The analyst problem, as usual, is grouping every type of semiconductor chip into one category to make the analysis easier to understand (but invariably wrong). Looking at the demand for microprocessor chips and DRAM--two types of semiconductors that find heavy use in desktop and portable computers, there has been a drop in demand. Looking at flash memory, there has been a drop in the RATE OF INCREASE in demand. That is, demand for flash memory is still increasing at a rather strong rate, but it isn't quite as strong as it was a year ago. In 2000, OEM camera manufacturers were worried about adequate supplies of flash memory, so it appears they ordered more than they needed to make sure they had enough. Then late in 2000, as growth in demand for digital cameras subsided, the camera manufacturers were caught with too much inventory and probably cancelled ongoing orders to adjust their own inventory, leading flash memory producers to make similar cutbacks and inventory adjustments.
The point is that overall demand for flash memory chips is increasing, and in fact is the only really healthy segment of the semiconductor sector. By grouping all semiconductor companies together, analysts are missing some rather good investment opportunities in the strongest growth part of the sector. It seems to me that the three companies which are able to benefit most from flash memory are Toshiba, SanDisk, and Advanced Micro Devices. Only SanDisk is the pure play here, as any earnings increases in the other two would be dwarfed by the rest of their business. AMD appears to be taking a commanding position in embedded memory, particularly for applications in set top boxes, routers, cell phones, etc. But revenues from these sources are still small compared with AMD's microprocessor business for the PC market.
Art |
