【出版时间及名称】：2010年1月台湾内存条行业研究报告
        【作者】：摩根斯坦利
        【文件格式】：PDF
        【页数】：49
        【目录或简介】：

Where we Differ? The market seems to be pricing in a
more profitable 2010 for Taiwan DRAM, and not
focusing on the need for debt repayment, while
expecting 2009’s strong stock price rebound to continue.
In 4Q09, stock prices rose 30-70% vs a 30-40% rebound
in DRAM prices. DRAM stocks appear toppish near term
as DDR III pricing is peaking and DDR II is declining
given the DRAM price/box is nearing 10%. We suggest
booking profits into January’s rally and revisiting in 2Q10
for a 2H10 recovery.
Our Pick. DRAM backend foundry Powertech is our top
pick on strong volume growth in 2010 (75%+ bits growth
for Taiwan). We expect DRAM price to decline gradually
in 2010 after a sharp rebound (~200%) in 2009. We
downgrade Inotera/Winbond to EW after 40-60% run-up
in 4Q09. Inotera is raising NT$14bn (~US$440mn) in
1Q10 (16% dilution). We remain EW on Nanya on
valuation and UW on Powerchip on weak balance sheet
(350 net % debt/equity; US$133mn ECB due 1Q10).
What’s New? We expect a correction in DRAM prices: 1
gigabit DDR II spot price from the peak of US$2.75 in
November 2009 to below US$2 mark by late 1Q10; DDR
II contract price has shown its first reversal since it
started to rise in February 2009. By late 1Q10, we think
DDR III prices will follow DDR II for correction.
What’s Next? DRAM makers’ view: 1) 2010 will be
good – Supply stays in check after multi-year underinvestments,
bottleneck equipment on allocation, yield
loss on technology migration. Demand is strong driven
by PC upgrade cycle on Windows 7 and resumed
corporate IT spending after 2009’s slump. Our view:
Demand expectation is high for 2010 leading to eventual
higher supply growth. In the past even years (World Cup,
Olympics) when demand expectation is high, DRAM
stocks have higher correction risks (1996 semiconductor
downturn, 1998 LTCM crisis, 2000 Y2K bubble burst,
2002 double dip risk, 2004 China macro management,
2006 Vista hype, 2008 Lehman). Macro impact on
government stimulus exit is an unknown risk beyond
microscopic framework on DRAM.