Reader,

This is our second edition of BlueLake Partners' Newsletter - The Perspective from BlueLake©. The Newsletter focuses on and analyzes trends in the sectors we follow: Software, Semiconductors and Materials, Enterprise Storage Networking, and Communications. We hope you find it informative and thought provoking and we welcome any suggestions or thoughts you might have on the content. Please feel free to pass it along to others that you think would find it interesting. To Unsubscribe or Subscribe, please click here.

Sincerely,

BlueLake Partners, LLC

As we are now in mid 2002, what are the trends in the industry and what does it mean for us going forward.

Where are we?

Foundries
First of all, the success of the foundry model is clear, at least for the leaders. TSMC and UMC have the scale to support advanced process development and fab costs that other participants simply cannot match. Even large Integrated Device Manufacturers (IDM's) such as Motorola with their "asset-light" semiconductor manufacturing strategy (a strategy the limits internal manufacturing assets and uses outsourcing manufacturing strategies) see the benefit in limiting their manufacturing base and working in close partnerships for both advanced process development and production. Market research firm Gartner Dataquest had previously forecast that chip foundries will account for 40% to 50% of global semiconductor output. So foundries have replaced many IDM's as equipment purchasers resulting in a concentration of equipment purchasers. Most recently, it has been the strength of the semiconductor foundry industry expansion in Mainland China that has been one of the brighter lights. However, as Morris Chang, chairman of TSMC noted in February 1999, "The foundry industry will have its share of winners and losers, just like the rest of the semiconductor industry." The first round of false starts for Southeast Asian foundries (Wafer Technology and InterConnect Technology in Malaysia as well as Thailand's Submicron Technology) and the inconsistent performance of other foundry participants reinforces that execution and scale are still critical factors.

300mm
After many years, 300mm is finally on the production scene. It is expected that fully 50% of future equipment sales will be 300mm systems. But while 300mm may represent half of equipment sales, during 2002 it will only represent about 7% of total wafer demand according to a recent report from VLSI Research, Inc., against 50% for 200mm. As Steve Appleton, CEO of Micron Technology, noted in July 2002, “We [Micron] have to be prepared to do 12-inch wafers. But it doesn't appear that 12-inch will be competitive with 8-inch for a couple of years.” In addition, there are dynamics that come with 300mm systems. First, they are considerably more expensive due to both the complexity of the larger wafer size and, more importantly, because 300mm is automated. A 300mm wafer boat is too heavy for human intervention, and fab utilization and process control demand automation.

Second, the population of buyers is much smaller for 300mm equipment for two reasons – expense and volume. Due to the high cost, only the largest and best capitalized IDMs and foundries can afford 300mm lines. Additionally, only the high volume IDMs and foundries will actually reap the benefits of the enhanced unit economics of the 300mm wafers. This has impacted the equipment suppliers as the largest suppliers consolidate their hold on the major customers with organizations that can provide a higher level of support and more complete solutions. Many second and third tier suppliers are now focused on “niche” applications, places where there is inadequate volume to support the large competitors.

Technology
Technology remains a driving factor and will push more volume to the foundries. Notwithstanding rumors about UMC holding back up to a third of their capital expenditure budget because it is technology related and “may not be necessary” in the current environment, the drive for greater performance and productivity is unrelenting. Copper and Low-k are firmly on the roadmap as well as solutions for 130 and 90nm devices. While over a decade ago some pundits said that X-ray would be the only way to continue with shrinking geometries and current experts look to EUVL (Extreme Ultra-Violet Lithography), projection ebeam and other approaches, optical lithography continues to drive to smaller geometries. This forces equipment suppliers to develop increasingly complex solutions and pushes IDM’s towards the leading foundries for manufacturing processes they cannot afford to develop in house. The cost and performance advantages of the smaller geometries will punish those who fail to remain competitive with their peers and in the short term (three to five years at least and probably more), it will be optical processes that manufacture production product.

The Search for the Next Driver
If history tells us anything, it tells us that the Semiconductor Industry is cyclical……period. Unfortunately, limited visibility of what will support the next up cycle has shaken the confidence of both industry management and investors. The sharp sell off in the financial markets in the late spring and early summer was in response, not so much to the Enron and WorldCom disasters, but to a lack of understanding of why the foundries were reporting ever increasing utilization and equipment suppliers were reporting improving bookings and book-to-bill ratios amid a general consensus that an order pause would occur in 2H02. The PC and telecomm markets, the drivers in recent history, at the macro level simply do not support the thesis that a recovery is under way. Recent news stories paint a confusing picture, on the one hand we have recently heard:

• TSMC reports an 83% increase in year/year revenues
• UMC reports a 58% increase in year/year revenues

and on the other hand:

• Intel reports it is seeing softness in PC demand
• Nokia is lowering their growth expectations from 15% down to 10% and
• Both TSMC and UMC have announced they expect lower fab utilization
   in the third quarter compared to the second.

Notwithstanding selected media frenzy, economic indicators point to an improving economic climate, though not a super-heated one. Further, it is unlikely that the “killer app” that everyone is looking for will materialize until an up cycle has firmly revealed itself.

Leading Indicators
After declining 15% in 2001, the worldwide blank silicon wafer market is expected to grow 7% to 3.997 billion square inches (bsi) in 2002, according to a new report from VLSI Research Inc. At the Advanced Reticle Symposium in San Jose in June, Klaus-Dieter Rinnen, chief analyst and director of market researcher Gartner Dataquest’s semiconductor manufacturing group said: "It’s official, the semiconductor industry is in recovery." Noting that the rolling 12-month average of chip sales has finally reached an inflection point and is starting to rise. He also noted that the industry has now seen two quarters of sequential improvements. According to the Scottsdale-based market research firm IC Insights, Inc., worldwide fab utilization rates are projected to jump from an average of 72.2% in 2001, to 83.5% in 2002. But worldwide fab capacity will remain flat in 2002 over 2001. “Although MOS capacity is forecast to be flat in 2002, MOS wafer starts are forecasted to surge by 16% in 2002,” according to IC Insights. Recent Semi Book-to Bill and raw bookings and billings figures all point to improving conditions but against a backdrop where there is little confidence of sustained momentum.

Where are we going?

This brings us back to the impact of these trends on the industry. The universe of chip manufacturers willing and able to buy leading edge technology is becoming more concentrated, creating a smaller number of purchasers making larger buys of equipment. This smaller population of buyers recognizes and rewards the leading equipment suppliers who have the financial stability to maintain momentum during down cycles and continue the funding of necessary R&D and customer support organizations. Given that chip manufacturers tend to select a single “tool of record” for any process step, the limited number of large volume “design wins” inevitably falls to the leaders in the respective process segment marginalizing the smaller suppliers. At the same time, the economic slowdown and cost of product development for equipment suppliers has compounded the thought that size matters.

Size Matters - Market Valuations By Sector
As of 12/31/01 - Ranked by LTM Sales - $s in Mils, except per share amts.

Deposition Price 12/31/01 Enterprise Value LTM Revenue Ent Val/ LTM Revenue Applied Materials $ 10.10 $ 29,465.4 $ 7,343.2 4.0 Novellus Systems $ 39.45 $ 5,699.3 $ 1,582.0 3.6 ASM International $ 19.51 $ 967.9 $ 643.5 1.5 Applied Films $ 31.25 $ 258.0 $ 137.1 1.9 Trikon $ 11.75 $ 138.1 $ 119.1 1.2 Genus $ 2.43 $ 55.0 $ 58.7 0.9 Average Top Players     3.8 Average of others     1.4 Premium     178% Clean Etch/Strip/CMP Applied Materials $ 10.10 $ 29,465.4 $ 7,343.2 4.0 TEL $ 48.87 $ 8,560.5 $ 4,694.0 1.8 Lam Research $ 23.22 $ 2,873.4 $ 1,554.3 1.8 Axcelis $ 12.89 $ 1,123.9 $ 504.9 2.2 SemiTool $ 11.48 $ 286.7 $ 256.6 1.1 Mattson Technology $ 8.81 $ 319.2 $ 229.4 1.4 FSI International $ 9.22 $ 187.2 $ 199.2 0.9 Speedfam IPEC $ 2.98 $ 166.5 $ 177.1 0.9 Tegal $ 1.34 $ 10.6 $ 30.2 0.4 Average Top Players     2.6 Average of others     1.2 Premium     121%   Source: SEC Documents and BigCharts.com

The market will reinforce its trend of leaders (at best two to three suppliers in any given space) and niche players. The capital markets reflect this reality as the valuation metrics, such as P/E and Price to Sales Ratio (PSR), between the leaders and the rest of the pack have significantly different valuation multiples. This gives the leaders a huge financial advantage, in addition to their size advantage in the industry and niche players are challenged to support them. Selected participants are trying to capitalize on this as witnessed by the Brooks/PRIA and the recently announced Veeco/FEI mergers. A further example of this dynamic is the lithography segment, the most expensive part of the semiconductor equipment market. The three leaders are ASML, Canon and Nikon. Even with over $100m in revenue in the last year, Ultratech Stepper is a niche player that recently has developed a strong penetration into the packaging market.

Yet there are still some segments of the industry that are fragmented and have yet to consolidate. These segments still exist for a variety of reasons such as low technology barriers to entry, the absence of an aggressive leader to affect a consolidation effort or the absence of other factors that allow suppliers to hold market share in the face of competition. Given the tendency of the industry towards consolidation, it is a question of when the necessary factors (including economic necessity) will coalesce to force consolidation. Applied Materials’ entry into the wet cleaning arena with their Oasis system has highlighted this segment. But even with Applied participating, it is still a clear example of a segment yet to consolidate. As we look at other sectors, we haven’t seen the kind of M&A activity one might expect when companies are financially challenged and the strong see more attractive valuations of potential targets. The exception has been in fab automation with the Brooks/PRIA merger at the high end/large size as well as smaller acquisitions by Brooks, Asyst and others in the space. Ironically, this may be because the strong are protecting their balance sheets and are rewarded more for buying when valuations are high than when they can get a better bargain but less immediate revenue and earnings additions. Or, it may be because the strong see no benefit in acquiring the products and revenue of smaller competitors that are losing money and suffering the pain and costs of integration. This almost forces the targets to hold back for the higher valuations that become available when the industry becomes frenzied.

There will be room for new technology innovation startups, but they will quickly be consolidated into existing leaders as the innovators struggle to garner funding, traction with customers and the burden of supporting a worldwide customer base that expects instantaneous field service and continuous process development. And in an industry that despite its high tech nature loves evolution rather than revolution, new technology is slow in adoption, giving the industry leaders an opportunity to respond with internal resources (and existing customer relationships) rather than forced acquisitions. X-ray is the ultimate example of a solution that has been pushed aside by evolution of the prevailing technology. This means that many innovators will be valued (if they are valued at all) on the basis of the cost of development and time to market opportunity (which in the current market is a relatively low valuation when compared to the recent past) and their current market traction rather than the size of the potential opportunity and the “technology”.

The industry is better poised for the recovery. Companies have taken the hard actions that were necessary to resize themselves. Almost every supplier says that the products for the next generation production lines are available. The stock market has gotten over the false start that coming off bottom elicits and is poised with “value” investors for the real improvements that will come.

Where is the “cycle”?

Finally, as the industry analysts are now professing, the valuations are in line with visibility, the economic data is improving and the cycle should improve after the current pause. So while SEMICON 2002 may have been muted, it simply means we haven’t gotten to the excesses that reflect reaching the top of a cycle. And that is a good thing.

As history tells us, there will be a next up cycle. But with no killer application identified on the horizon, there is uncertainty in both industry management and investors as no one has visibility of when we will see a sustained, strong up-cycle. We do believe the industry is stabilizing, although there may be further false starts before we see a strong upturn.