From Soundview (PDF file contains charts and additional info)
PDF found at:
research.soundview.com
2003 Semiconductor Manufacturing Outlook
January 6, 2003
300mm Production Ramps Should Come Into Fruition in 2003
2003 looks more and more likely to be the year in which 300mm production comes into its own. We are
tracking between 20 and 25 300mm projects that could come into fruition in the next several years. In our
view, chip designers could begin to migrate designs to 300mm as the process becomes more stable. Intel will
continue to invest in 300mm, but we are also looking for both TSMC and UMC to begin to build out their
respective 300mm facilities to much higher production volumes. Additionally, we believe that Japan chip
companies are looking at 300mm as a potential springboard to bring them back into a competitive position vs.
the rest of the world. Elpida, Toshiba and Trecenti should all see ramps of production capacity in 2003. In the
United States, both IBM and Texas Instruments remain committed to 300mm in the nearer term, while Micron
is looking for strategic reasons to get more aggressive on 300mm. In our view, more than 50% of the equipment
shipped in 2003 will be for 300mm. In 2002, we believe this percentage was closer to 30%.
The Challenges Remain Vast and the Opportunities Significant for Equipment Companies
After a decade of extraordinarily robust spending, the industry needed to retrench and work off its excesses.
We believe that fab closures, as well as continually increasing chip volumes should enable utilization rates to
migrate higher in 2003. We believe that about 7%−10% of fab capacity has or will be shut down. Most of this is
trailing edge technology, and this is a typical outcome of overspending. Even so, technology continues to
advance and leading edge technology must still be put in place if the semiconductor roadmap is to be adhered
to. Thus, 300mm, copper, and 0.13/0.09 micron linewidth production ramps will continue to be necessary.
In our view, the industry is still a double−digit grower, with the likely long−term growth rate at 10%−15%.
While this would be slower than the prior view of 15%−18%, we believe that the industry’s ability to be more
profitable in up and down turns suggest that the prior multiples applicable to the group can remain intact.
Our Favorites
Given our view that robust growth can return to the semiconductor manufacturing space, how do we best play
the opportunities? In our view, those that are best positioned to take advantage of the 300mm transition, the
move to copper and other new materials, as well as the continued trend toward outsourcing are the companies
that can benefit most in the next growth cycle.
In the large cap front−end companies, we would rank our favorites as Applied Materials and KLA−Tencor.
Applied appears to be executing well in almost all areas, with only its electroplating segment in the process of
new product design to compete against Novellus. In the process diagnostics area, the company’s share in
300mm could be substantially higher than 200mm, and we believe that its inspection tools are competitive with
KLA’s, although its product offerings are still somewhat limited. Thus, we believe that Applied can at least grow
with the market in this upturn, with faster than market growth predicated on success with process diagnostics
as well as continued penetration of its "Total Service Solutions" offerings. All in all, we would look for Applied
to reach about 80% of its prior fundamental peak in this upturn.
KLA−Tencor is one of the large cap names that we believe can reach higher fundamental highs vs. the last
cycle. From our vantage point we believe that process diagnostics will continue to be one of the highest growth
areas in this cycle. KLA remains the dominant player here, and although Applied is more competitive and some
share loss will result, again it is against a small percentage of KLA’s total served market. Thus, the company will
offset any small share losses by the growth of the overall sector.
In the smaller cap side, we continue to have faith that flat panel spending will remain robust and Photon
Dynamics is well positioned to take advantage of this. The company continues to enjoy dominant share in its
test systems for flat panel manufacturing, and we believe its PCB and high−quality glass businesses are finding
bottoms. The flat panel investment patterns will continue to be volatile, however, given that there is a very
limited number of players. We still believe that $1.50−$2.00 of earnings power is achievable in this cycle as
additional Gen 5 and Gen 6 panel fabs are constructed. If a PC replacement cycle does take hold, we believe
that it is unlikely that new CPUs will be attached to CRT displays, but rather flat monitors are the more likely
decision by IT managers.
2003 Will Be Another Interesting Year, but We Believe It Will Bring More Smiles
The biggest obstacle to our recovery thesis is whether the overall macro economy recovers in 2003. While we
are not expecting a significant contribution from communication infrastructure spending, we do believe that
the PC and wireless handset markets are key to the business environment improving. Thus, given that the
equipment stocks trade off of order momentum, we believe that investors should be building positions now, to
take advantage of at least the next few quarters of order improvement. If end markets do not cooperate, then a
rally could fizzle, but we would be more constructive at that time. In our view, valuations are not extended at
these levels, and we believe substantial gains could be made. |
