本帖隐藏的内容
本帖隐藏的内容
We have revisited our global memory model and changed our major
assumptions given meaningful changes to both supply and demand. As a
result, we make significant upward revisions to DRAM revenue thanks to
increased ASP assumptions resulting from ongoing tight supply, while our
NAND assumptions remain broadly unchanged. Hence, we turn positive on
DRAM and remain relatively cautious on NAND.
DRAM should remain healthy due to structural changes: Despite strong
DRAM prices, we do not expect capacity expansion from the remaining
three players (SEC, MU camp, and SK Hynix), a difference from previous
cycles. We expect demand from server and mDRAM to offset muted PC
DRAM demand. Hence, the decline in blended DRAM prices should be
well below the historical average of 30%, and major DRAM makers should
continue to expand their margins.
mDRAM market will expand into PC: Despite a slow down in high-end
mobile devices, we think unit growth in both smartphones and tablets will
continue to drive mDRAM demand. Also, LPDDR3 will break into the
Ultrabook memory market on the back of higher bandwidth and efficiency
on top of lower power consumption. Hence, in 2014 we expect mDRAM
demand to comprise 38% of total DRAM demand which is comparable to
PC DRAM’s contribution of 40%
NAND will move into moderate oversupply in 2014: We believe NAND
supply/demand dynamics will improve in 2H13 due to seasonal upticks in
demand and impact from SK Hynix's NAND capacity conversion. In 2014,
however, we think NAND market will be in oversupply due to slowing
high-end smartphone market and relatively high supply growth from
capacity expansion. Hence, we forecast global NAND revenue growth to be
flat in 2014 after double-digit growth in 2013.
All about SSD for a game changer in NAND demand: Due to inferior
product mix changes in mobile devices, we expect NAND content growth to
decelerate in 2014. Given still high prices of SSD relative to HDD, NAND
prices should come down to stimulate NAND demand in NBPC market.
Although 3D V-NAND based SSD will enable denser, faster, and longer
lived SSD, it is likely to remain a small portion by 2014.
Stock implications: With changes in our major assumptions, we upgrade
SK Hynix to OW (see separate note published today) given that it has the
highest exposure to DRAM spaces, and maintain our Neutral ratings on MU,
Toshiba and SNDK. As the global number 1 DRAM maker, SEC will also
benefit but the impact on its earnings is likely to be relatively small given
that non-DRAM businesses (especially handsets) account for a meaningful
share of its total OP.