Art,
re: SSB's J. Joseph The upgrade came from none other than Jonathon Joseph, who earlier had bashed the entire group, causing SNDK to travel with the rest into a black hole which appeared to have no bottom. Suddenly he's decided that it can't be all that bad, since semiconductors are the engines that drive economic and productivity gains throughout the world. Do you think that the portfolio managers who rely on these analysts have already added semiconductors to their accounts? Do you think that the opinion just made public will help protect those accounts? All I'm saying is that you can depend on analyst comments to be short on logical content and long on expediency.
You seem to imply that his comments last summer were the reason that the semiconductor stocks tanked. The reason they went down was that the fundamentals went to hell. Joseph was researching the channels better than any other analyst (or any poster on SI that I noticed) and made the bold call, when everyone else was bullish, that a cyclical peak had been reached. He was 100% right, and if you or I had listened to him, and taken a defensive posture, we could have saved a ton of money. But he certainly didn't cause SanDisk or any other stock to go down, the lack of sales caused the stocks to go down.
Now is he making a positive call on the sector, not because "semiconductors are the engines that drive economic and productivity gains throughout the world", but because things are so bad he doesn't believe they can get worse, so he feels we are at the bottom of the cycle. Time will tell if he is right, but again it's a ballsy call.
IMO he is doing his job better than any analyst on the street. What would you have him do, only issue good reports?
re: You express some skepticism about lowering the cost of chip production without corresponding improvements in technology, such as going to .18 or .13, and moving from 200 to 300 mm wafer size. While these new technologies represent quantum leaps in productivity, they are not the only source of improvements. What I was referring to was the concept of "yield," which Harari has mentioned often in his conference calls. You open a new factory and you're lucky if you can use even 10 percent of the wafers. As things get under control, there is a learning process resulting in fewer blemished wafers. As yield goes up, the average unit cost goes down. This alone has reduced chip cost more than anything else, since there are so many new fabricating facilities that have begun production within just the last two years.
The reason the yields per wafer go so low is bacause they are switching to a new, finer process. You are right that with time they learn to tweak the new process to get better and better yields. If they stayed on the current process, yield would max out. Only by going to a new process (smaller die per unit) can they increase potential yields per wafer and dramatically reduce costs.
John |
