Hi Slacker,
This is just my analysis and opinion (related to SSB call) and I could very well be wrong.
1) I agree, there is a possibility that NAND could be in oversupply for a short duration (couple of q's) after the holiday season but I disagree with the degree of impact on SNDK. There are a couple of reasons for this- SNDK has a built-in cushion captive/non-captive. The moment they see oversupply, they reduce their non-captive purchases and that itself adds several GP %ge points which neutralizes ASP decline. Secondly, they are the lowest cost producer for at least next 2-3 Q's so on a relative basis it hurts others more than it does to SNDK, so it is unlikely that Samsung is going to increase production and not switch some of the lines to DRAM (where also they have cost and process leadership and major producer status). By increasing the NAND production and then dumping it in the market, Samsung hurts themselves more than they hurt Sandisk/Toshiba. Financially, it doesn't make sense unless they want to go after marketshare. I did a back-of-the-envelope calculations for eps estimates for SNDK and it seems that there is fair chance that they can make or not make 1.65 depending on how the demand shapes up from cell phone market. I used 2.25-2.5B sales, and used a lower GPM of 37% (vs 41.2 % last q just to be on safeside), opex of 15%, 2% dep/amort and 37% tax bracket (which I hope they are able to somehow bring it down, I dont know how though).
2) The unit sales may be off by 12% but more higher-end camera's are getting sold and some vendors are bundling much higher capacity cards (Canon for ex. when I bought my 4MP camera, I got 32MB card now they have 256MB card). Also, for 5-11MP camera's you need much higher density cards. So it is wrong to focus just on unit shipments.
3)What diminished marginal benefits? 90nm transition by Toshiba/Sandisk is the one that will save their butt and put them in driver's seat for the past and coming 2-3 quarters. Samsung doesn't even start 70nm 4GB SBC production till 1Q05 so there is no match between them and Sandisk/Toshiba here. If ASPs decline, this is what will help them keep their margins high.
4)Hopefully by then, Cell Phone market will kick-in!
5)Unit shipments for cell phones is 650mm/year much bigger than DSC which had 45mm units last year and are expected to be 65mm this year! The initial card size requirement is going to be small but will accelerate fast and remember the old cards cannot be used in cell phones as they are all RS formats!
As for your worry about Sandisk not having capacity- Eli has talked about them already working on 70nm and 55nm (last conf call). Also if there is a glut, why would we want more of our own capacity? If Sandisk can produce exactly as they can sell, our margins would be higher than buying some from outside! Beautiful model, isn't it?
Finally, last time I checked Citicorp (SBarney, et al) owned 10% of the stock so I fail to understand the timing of this call just before a nice eps announcement (or it doesn't take a rocket scientist to understand the motivation behind this call).
That's my 2 cents worth,
Pam
PS: As for the decline of 60-70% in Flash memory prices, these analysts do not distinguish between NAND and NOR markets. I wonder why? |
