly, our own direction is to not add a substantial amount of new capacity ourselves. To the extent that we are going to invest in the future, it’s going to be more in technology transitions, definitely in differentiation adding future star products, creating new markets, and trying to move more and more to a captive/non-captive mode where we can purchase some – become a consumer of other people’s flash rather than create excess supply.”
What will happen with Samsung on re-licensing is anyone’s guess. It appears that negotiations are intense, which I suppose is better than no talk at all.
FWIW my read is that SanDisk has put its best offer on the table and is now preparing for Plan B which would be legal action if an agreement can’t be reached by August 14, which is when Samsung’s current license expires.
“Paul Coster: Let me sort of focus on two more things and I’ll hand it off to the audience. Middle of this year a key point coming up, at least from an investor perspective, which is the rollover of the contracts, or otherwise, with Samsung on royalties on the licensing agreement. Can you talk about that in any way that will help us think it through the risks and rewards associated with that singular event?
Eli Harari: It is a singular event, because it’s an important part and IP is an important part of our strategy, although not the only thing that drives the company. We are in negotiations, intense negotiations I would say, with Samsung to try to come to a mutually acceptable agreement to renew the license when it expires on August 14, and that is really all that I can say…
Every negotiation has got to continue up to a certain point. Beyond that, you conclude that you are either going to have an agreement or you don’t. If you don’t, you don’t. When – if we don’t have an agreement by August 14, then the agreement terminates and they are unlicensed from that point on.”
Captive vs Non-Captive NAND Supply
SanDisk has been on record now for some time that it plans to move back to more of a captive/non-captive mode where SanDisk would purchase NAND from third parties- as needed.
I found the following comments most revealing. Apparently as needed translates as as needed for the retail market. OEM appears to be where SanDisk is planning to make its stand on GMs and business stability.
“ Eli Harari: That’s a good point and we’ve always said that the margins on non-captive where we have to buy, on the contract market, flash and then package it into a card and sell it, of course is going to be less than in a vertical integration situation even though today it’s opposite because people sell below cost.
Even after the restructuring, please understand that our captive capacity is still very substantial. It adds up to close to 1.5 million wafers a year of 300 millimeter capacity, all at 43 nanometer-capable of 32 nanometer. That, I believe, is the equivalent of Micron plus Intel plus Hynix combined, so it’s not small. It’s very substantial and when OEMs look at our captive supply after the restructuring, they see that we have a very, very strong supply base. The retail, basically today, benefits from that supply base, but in the future retail will have to be one, just like Kingston or PNY, or our competitors where they go and buy non-captive and we’ll go and buy non-captive and if we do a good job we will be able to compete.
The OEM is where we believe we can generate the margins in the future and the stability of the business. But it’s a combination of the OEM and the retail that gives us that flexibility to use non-captive, because you cannot use non-captive and be an OEM supplier. You cannot. You cannot quote, three quarters out, pricing when you don’t know what it’s going to cost you in the non-captive market.”
Although not addressed specifically, reading between the lines SanDisk was also talking about SSDs. Those who want to be significant SSD players will need to be NAND OEM players who have substantial captive NAND supply. Makes sense to me.
2011 sounds like its shaping up as the year when SSDs break through into the mainstream notebook and netbook markets and 24 nanometer will be what it takes.
“Eli Harari: Yes, the SSD market is a nascent market, in its very early stages, very price elastic, but only up to – only from a certain point on. You need to get a certain threshold price in order for it to take off – and if you’re not at that price point, you could play in the enterprise but not in the notebook or netbook market. So you are right, as the price goes up – I mean the spot price or contract price now for 16 gigabit chip, 2 gigabyte, is around 4 bucks, more or less, give or take. So it’s $2 per gigabyte; at the system level it will be slightly more than $2 per gigabyte and that’s – that means a 64 gigabyte SSD would be 130 bucks or so, which is too high. The competition for netbooks is 30 bucks for 80 gigabyte drive.
In enterprise, the cost is okay. They can afford to pay $2, $2.50 per gigabyte, no problem. The challenge for the NAND manufacturers is that as much as we want the the SSD business to take off, we don’t want to subsidize it. We can only go that far as far as getting it off the ground, and therefore it needs to wait, really, until the next two technology nodes. 32 nanometer- our costs will get a lot better. And 24 nanometers is where we can meet what the market needs, which we mostly think is around $1 per gigabyte- and still be profitable, still generate our margins.
So it’s still 2010, 2011 for the time it takes off. Once it takes off, the biggest issue will be that there will just not be enough capacity. There will just not be enough capacity.”
In a nutshell, it appears that thanks to mobile, SanDisk is optimistic about the next couple of years- before SSDs take off. Then there won’t be enough capacity for the next leg up driven by SSDs. No wonder Eli was in such a good mood for JPMorgan.