|
In other words, demand for semiconductors is exploding. Part of that is due to a usual suspect: PC sales are on the upswing again. But much of the new demand is coming from a boom in some disparate areas, from wireless phones to new Internet access devices to high-end Web servers. Chips are hot, from flash memory to DRAM (dynamic random access memory) to traditional CPUs (central processing units). Investors trying to take advantage of this trend have a cornucopia of materials companies, parts suppliers, equipment makers, design houses, and plain old semiconductor manufacturers to choose from. But many of these companies are tied to one niche area of the market. Even the big daddy, Intel, is primarily a manufacturer of CPUs, with a few higher-growth but much smaller lines of other chips. BIG THREE. There's one exception, however: Varian Semiconductor (VSEA). It sells its product to the entire industry. Not only is this company a play on the exploding growth in all kinds of chips but its new product family is a technological breakthrough that might just knock its competitors down for the count. And the most amazing thing about the stock may be how cheap it is compared to other manufacturers of chipmaking gear. To understand Varian, first flip through this Cliff's Notes of semiconductor manufacturing. Making a chip involves as many as 250 different processes, starting with a bare wafer of silicon, which is then heated, cooled, pulled, bent, and burned in every way imaginable. One of the most fundamental steps in the manufacturing process is impregnating the flawless silicon with intentional abnormalities, around which the structure of the chip will eventually be built. This process is called ion implantation and is an incredibly complex mix of high-energy physics, molecular chemistry, and robotics. As the size of the channels on chips continues to get smaller (the industry is now on the verge of building chips that have 180-nanometer-wide channels, vs. the current norm of 250 nanometers), the art of implanting ions gets even more complicated. There are three significant ion-implant makers. Varian Semiconductor (which was split from its parent company, Varian Associates in April, 1999), Eaton (ETN), and Applied Materials (AMAT). Varian and Eaton each have about a 40% market share, and Applied Materials has about 18%. GOING SOLO. But analysts are expecting those numbers to change soon, thanks to Varian's new product line. Called VIISta, the new ion implanters work on one wafer at a time, unlike most other machines, which do batches of 18 to 20 wafers. Although this might seem like a productivity setback, it is in fact a leap forward. Doing each wafer individually guarantees uniformity and increases quality control. More importantly, the new ion implanter is able to tilt the wafer at a greater angle and with more control, which is a necessity for high-energy processes and for the new generation of 180-nanometer channel chips. "Historically, the chip industry made chips in the batch system, but as things have gotten smaller, the need to get precise uniformity has grown," says Manoj Nadkarni, editor of Web site chipinvestor.com. "Everything else in the process has switched to single wafer production, but ion implantation is the last step to still use the batch system. Varian has solved the technological hurdles with its new line. This is the way that everybody will go." In addition, the VIISta line of ion implanters is a "streamlined design," says Suresh Balaraman, an analyst with Adams, Harkness & Hill, who has a strong buy rating on Varian. "It's a machine that has a much simpler design, so it costs a lot less for them to make it. The company earned 35% to 40% gross margins with previous machines, but this one has the potential to see margins north of 50%," says Balaraman. GREEN LIGHTS. Varian Chief Financial Officer Ernie Godshalk agrees in general with Balaraman's estimate of future gross profit margins, but hedges slightly. "If everything goes perfectly, we can make 50% on these machines," he says. "But things rarely go perfectly, and labor costs aren't getting any cheaper, so a more reasonable margin to expect is in the mid-forties." That's still a dramatic improvement over the company's current gross margin of 38%, and would put Varian close to the elite semiconductor manufacturing equipment makers, such as Novellus (NVLS), which earns as much as 55% in gross margins on its products. The biggest obstacle to Varian's near-term growth now is its ability to manufacture enough of its new implanters. "Right now, all the lights are green, and we're doing our best to meet demand. We have a six-month backlog, and frankly that's too much," he says. He points out the company has increased production at its U.S. plant so that it's operating 24 hours a day, seven days a week. In addition, it has installed new equipment to speed production and reopened a previously shuttered plant in Korea. Meanwhile, Applied Materials is breathing down Varian's neck. It has a new single-wafer implanter that's still in the testing phase and might be able to compete with Varian's by yearend. But CFO Godshalk isn't sweating. "Applied Materials is one of the best companies out there, and they are a quality competitor. But this is their first attempt at such a machine, and I have my doubts that they'll get it right on the first try. I know how complicated these things are, and I know hard it is," he says. FINANCIAL BOOST. Nadkarni agrees. Until its new device comes to market, Applied Materials won't have any single-wafer machine to offer to customers to compete with Varian's. That's why Nadkarni rates Varian a strong buy and has even used his own personal funds to buy the stock. "This isn't rocket science -- it's far more complicated than rocket science," he says. "I can imagine few things with higher barriers to entry." He figures Varian will have several quarters to whittle away at the high current implantation segment, which is currently owned by Applied Materials. Varian's new product has already had an immediate impact on the company's financial statements. In its fiscal-year second-quarter earnings report, which closed on Mar. 31, the company announced that revenues had grown by 42%, to $141 million, over the first quarter of this year. (It's difficult to compare the company's results to the second quarter of 1999, since it was still a separate division of its parent company at that time.) In addition, the company nearly doubled its cash position, to $96 million. Analysts expect the company to make $1.91 per share this fiscal year and $3.08 in 2001. That gives the company a price-to-earnings ratio for 2000 of 22.7. A more important valuation tool to use for equipment companies, says Nadkarni, is the price-to-sales ratio. "If you look at price-to-sales and compare it to other successful peers, it just doesn't make sense," says Nadkarni. Novellus and KLA Tencor (KLAC), have price-to-sales ratios hovering around 10. Varian's is 4.9. "This company is drastically undervalued by the market right now." We'll see how long that lasts. Jaffe writes about the markets for Business Week Online What do you think about Varian's valuation? Let us know at our Ask Sam Jaffe Interactive Forum EDITED BY DOUGLAS HARBRECHT |
