PREMIUM SEARCH Search by job title, geography and build a list of executive contacts
If they remade the 1960s movie The Graduate, the "one word" that Dustin Hoffman's self-appointed mentor would whisper into Dustin's perturbed young face wouldn't be "plastics." Nope, it would be: "infrastructure."
If that word leaves you as befuddled as Hoffman's character appeared to be over plastics, you aren't alone. Infrastructure has become one of those catch-all terms that's hard to define when it comes to the Internet. In its broadest sense -- which is pretty broad -- it means anything that helps two entities connect via the Web. That includes the software as well as the hardware -- processors, physical networks, telecommunications equipment -- that enable this interaction.
Infrastructure is the latest fad for Web investors, bigger even than "business-to-business" and light years away from the "business-to-consumer" Internet plays that lately have been going the way of the poodle skirt. Infrastructure is also the mantra of venture capitalists. Only a few weeks ago, venture firm Safeguard Scientifics announced that it was turning its back on its incubator strategy -- of investing in Internet startups -- in favor of building an infrastructure empire.
NAME GAME.
Trendiness notwithstanding, analysts and fund managers alike warn that investing in the Net's infrastructure is no defense against corrections in the Nasdaq. In fact, far from being a hip concept, Internet infrastructure is in reality pretty dull. When investing in it, the experts argue, you should stick with known names that have proven earnings and cash flow, and -- oh yeah -- plenty of revenue growth potential. These are the companies that have the wherewithal to adapt to future unforeseen developments in infrastructure technology. And by the way, an infrastructure portfolio won't come cheap: These stocks are richly valued these days, which makes it important to be selective about picking the ones with the least risk.
"Infrastructure stocks are slightly less risky than B2B," says Christopher Traulsen, a senior analyst at research firm Morningstar.com. "But they are not the safety net some people seem to think."
The safety-net argument goes like this: The companies that made the most money off the Gold Rush in 1849 weren't the prospectors, but the ones who sold them the picks, shovels, and Levi's jeans. "I retch when I hear that argument," says Pat Dorsey, the director of stock analysis at Morningstar Inc. "The '49ers were not paying $40,000 for a pick that was going to be obsolete in three months."
ONE-HIT WONDERS.
So how can an investor avoid the potholes? First off, by steering clear of startup networking companies that are trading at crazy multiples of earnings or sales, according to Dorsey. Another company to avoid is the one-product wonder: If the company you're looking at does only one or two things, analysts say, it's vulnerable to premature obsolescence.
One example Dorsey points to is Juniper Networks (JNPR), the first company to ever eat into Cisco's market share for so-called terra-bit routers, the chunks of hardware that manage the data flow between narrow and broadband cables in a network. Dorsey would bet on Cisco, not Juniper. "Yes, [Juniper was] the first company in years to take market share from Cisco, but they just do terra-bit routers," points out Dorsey. "All somebody has to do is come up with a better terra-bit router and they're toast."
That's a pretty strong statement to make about a company that took 20% of Cisco's market in just two years, when companies such as Lucent, Nortel, and Cabletron couldn't make a dent. But even fans of Juniper, such as Brian Miller, equity analyst at Invesco Funds Group's telecommunications fund, acknowledge that Juniper needs to make acquisitions to broaden its product line.
PACKET-BASED WORLD.
Still, Miller is wary of infrastructure companies that are having trouble making the transition from a circuit-based telecommunications world to a packet-based world that uses optical technology. In a circuit world, each connection has a dedicated line between caller A and caller B, which lies dormant when A and B aren't talking to each other. In a packet world, when a connection isn't in use between callers A and B, it can be used by callers C and D. While the circuit method is almost 100% reliable, the packet world, though less reliable, is much more efficient.
Miller and other telecom analysts believe the future lies with packets. That's why Tellabs, Lucent, and Alcatel are all companies that Miller is "keeping a close eye on," as they try to adapt to the latest technology. His favorite infrastructure bets are companies like JDS Uniphase, SDLI, E/Tek Dynamics, and Corning, all suppliers of parts of the network architecture. "No matter who wins among the network providers, [these companies] will be around to supply them with components," says Miller.
Their high stock prices don't put him off, either. "Expensive is an issue among these stocks, but with the volatility we've seen in the markets, you have to focus on companies that are earning profits with high cash flows and high returns on equity," says Miller. "These companies are self-funding."
AMD'S CASE.
It's hard to get away from expensive when buying a piece of the Internet infrastructure, but you can diversify. For instance, semiconductor manufacturers offer another way of betting on Net technology. "The only worry in these stocks is valuation," says Sudeep Balain, a senior technology analyst at Chase H&Q.
Balain sees plenty of upside in the old blue-chip chipmakers Intel (INTC) and Advanced Micro Devices (AMD). He has a 12-month target of 175 on Intel, which currently trades at 119 9/16. With chips that are now as fast as any that Intel has to offer, Balain believes, AMD will be trading at 150 within 12 months, up from its current price of 90 1/2.
AMD has momentum in its favor, having zoomed from 50 on Mar. 22 to its current price. But Balain is also predicting revenues of $5 billion in 2000, vs. $2.8 billion in 1999, and earnings per share of $5 this year, partly due to the use of tax credits, vs. an $89 million loss in 1999. In 2001, he expects revenues to reach $6.5 billion and EPS to be at $5.50 without using tax credits.
As for companies in telecom equipment manufacturing, Balain likes Xicor (XICO), which makes microchips for cellular phones. He expects the stock, now trading at 17, to hit 35 within a year. Given his 2001 earnings estimate of $1.15 for 2001, Balain figures the stock is pretty cheap. "It's trading at less than 15 times fiscal 2001 earnings," he points out. Competitors such as LSI Logic are trading at 25.
UNDERVALUED?
Of course, you can always delve into broadband networks as an infrastructure play. In a report called "Telecom - Internet Infrastructure Services," published on May 1 on its Web site, Morgan Stanley Dean Witter listed 12 stocks that its analysts, contrary to many of their brethren on the Street, believe aren't fully valued. From Covad (COVD) to SBC Communications (SBC) to Teligent (TGNT) to Broadcom (BRCM), all are positioned to "reap the rewards of broadband investment," the report declares. The stocks on the list have anywhere from 25% (Broadcom) to 182% (NorthPoint) of upside potential, according to Morgan Stanley.
If it's starting to sound like "infrastructure" is too broad a term to define an investment strategy, you may be right. "It's true that a lot of companies will participate in the secular growth of the Internet," says George Nichols, an analyst at Morningstar. "But how your portfolio does depends on which ones you pick." Whichever category you may choose to invest in, the consensus advice is, stick with the leaders in that category, and don't get fancy.
Popper covers the markets for Business Week Online
Get BusinessWeek directly on your desktop with our RSS feeds.
Add BusinessWeek news to your Web site with our headline feed.
Click to buy an e-print or reprint of a BusinessWeek or BusinessWeek Online story or video.
To subscribe online to BusinessWeek magazine, please click here.
Copyright 2000 , by The McGraw-Hill Companies Inc. All rights reserved.
Printer-Friendly Version
