06:36am EDT 26-Apr-02 Lehman Brothers (Niles, Daniel 415-274-5252) INTC AMD
Intel Corp: Great Execution But Demand Remains Issue (part 1 of 2)
PRICE: (USD 29.09)
EPS (FY Dec)
-----------------------------------------------------------------------------
2001 2002 2003 % Change
Actual Old New St. Est Old New St. Est 2002 2003
1Q 0.16 0.15A 0.15A 0.15A 0.22E 0.22E 0.22E (6) 47
2Q 0.12 0.15E 0.15E 0.15E 0.21E 0.21E 0.23E 25 40
3Q 0.10 0.18E 0.18E 0.18E 0.25E 0.25E 0.27E 80 39
4Q 0.15 0.21E 0.21E 0.22E 0.30E 0.30E 0.32E 40 43
-----------------------------------------------------------------------------
Year 0.54 0.70E 0.70E 0.70E 0.98E 0.98E 1.03E 30 40
-----------------------------------------------------------------------------
P/E 41.6 29.7
Market Data Financial Summary
-----------------------------------------------------------------------------
Market Cap 202699.1M Revenue FY02 $28.5B
Shares Outstanding (Mil) 6968.0 Five-Year EPS CAGR 25.00
Float NA Return on Equity NA
Dividend Yield NA Current BVPS $5.23
Convertible No Debt To Capital 4.12%
52 wk Range 36.78 - 18.96 Disclosure(s) A/C
-----------------------------------------------------------------------------
-----------------------------------------------------------------------------
-----------------------------------------------------------------------------
Rating Target
----------------------------------------
New:3 - Market Perform New: NA
Old:3 - Market Perform Old: NA
-----------------------------------------------------------------------------
INVESTMENT CONCLUSION :
* There were two takeaways from the analyst meeting in our view: 1) Intel is
executing the best it has in several years, but 2) there are no signs of an
IT spending pickup. However street consensus calls for a 7% q/q revs increase
in Q3 and 11% in Q4. We therefore remain cautious heading towards the summer.
SUMMARY :
* Convergence of computing and communications is driving integration and
Intel plans to build wireless connectivity into every chip in the future.
They want to go from #3 to #1 in marketshare for comm silicon over the next
several years. Intel also expects half of all the incremental processors
sold over the next five years to be into the Asian region.
* 300mm reduces cost per unit by 30% over the next two years and Intel
seemed to indicate they only need to grow cap ex by 10% or less in 2003 vs.
2002 unless street forecasts for demand are exceeded.
Intel currently sees very few signs of improvement in the IT sector but
expects it to recover before telecom. Given the recent downturn, the high
development costs for the latest 0.13u 300mm technology and the need for high
volume markets to drive economic feasibility, Intel believes they are
uniquely positioned to benefit when it does improve. They point to the
numerous other semiconductor manufacturers who have had to partner recently
as an example of this.
They believe that the world economies are still dependent on information
technology and this can be seen in the convergence of computing and
communications in every device. This is driving the integration of these
capabilities into every piece of silicon which Intel aims to provide. The
combination of the internet being delivered at broadband rates whether it is
wireless or wired should drive help drive this integration. Intel had a
working demonstration of their Banias next generation mobile platform to
illustrate this. It is basically a microprocessor engineered from the ground
up for low battery usage but with built in 802.11 connectivity and the
ability to make a VoIP call from your PC. The silicon was two weeks old and
the product is expected to be released next year.
Intel plans on integrating self detecting, auto configuring connectivity into
every piece of silicon they sell. They believe the analog circuitry is
trivial but that it takes up a lot of space while the digital circuitry is
easy and takes up minimal space. Using MEMS technology, they plan on
reducing the analog space. They believe that `free real estate` as processor
move to 0.13u on 300mm will enable radios on all Intel Silicon and have
dubbed it `Radio-Free Intel`. This feature should be in desktop processors
starting in 2003. Intel is also working with developers to have them design
once on Intel hardware and give them the ability to run it on any Intel
device. Intel believes that they are the third largest communications
silicon provider today with the goal of being #1 over the next several years.
About 30% of the R&D dollars are being spent on communications technology
development.
Intel had a demo on hyper-threading which will allow Intel to have two
virtual processors in one physical piece of silicon. Intel demonstrated this
technology last fall at IDF. They believe that in the server environment
there is a 30% performance increase and 10-30% increase in mainstream desktop
environments.
They believe these new technologies and focus on IT spending will be
particularly important in the Asian markets which have grown from 29% of
their revenue in 1991 to 40% of revenue in 2001. Intel believes that half of
all the incremental processors sold over the next five years will be into
this region. They also believe that emerging markets no longer buy the
trailing technology and point to China as an example of the first country to
move hard to PIV. They believe the market definition is changing from
emerging and mature markets to the 1st time buyers and experienced buyers
market.
Intel believes that 300mm rolls back costs by a decade and seems to indicate
they only need to grow 10% or less on capex in 2003 vs. 2002 unless street
forecasts for demand are exceeded. This was probably one of the most
interesting segments for us in the presentation given Intel had not even
given a hint before as to capex plans for 2003. Basically, they showed a
chart with capex as a percentage of COGS decreasing very slightly from 2002
to 2003. If we use our model where COGS increase by 10% on a revenues
increase of 15% in 2003, this implies that capex in 2003 is likely to grow by
only 10% or less. If our assumptions on gross margins increasing to 54.5% in
2003 are too high or Intel grows revenues faster than our assumption of 15%,
our cap ex forecast calculations would obviously be exceeded. This was in
the context of a long-term chart with 200mm construction running 49% of COGS
from 1992-1995 and then running at 35% of COGS from 1996 to 1999. They
expect 300mm construction to average 47% of COGS from 2000 through 2003 and
then to drop to below 35% for the next few years.
The goal is more capacity for less money with 300mm. On an oversimplified
basis, it costs Intel about 55-60% more to manufacture a 300mm wafer versus a
200mm wafer but they get 240% more die per wafer yielding roughly a 30%
reduction in die cost. On the lithography front, Intel plans to crossover to
0.13u from 0.18u in microprocessors in the current quarter. Intel also said
their shift to 300mm and their belief that they get 1.5-2.0 times better
total yield on their processors negates the die size advantage of their
closest competitor.
Though Intel is spending less this year than last on capex, a much greater
percentage is being spent on 300mm and 0.13u and below. In 2001 Intel spent
22% of total capital and 27% of fab capital on 300mm and 60% of total capital
and 78% of fab capital on 0.13u and smaller. In 2002, Intel is spending 56%
of total capital and 72% of fab capital on 300mm and 73% of total capital and
95% of fab capital on 0.13u and smaller. 300mm will be about 60% of desktop &
portable microprocessor production by Q4:03. With regards to Intel
announcing that they were ramping up their Ireland facility yesterday, this
was already in their capex forecasts for this year.
Figure 1: Intel`s Capital Expenditures
Source: Intel
At the end of the day Intel seems to be executing the best we have seen in a
while but the problem remains demand and high expectations. We note that
Intel right now is not counting on an IT spending pickup. We note however
that street consensus calls for a 7% q/q revenue increase in Q3 and an 11%
increase in Q4. We note that the average over the past five years has been
7% q/q growth for both quarters. In addition we believe that a P/E multiple
in the high 20s seems to discount some EPS upside to these numbers based on
an IT spending pickup. We therefore remain cautious as we head towards the
summer.
On a separate note, we gleaned some additional IT industry demand information
from Ingram Micro (IM - Not Rated) which reported Q1 earnings yesterday.
Ingram Micro is the largest global wholesale provider of technology products
and supply chain management services. The company operates in 36 countries
with more than $25 billion in sales in FY2001. IM announced revenues that
were in-line with consensus (down 9% q/q, down 22% y/y) and EPS of $0.10
which beat consensus of $0.06 due to margin improvements. The company noted
that corporate resellers continue to be relatively weak while VARs were
relatively strong. In regards to demand IM stated that the IT industry
continues to be affected by slow demand as companies defer spending and they
do not expect strong demand to return in the second quarter. The company`s
revenue guidance was for a modest sequential decline in Q2 (-5% using
midpoint of IM`s range), flat Q3, and a mid single digit uptick for Q4. This
is much lower than the sequential changes of 3%, 6%, and 9% implied by the
current consensus revenue estimates for Q2, Q3, and Q4 respectively.
Figure 2: Intel`s Income Statement
Company Description :
Intel is the largest manufacturer of semiconductor devices in the world.
Intel`s principal products are microprocessors that serve as the CPUs of IBM
compatible personal computers.
Related Stocks :
Company Name Ticker Price (04-25-2002) Rating
Advanced Micro Devices AMD ( 11.69) 3 - Market Perform |
