gs: US Technology Semiconductors: Samsung
'06 capex cut underscores our negative view
on SPE; DRAM spending will likely come
down from 8-yr highs and capital intensity
continues to decline
Samsung guided its ?06 semi capex budget down 11% Y/Y, below expectations that called
for flat to slightly higher capex in ?06. Our bottom-up capex model now stands at flat Y/Y
in ?06. With the SPE stocks having enjoyed a strong rally in recent weeks on expectations
of a strong ?06 fundamental environment driven by a +5-15% Y/Y increase in global
capex, the stocks should be weak on the news that the world?s largest spender will reduce
its capex in ?06. Samsung?s capex budget cut underscores why we are Cautious on SPE,
as: 1) we expect DRAM spending in ?06 to come down from 8- yr highs in ?05, 2) the
company commented that its capital intensity will decline in ?06, which we believe is the
biggest issue facing the SPE industry, and 3) management noted that 2006 capex is
heavily front-half loaded, implying that most of the orders associated with ?06 will likely
be placed by the end of March. We maintain our Cautious coverage view.
SAMSUNG'S 2006 SEMICONDUCTOR CAPEX GUIDED DOWN 11% Y/Y, BELOW
EXPECTATIONS. Samsung reported its Q4 earnings late on Thursday evening. The
company's 2005 capex budget came in at $6.2B, 5% above original expectations. More
importantly for the stocks, Samsung guided its 2006 semi capex budget (including
memory and System LSI, but excluding LCD) down 11% Y/Y in 2006 to $5.5B, below
Street expectations, which had called for the company to guide its capex budget flat to
slightly higher Y/Y. Although 2005 capex was a bit higher than originally expected,
aggregate capex for both 2005 and 2006 is below expectations.
In 2006, memory spending (which includes DRAM and NAND flash) is expected to
decline 13% Y/Y and system LSI spending is expected to decline 1% Y/Y. Samsung also
guided its 2006 TFT-LCD capex down 17% Y/Y to $2.3B from $2.8B in 2005. Recall that
we include spending for TFT-LCD in our bottom-up capex model as both Applied
Materials and Tokyo Electron sell equipment used in flat panel manufacturing facilities.
Other companies, including Advanced Energy and MKS Instruments, also have exposure
to TFT-LCD. Total capital spending, including spending for semis and TFT-LCD, came in
4% higher than anticipated in 2005, but was guided down 13% Y/Y in 2006. See Table 1.
Table 1. Samsung's 2005 and 2006 capex. US$ millions; we use an exchange rate of
1USD = 1,028KRW 2005 planned 2005 actual difference 2006 budget Y/Y % Chg Semi
$5,848 $6,160 5% $5,479 -11% Memory $4,291 $5,148 20% $4,457 -13% System LSI
$1,489 $963 -35% $954 -1% TFT LCD $2,783 $2,793 0% $2,306 -17% Total $8,563
$8,904 4% $7,717 -13%
Source: Goldman Sachs Research estimates, company data, www.XE.com.
SAMSUNG'S 2006 CAPEX CUT UNDERSCORES WHY WE ARE NEGATIVE ON
THE SPE SECTOR. As has historically been the case, we would expect some confusion
around Samsung's capex release for two primary reasons. First, the company provides its
capex budget in Korean Won, so different analysts on the Street are likely to be using
different exchange rates in translating the budget to US dollars (note that in our analysis
above, we use an exchange rate of 1 USD for 1,028 Korean Won). Second, different
analysts are also likely to choose to either include or exclude spending for TFT-LCD in
their capex estimates.
Our bottom-line is that the capex announcement is below expectations, with the company's 2006
budget cut underscoring why we are negative on the SPE sector, as: 1) we expect DRAM capital
spending to decline in 2006 from the 8-yr highs it reached in 2005, as DRAM companies that have
suffered a significant decline in DRAM prices in full-year 2005 attempt to alleviate some of the
excess supply that is expected to continue to plague the industry in 2006 (note that Samsung
management indicated that it expects DRAM to be in excess supply in 2006 after some tightness in
Q1). Importantly, we think Samsung's memory capex cut in 2006 could be a precursor to other
DRAM companies potentially cutting their 2006 capex budgets, which other DRAM companies (i.e.
Powerchip and Elpida) have implied is possible, although official budgets have not yet been
released. 2) Samsung management commented on the earnings call that the capital intensity of their
business will decline in 2006 (i.e. the company says they will get similar output in 2006 even
though they are spending fewer dollars), which we believe is the single biggest issue facing the SPE
industry. We believe that declining capital intensity is driving slower SPE industry growth, as capex
as a percentage of semiconductor revenues has continued to decline over time. And (3) While bulls
will likely argue that the company's commentary that wafer fab equipment will increase as a
percentage of capex in 2006 is a positive, we would argue that the company's indication that its
2006 capex budget is heavily front-half loaded (even more so than in 2005), implies most of the
equipment orders associated with its 2006 capex will have been placed by the end of March,
considering that lead times for equipment are at least 3 months. Again, bulls may argue that
Samsung's commentary is good for March quarter bookings, which we believe is certainly true.
However, we believe that it's nearly impossible to argue that these orders aren't priced into the
stocks given that the group is trading at ~23x last cycle's PEAK earnings. More importantly, if most
of the 2006 equipment orders from the world's largest capital spender are likely to be in the books
by March (if they aren't already), then how could one argue for the sustainability of an order
increase throughout the year?
While most of the analysts on the Street may try to turn a capex cut from the world's largest capital
spender into a positive, the we believe the SPE stocks should be weak on the news that the world's
largest capital spender is reducing its capex by more than 10% Y/Y, particularly in light of the
recent run-up in the stocks on expectations that 2006 global capex will increase +5 to +15% Y/Y.
There is no change to our view that aggregate capex in 2006 is likely to be flattish Y/Y, and we
don't believe a flattish capex environment will be good enough to support the SPE stocks, which in
some instances are trading within 15% of their previous peak share prices and at more than 30-35x
normalized earnings and above 20x previous peak EPS.
Each of the analysts named below hereby certifies that, with... |
