Maturing Industry? No mention of ACLS but the discussion was worthwhile.
Wall Street Transcript
Market Maturation a Key Topic of Wall Street Transcript Semiconductor Equipment Review
Wednesday April 13, 10:21 am ET
67 WALL STREET, New York--April 13, 2005--The Wall Street Transcript has just published its SEMICONDUCTOR EQUIPMENT REVIEW, a report offering a timely review of the sector to serious investors and industry executives. This 40-page feature contains commentary from 3 leading industry analysts, plus in-depth interviews with top management from 12 firms. The full issue is available by calling (212) 952-7433 or via The Wall Street Transcript Online.
There are some mega trends from the demand side that will persist for a long time, such as digital TV, digital content, audio and video. Technology inventories are being cleared out of the supply chain for chips and we are seeing more strength in equipment orders from major chip companies in the logic and memory market segments. Topics include: Trends driving the industry, new innovations and technology, the Taiwanese players, industry spending, process firms, and the ramifications of more devices containing more silicon content, Stock picks, Companies to Avoid.
Companies include: Intel (INTC). Taiwan Semiconductor Manufacturing (TSM); United Microelectronics (UMC); Samsung; Applied Materials (AMAT); Novellus Systems (NVLS); Lam Research (LRCX). KLA-Tencor (KLAC); Nanometrics (NANO); ADE (ADEX); Rudolph Technologies (RTEC); ICOS Vision Systems (IVIS); Teradyne (TER); Semitool (SMTL); Aetrium Incorporated (ATRM); Brooks Automation (BRKS); Cohu, Inc.(COHU); FEI Company (FEIC); inTEST Corporation (INTT); Kulicke & Soffa Industries (KLIC); MEMC Electronic Materials (WFR); QuickLogic Corporation (QUIK); Genesis Microchip (GNSS); Silicon Image (SIMG); Trident Microsystems (TRID); PortalPlayer (PLAY); SigmaTel (SGTL); Napster (NAPS); Diodes (DIOD); Apple Computer (AAPL). Analysts include: Gerald S. Fleming of WR Hambrecht + Co, Gavin X. Duffy of A.G. Edwards & Sons, Christopher A. Chaney, of Stanford Group Company
In the following brief excerpt, Gerald Fleming discusses the maturing nature of the semiconductor equipment sector, and his expectations for long-term demand growth.
TWST: When you talk to institutions, what's the prime question or concern you're getting? Is it this maturing theme?
Mr. Fleming: I think it is related to the maturing theme. In a maturing industry, the big question is, "What's the right level of spending on equipment for a semiconductor company?" Historically, the capital expenditure for semiconductor companies was about 22% of their revenues. When you grow less rapidly, that number comes down, and there's debate among people on the Street (my competitors and myself) as to what the right level is.
Our feeling is if you come down from 15% growth to 10% growth, the cap ex to sales number probably comes down to somewhere around 20%. If you spend at a higher rate than that, you're adding more capacity than is necessary to keep up with demand and then chip companies end up in a price war. If you spend less than that, you haven't got enough capacity and then firms rush to add capacity and that's what creates the cycles.
We think right now, with the industry spending at about 20% of revenues this year based on our forecasts, spending is at roughly the right level and chipmakers are likely to continue to invest in technology. But there likely won't be a huge surge in capacity spending.
TWST: Which is good news.
Mr. Fleming: It's good news and it's bad news. For a chip company it's good news. It basically puts them in a position to run a business where they can generate cash flow and hopefully make a profit. For the equipment companies, it can provide steady growth but not dramatic growth.
TWST: Are the managements adapting to a slower-growth way of thinking?
Mr. Fleming: I think they are. I think that's why we see the consolidation situation. You've had a company like Applied Materials, which is the 800-pound gorilla in the space, start to change its business model to where, rather than just selling equipment, they're focusing on building a much larger service and consumables type of a business. That's not only supporting the very large installed base of their tools that are out there in the field, but also supporting other tools. I think that's another sign that the maturing is happening.
But do you need 60 semiconductor companies? I don't think so. Customers want at least two choices for every product that they're buying. There are a few areas where there are three or four. I think the one area that's probably the best example is electronic test equipment, automated test equipment, where you have Teradyne, LTX (LTXX), Credence (CMOS) and a number of more recent startups. The big guy in the industry is Advantest (ATE) in Japan. People don't need a half-dozen suppliers. In our opinion, that's why those companies don't make much money.
TWST: Given this more modest outlook, what are you telling investors to do in the space?
Mr. Fleming: We're telling them that the best place to be is in the inspection and metrology area - process control - because it is getting tougher to make a chip. There are something like 500 process steps in a typical 90-nanometer chip. The number of transistors that are on each chip is increasing and the feature sizes are getting smaller. As the number of critical layers in the process is continuously increasing, so is the probability of having a failure increasing very rapidly. How do you fix the problem? You put in more process control equipment.
In the early 1990s, about 5% of a chip company's equipment purchases were in inspection and process control. By the end of the decade it was 10% and today it's up to about 14% or 15%. Projections are that it's probably going to increase to about 20% at the end of this decade, so you've got solid growth. This equipment gives people a terrific payback. If you're producing Pentium chips and you can increase your yield by 1% by buying a $10 million machine or buying several $10 million machines, you may get a payback in weeks or a few months. KLA-Tencor and these smaller companies are able to charge a pretty handsome price because of the value the machine adds.
TWST: What names do you favor in the space?
Mr. Fleming: Within the process sector, we think you have to look at companies that are reinventing themselves in a positive way. There are two extremes there. One is Lam Research, which for years was number two or number three in the oxide etch market. They had some of the lowest margins among the larger companies with good products. Basically, they stopped competing on a price basis, instituted a lot of changes in terms of going to more outsourcing and accelerating their build cycle. They've gone from having some of the lower margins in the industry to having some of the better margins in the process equipment sector.
The other extreme was Novellus, which lots of people like because they were the ones that pioneered equipment for producing copper wiring as opposed to aluminum wiring, which gave chipmakers better performance. Meanwhile, they were out acquiring companies that had lower returns and smaller market share positions. They've seen their margins going from being the best in the industry to where they're below Applied and Lam today. We don't think that's fully reflected in that stock, so we have a buy on Lam Research, and while we have a hold on Novellus Systems, it's certainly a name that we wouldn't be adding to.
TWST: It's not an enthusiastic hold.
Mr. Fleming: Right. There are two other buys on our list, including ICOS Vision Systems, which is a company based in Belgium that has a lot of vision and vision systems and software that can be used to control other tools. They've acquired several operations in the business of inspecting and assembling chips, and have put their vision technology into them and have created industry-leading companies. The company has a 61% gross margin in spite of the fact that the back end of the business is incredibly cyclical, much as the electronic test equipment business is. I think last quarter, their sales were down 25%-30% and yet they held gross margins at 61%. That's pretty darn impressive.
They're now moving into the back end of the wafer processing business. They bought a division from Siemens (SI) in Germany and put their inspection technology into it. They're actually shipping a new product in that area that we think has pretty good potential. That's one we have a buy on.
The other firm that we have a buy on is Semitool. It's in Kalispell, Montana, which is an odd place for a semiconductor equipment company. It makes cleaning and electroplating equipment. When people think of electroplating to put down the copper wires on a microprocessor, you immediately think of Novellus, which is the dominant company in that part of the market, but these guys do the electroplating to put bumps on chips.
One of the new trends in the semiconductor industry is that instead of packaging chips using what are called wire bonders that connect fine gold wires from the chip to the circuit board, you build the chips and then the last step is to plate on a bump, which could be some sort of an exotic material or it could be solder, and all of these bumps have to be identical. There might be hundreds of them on a chip. Then you take the chip and you flip it onto the circuit board, which has electrical contacts, and you solder it.
It enables you to package devices in an electronic product much more densely than you had historically. You get better electronic performance and you also get lower cost. That area has been around for quite a while, but it really seems to be gaining some traction right now. Semitool's product is an enabling technology for those people. They also make cleaning equipment and are doing quite well in that market.
The negative in it is that it's a small company that competes in that process equipment space, and the chipmakers like to deal with big guys. They could be a consolidation candidate. I think that about covers our current list of equipment producers.
. |
