Zeev,
You stated...
"I think that the model for Flash was set in the 1995 to 1998 period for DRAM. The critical question is more on what portion of this volume does SNDK get royalties, IMHO. Since capacity additions are usually in quite big "quantas", even very rapid growth in end demand can be rapidly absorbed (some of the big one announced quadrupling capacity, if memory serves), meaning that there will be severe margin erosions. However, if SNDK royalties grows at only half that rate (a factor of 5 from the current base of about $50 MM/year?), we are looking at royalties revenues of $250 MM, which after taxes should be in the neighborhood of $175, or $2.5/share. To that you'll have to add whatever profits (or losses) their actual direct sales will add. Remember that the DRAM industry, in 1998 lost about $5 billions because of excess capacity, despite a growth rate of bit counts in the 70% to 90% annually."
I think your statement is an oversimplification, Zeev.
It reads worse than the safe harbor statements of the annual report. You need to separate the relationship between royalty payments, increasing incremental worldwide production "quanta", possible overproduction,... by the type of flash produced. I didn't see any attempt to tease this out in your post.
Second, we have just had our first profitable quarter based on operations alone (with royalty payments removed from the equation) and now you are suggesting that direct sales may lead to losses despite the excellent bookings, announced tripling of capacity, the migration to 256Mbit/D2 and upcoming die shrinks. I think that is unfair. It does not reflect any of the guidance given at the Shareholders' Meeting.
Finally, all royalty agreements are confidential. Any future estimates one may present are based purely on conjecture. Also, the '987 patents do not contribute significantly to the current projections of $12,000,000.00 t $12,500,000.00 per quarter for the balance of the year. If the '987 patent defenses against Lexar (and the companies it has supplied) and Mitsubishi are successful, royalty payments will be substantially higher than these estimates.
Zeev, you yourself have pointed out the manufacturing efficiencies SanDisk is currently implementing and the fact that they seem to be proceeding in a fairly consersative fashion with the current die shrink. You also did not comment on the feasibility of higher orders of MLC technology as described in Dr. Harari's last interview that we posted this week.
Are you a believer in "market elasticity" for flash memory (which is currently in its infancy in the consumer realm)? Any comments on oversupply must address this concept. Do you think you will own more removable flash memory megabytes than you own DRAM megabytes in your desktop and laptop? Help us put things into perspective rather than casting doubt on flash's prospects.
Why are you so down on SanDisk's prospects???
Ausdauer |
