The Spansion Expansion

Having lost so much sleep over taking money out of one investment and putting it into Spansion (SPSN), I'm now ready for this stock to go way up.

No fooling. With SP1 running 2000 wafers/week, I think the chance that STMicroelectronics (STM) can bring M6 online is low. This week's news from Hynix suggests that their JV fab won't be available for NOR either. And Samsung's NOR is pure 200mm.

The fact is, no sensible investor would put money into NOR now that Spansion has done so. Say M6 is equipped: It's now late 2008. SP1 is probably running 4000 wafers/week. It's about to transition to a 45nm process node. No other NOR or DRAM manufacturer is even talking about that node, that soon. And all this is happening on a fundamentally cheaper technology. Can you see it? A nightmare!

What ST is doing right now looks like smoke and mirrors to me. Consider it: they're telling the press that Catania will be 300mm ready, and soon. They must be telling Francisco Partners that Phase Change RAM is a revolutionary new architecture. They even told told me to pay no attention to the man behind the curtain!

All the while they're telling their own shareholders that ST is nearly free of NOR and $300 or $400 million richer into the bargain. It's magical really. Everybody wins.

Though I could be way wrong and just wishfully thinking, I can't help seeing Numonyx fading out like a baby bell would. I can't help seeing one third of a $7 or $8 billion dollar market being carved up, say, half half between Spansion and Samung. I just can't help it.

Disclosure: Author holds a position in SPSN

Comments

[graize chen:]

not so sure, of course Numonyx will has her issue,but also spansion. the problem is not whether you got a 12mm fab or not, the key point is that who can survie in this cut-throat Nor flash competition. we should keep it also in mind that Spansion has been losing money for how many quarters? spansion is also loosing money, dangerous!

personally i think the Nor flash market will be dominated by 3:Numonyx, Spansion and Samsung, (by market share), then let's see who will smile?

2008 Mar 12 10:27 AM

[GCW:]

At the low volumes of SP1, there is very little economies of scale. That plus the fact they are over 1 year behind Numonyx on the process technology means that the purported cost savings are questionable.

They do, however, have an excellent marketing department.

2008 Mar 12 11:00 AM

[Omer Pelman:]

thank you!

but i don't agree they are 1yr behind. they were. but they are now caught up or just behind on the process. which means well ahead on the die size, because of the charge trap.

2008 Mar 13 05:10 AM

[graize chen:]

as i know, numonyx already had the 65nm MLC, and spansion had 90nm mirrorbit, frankly speaking, i don't knwo which one will has smaller die size.

2008 Mar 13 10:20 AM

[MemoryExpert:]

There is no question that Spansion is out in front of Nor vendors in fab technology. But the problem is that only a small portion of their portfolio will migrate to the new process. And that is only the high density parts. The low density parts don't see any cost reduction when they shrink and the high density parts have to compete with NAND pricing. All parts also have a long qualification cycle they have to go through before customers will use them which will push out any real cost benefit from the new fab 6-12 months. While I agree Spansion is in relatively better shape than Numonyx, it's kind of like saying GM is in better shape than Ford. I wouldn't want to invest my money in either.
For Spansion's stock to turnaround there would have to be a major shakeup in management and a significant restructuring around a smaller lower density market. And they better do it fast cause with their debt load, cash position and burn rate they are headed for bankruptcy in 8-12 quarters.

2008 Mar 13 02:18 PM

[Omer Pelman:]

i think there are a number of upsides, which makes bankruptcy unlikely, in my opinion.
1) they have been taking share from Numonyx et al. and will continue to take share from any 200mm producer. with only have 1/3 of the market right now, what's to stop them being the Samsung of NOR?
2) they have a data product that they at least believe can compete with a low density nand.
3) they have the eclipse product, which looks to me like a very cool thing. it gives the oem's the chance to build mutiple products around the same "engine," if you will, by just changing the software. that led to savings at all the automakers, and it strikes me as a great value-add for phones too.
4) SMIC is finally making progress on a NAND device using Spansion's/Saifun's IP.
5) DRAM market is a potential target if Spansion manges to shrink to 45nm by '09.
Of course there is always the risk nothing works. But with so much potential it looks less like a Ford and more like Boeing before the announced the Dreamliner.

2008 Mar 13 04:33 PM

[FABulous:]

Spansion will dominate NOR provided they quickly ramp their 300mm fab in Japan (SP1). This is gives them a huge advantage over the competition, not to mention their technology allows them to move towards more advanced technology nodes. Look for Spansion to make some waves in the next 6-12 months.

2008 Mar 14 02:58 AM

[jsam:]

Good thoughts and analyses, with one key tragic flaw. I agree numonyx is not headed anywhere (though it will take a very long time to fall). I agree Phase Change is dead (anything after 20 years of development, that's still not manufacturable, ain't got it). I also agree that Spansion is in the leading process position.

The tragic flaw? Exactly that: Spansion is in the leading process position. Therefore, they should also be the cost leaders... but they are not. This leads to one conclusion only: that the costs of ramping mirrorbit are staggering. It's not an easy technology by any stretch, and the costs of research, development, and ramping to high yeild can extend through the life of the product itself. Being a leader in fab technology is often associated with being a cost leader; but this is only true if the technology is manufacturable.

2008 Mar 14 03:19 PM

[MemoryExpert:]

Here is why your points are just wrong:
1) they have been taking share from Numonyx et al. and will continue to take share from any 200mm producer. with only have 1/3 of the market right now, what's to stop them being the Samsung of NOR?
Samsung is a broad based memory supplier capable of mixing technologies and weathering ups and downs in different markets by having exposure to different markets. Samsung can shift production and investment between Nor, NAND, and DRAM as the market dynamics shift. Samsung has a lot of money in the bank and relatively little debt. Since the Nor market is shrinking nearly 10% per year they would have to gain 3% market share every year just to stay flat. Eventually that game runs out. The Nor market is starting to look like the EEPROM market.
2) they have a data product that they at least believe can compete with a low density nand.
Most of the profits in NAND are high density. That's why the cost curve bottoms out below 4Gbits.
3) they have the eclipse product, which looks to me like a very cool thing. it gives the oem's the chance to build mutiple products around the same "engine," if you will, by just changing the software. that led to savings at all the automakers, and it strikes me as a great value-add for phones too.
A good idea that has no second source and if successful would take quite a while to shift over. Only software changes is a huge barrier, particularly if there is no clear benefit. Eclipse also does not deliver the programming and erase speeds of NAND. Although that may not be absolutely a killer. I challenge you to find any phone in the world using this stuff before you bank on it to turn the company around. Even any phone manufacturer willing to endorse and announce plans to use it.
4) SMIC is finally making progress on a NAND device using Spansion's/Saifun's IP.
Haven't heard that but until a product is in the market it will only add to expenses not the top or bottom lines.
5) DRAM market is a potential target if Spansion manges to shrink to 45nm by '09.
Except for the whole latency, bandwidth thing they could be cost competitive at 45nm in '09 vs. 65nm DRAM. So they could sell 1Gbit high end Nor to compete with DRAM for what $1.30? Have you seen DRAM prices lately?
Of course there is always the risk nothing works. But with so much potential it looks less like a Ford and more like Boeing before the announced the Dreamliner.
What was the last time Boeing lost money for 16 quarters straight after announcing they would be profitable time and time again?

2008 Mar 14 04:22 PM

[KT:]

Looking at the NOR market you have: (1) Spansion, (2) Intel, (3) ST Microelectronics, and (4) Samsung. While its not certain how Samsung is faring, its a well known fact that Spansion, Intel, and ST Microelectronics are all losing tons of money on NOR. Intel said it missed its earnings by $0.02 due to $0.025 loss related to NOR. If you compare quarterly loss rate, Spansion is lossing about $65M while Intel lost like $250M. Intel and STM are sick of the losses so they are getting rid of these two companies. Spansion has been separated from AMD for years; when Numonyx is formed it will also have to fend for itself. There will a mutual understanding among these two new spin-offs that prices have to rise. Samsung is a wildcard with its NAND-based fusion memory OneNAND. It looks like Samsung got some good NOR penetration in Q1/Q2, but this reached steady state in Q3/Q4. We need some sane pricing in the NOR market and I think that will come soon. Spansion/Intel/ST Microelectronics (NOR) are all under water, but Spansion is more efficient and will break the water first once prices start to rise.

2008 Mar 16 11:19 PM

[consider th...:]

best of luck to you on this one. i think the equity is a zero. have you built a financial model on this company?

let's do some simple math...
1Q08E:
EBITDA 70
Cash interest -30
Capex -250
Cash taxes? 0
FCF -210
Cash 12/31 416
Cash 3/31 206

uh oh, they need $200mn cash to run this business. fast forward through the next few quarters, when the mandatory 17mn/qtr spansion japan loan amortization kicks in and things get tight quick. sure capex is coming down dramatically, but i'm afraid that won't be enough. they'll be tapping their revolver by mid-year, an event that doesn't usually signal an equity rally. and if the bleeding continues beyond that? let's see, their credit facility is only 175mn, they can't issue more debt (secured notes issued just last year already trading at 70) and I don't need to say what the effectiveness of issuing stock at $2 would be.

semi industry success takes more than wafer size increases and node shrinkage. it also takes being in the right market at the right time with the right product and right capital structure. they might have a good product, but their market sucks, the timing sucks and their capital structure is not suitable. 1 out of 4 doesn't get my money.

with so many companies out there to invest in, how on earth somebody singles this one out as a great idea is beyond me. if you really want to own spansion equity, you should buy the bonds and clip the 11.25% coupon until they file for bankruptcy - your bonds will end up being exchanged for new equity after it gets restructured anyway...

2008 Mar 19 02:54 PM

[rbtbob:]

It's all about the PCM > > > Numonyx > > > Ovonyx ---duh!

2008 Mar 27 10:36 PM

[MemoryExpert:]

Uh Oh...I'm guessing you will be taking a hefty loss tomorrow.

2008 Apr 08 08:46 PM

[BABUYAYA:]

your anaylsis is all wrong

2008 May 14 01:09 PM

[MemoryExpert:]

For your sake I hope you listened to me back in March.

2008 Oct 22 07:47 PM

[Redde Rationem:]

Reality:
PRESS RELEASE
Fitch Downgrades Spansion Inc.'s IDR to 'CCC'; Outlook Negative
Last update: 5:39 p.m. EST Dec. 12, 2008
(..)
there is a real possibility Spansion will default over the near-term, due to a combination of a challenging macroeconomic environment and the company's weak liquidity position. Fitch believes Spansion's survival over the next 12 months will depend upon a combination of better than expected demand and pricing environment,
(..)
vs.
Fantasy
"I can't help seeing one third of a $7 or $8 billion dollar market being carved up, say, half half between Spansion and Samung. I just can't help it. " (Omer Pelman, this web page)

2008 Dec 20 04:19 PM

[Stocks007:]

SPSN just missed their coupon payment today. Sometimes I hate being right (for your sake, this time).

Jan 15 02:35 PM

[patently-ab...:]

ah crap. Deeeeeeelisted, for the win.

May 05 09:04 PM