(FORTUNE Magazine) – There was a time when nothing on Wall Street was hotter than B2B. Investors were falling all over each other in a mad rush to buy shares in everything from obscure software vendors to online cattle bazaars. If you could brand the words BUSINESS TO BUSINESS on the back of an armadillo, there would be someone on Wall Street to sell it--and someone on Main Street to buy it. That was two months ago.

These days, ever so suddenly, B2B stocks are beginning to resemble their shell-shocked B2C cousins, with share prices down 70, 80, even 90%. Take FreeMarkets, which sets up online auctions for old-economy products like ball bearings, coal, and corrugated boxes. After going public last fall at $48, it rocketed all the way to $370 early this year. Today, FreeMarkets trades at $64. Ventro, another online market-maker, has been hit even harder, falling from a high of $244 to below $30 by late April.

Citing such high-profile shipwrecks might seem like a strange way to start a story about investing in B2B. But then, that's the point. These dramatic swings highlight just how risky picking stocks in the fastest-growing, quickest-changing segment of the market can be. Make no mistake--buying B2B right now is akin to setting out to sea after a big hurricane has passed through. There's wreckage aplenty to avoid in the shoals. Still, investors have what seems to be a golden chance now to bet on a business trend that is likely to be as far-reaching as it is capricious. The recent fire sale on tech stocks has slashed valuations on even the most promising companies, creating B2B bargains that are too tempting to ignore.

Even after factoring out the hype, it's clear that e-business offers huge opportunities for the select group of companies that do manage to thrive over the long term. Banc of America Securities analyst Bob Austrian estimates that worldwide B2B electronic commerce will grow from less than $20 billion in 2000 to roughly $13 trillion in 2004--a 650-fold increase. That's an incredible growth rate, but the cost savings inherent in B2B transactions explains why it may not be far off. On average, corporations spend $100 on paperwork alone each time they make a purchase, Austrian says. Moving those transactions to the Web could slash costs by 90%, saving billions in overhead each year. Not surprisingly, traditional firms are investing heavily in hardware and software so they can do more buying and selling via the Net. A recent Goldman Sachs survey of 42 FORTUNE 1,000 companies reveals that more than 40% plan to spend at least $500,000 each on new B2B software this year.

Problem is, there are now hundreds of companies out there, all competing for these seemingly abundant B2B dollars and all presenting themselves to investors as B2B plays. It has been called a land-grab, even a gold rush, and like the California gold rush of 1849, it's likely to create a few winners that become household names, while consigning scores of lesser companies to the Wall Street equivalent of Boot Hill. Even self-described B2B bulls like Merrill Lynch's Henry Blodget are quick to point this out, noting that investors could see close to 100% losses on some of the losers.

With this sobering caveat in mind, we went back to the gold rush metaphor in order to find the smartest ways to play the B2B wave. As was the case in 1849, we think, the real winners in this boomtown will be the companies that supply B2B prospectors with the Internet equivalent of picks, shovels, dungarees, and other essential supplies. This means software, which explains why our list of eight picks includes names like Oracle, Siebel, and Kana. What's more, handling those millions of transactions and ensuring 24/7 reliability throughout will require both powerful hardware and custom-tailored services. So we found the companies--IBM, Sun, PMC-Sierra, and Exodus--that do that the best.

With one notable exception, Internet Capital Group, we've stayed away from recommending the companies that directly operate the online bazaars. Although Ventro is likely to change how lab supplies are bought and sold, and eMerge Interactive could revolutionize the cattle market, it's still too early to know if these companies will ever rake in significant profits. As Morgan Stanley recently noted, the mother of all marketplaces is the New York Stock Exchange, with trading volume totaling $8.9 trillion annually. But in 1999, the NYSE took in just $75 million in profits--about half of what IBM makes in a week. There's another, and perhaps more compelling, reason to favor the B2B infrastructure stocks over the online exchanges--downside. Witness this spring's correction: While B2B infrastructure companies are off a modest 14% since February, the exchange stocks have plummeted 76%.

As with last year's successful e-portfolio ("Web Bets: 9 Ways to Ride the Net"; see box), we've divided our picks into three categories: the "bellwethers" (those infrastructure leaders that should benefit in a broad sense from the B2B boom); the "big swingers" (somewhat pricier players that may well come through with the big hits); and, finally, the "bull riders" (companies that should lead the charge if they can stay in the saddle long enough). For simplicity's (and honesty's) sake, call them risky, riskier, riskiest. And while many of them are expensive by conventional valuation standards--even after the latest bloodletting--we think the high prices are worth it. "It is hard to generate first-rate returns by owning second-rate companies," says Blodget. "And in this industry, you usually get what you pay for." Just ask the B2C bunch.

THE BELLWETHERS

For this category, we sought companies that had real profits--not just revenues--as well as the size, stability, and track record to survive the inevitable tech-stock downturn. And the three stocks in this category--Oracle, IBM, and Sun Microsystems--are all familiar (and much beloved) names, having generated average annual returns of 77% over the past five years. While this trio may not include any pure B2B plays, the e-commerce wave is likely to give each company another leg of profit and revenue growth, especially as bigger firms get into the act.

Our first pick, Oracle, is known both for its utter dominance of the database-software market and for the exploits of its yacht-racing, samurai-emulating, much-married CEO, Larry Ellison. But while the latter's flamboyance may generate headlines and megabytes of Valley cocktail-party chatter, the heart of this firm isn't Ellison. It's the products that regulate the lifeblood of e-business: data, data, and more data. Oracle's software lets buyers and sellers record the details of every transaction, while also allowing users to shape and extract specific information within these sprawling databases. Each B2B deal, for example, leaves a long trail of digital fingerprints, and everyone, from outside accountants to in-house inventory experts, needs access to some of that data. It's a critical niche in the Internet Age--Oracle's technology, for example, is used in all of the top ten B2C sites, and in nine of the top ten B2B commerce sites. And now, says analyst Melissa Eisenstat of CIBC World Markets, the company is using its strength in databases and its relationships with FORTUNE 500 giants to gain access to the broader e-biz market. Oracle, which already offers a broad suite of e-commerce software, is quickly becoming a one-stop shop.

These trends have set the stage for its explosive growth. Profits in 2000 are expected to jump more than 50% from 1999 levels. What's more, the company is in the midst of an aggressive effort to boost its own profit margins, from an already fat 30% to 40% over the next few years, by drinking the same B2B elixir it's selling to customers. Strong sales growth and widening margins, naturally, spell out the potential for upside earnings surprises--a factor that should help investors see past Oracle's forbidding estimated 2001 P/E of 84. Eisenstat, who believes the software maker could hit $96 over the next year, sums up what seems to be an emerging consensus on the Street: "If you want to invest in B2B, you've got to have Oracle in your portfolio."

Our next pick, IBM, was already a powerhouse in the 1950s when Oracle's Ellison was growing up on Chicago's scrappy South Side. As a matter of fact, historians of geekdom know that when he was a junior programmer in the 1970s, Ellison worked on one of the first IBM-compatible mainframes. Both he and Big Blue have come a long way since then. Although IBM still has a huge mainframe business, services and software have replaced Big Iron as IBM's profit drivers since the company's apparent stroke in the early 1990s. And in recent years CEO Lou Gerstner and his team have aggressively oriented IBM's service offerings toward e-commerce, a prescient move that is now paying off big-time--though it has yet to be reflected in the share price (see chart on previous page). In 1999, revenues in IBM's eServices unit jumped 60%, to $3 billion, according to analyst John Jones of Salomon Smith Barney, with growth in its electronic business infrastructure and eProcurement units totaling 200% and 300%, respectively. Indeed, by 2001 the $43 billion services division should overtake hardware as IBM's leading source of revenues.

At the same time, as companies ramp up their systems to accommodate the huge flow of e-data, IBM is likely to see added demand for its industrial-strength hardware and software. Although those areas aren't growing as fast as services, software and hardware throw off billions in profits, money that IBM is using to invest in faster-growing businesses. Profits, meanwhile, should continue to rise by at least 15% annually during the next five years, analysts say.

While such growth may sound paltry compared with Oracle's forecast rate, we haven't yet mentioned what is probably IBM's most attractive quality: Its stock is incredibly cheap. At a current price of $113, IBM is trading at 27 times this year's earnings, and only 22 times next year's. Not only is that multiple a far cry from the valuations of many popular tech stocks, it's also 10% less than the P/E of the average S&P 500 company. Given IBM's across-the-board tech strength --plus its big backlog of orders and its reassuring earnings visibility--IBM should trade at a 15% to 20% premium to the overall market, Jones says, which translates to a 12-month target price of $150, some 33% higher than today's price level.

If Oracle's specialty is software and IBM is a tech melange, then Sun Microsystems is a good old-fashioned hardware play. Sun's server and storage-equipment units account for more than half of the company's revenue and are Sun's two fastest-growing lines. It isn't hard to figure out why--when you click on any Website, there's a strong chance your computer is communicating with a Sun server, since they power the networks of 10 of the top 12 Internet service providers.

What's more, the push toward B2B practices like online procurement is encouraging the world's corporate giants to buy more servers to accommodate their own internal computing needs. Take Enron, which just announced plans to buy 18,000 servers from Sun over the next five years, a deal worth roughly $350 million. Bonanzas like this explain why Sun's most recent results surpassed even the most optimistic Wall Street projections, with profits rising 44% from the same period a year ago. Revenues, too, were remarkably strong, with sales rising 36%, to $4 billion. In a world where many highflying dot-coms have annual revenues of less than $50 million, $4 billion in a single quarter is serious money. And that's no beginner's luck. Sun, after all, has had a winning Net strategy ever since it started spreading the basic language of the Internet, TCP/IP (Transmission Control Protocol/Internet Protocol) back in 1982.

Like Oracle, Sun doesn't come cheap--it trades at 76 times next year's earnings estimates--making it a bit more vulnerable than low-multiple stocks like IBM in the event of disappointing earnings. "But what's compelling about Sun is how well it has done under fire," says analyst Andy Neff of Bear Stearns. "You name it--during the Asian crisis, in the Y2K lockdown, Sun didn't miss a beat. And as the company has grown in size, profit growth has accelerated and margins have expanded." Neff, who gives the DiMaggio-like hardware giant his highest rating, says, "Sun develops the initiatives that define where the Internet is heading."

THE BIG SWINGERS

If you're willing to take a flier on something a little riskier than Sun or IBM but still want real earnings and skilled management, then Siebel Systems is a natural choice. Indeed, devoted FORTUNE readers will remember that Siebel took top honors last summer in our annual ranking of America's 100 Fastest-Growing Companies. Founded seven years ago by Oracle veterans Tom Siebel and Pat House, Siebel Systems creates software that helps big corporations such as Procter & Gamble and Ford coordinate their sales, marketing, and customer-service efforts. With Siebel's software, a salesperson in Dubuque can check and instantly see the status of all the different leads she's pursuing, while also calling up product specs and other data. Siebel's software is Web-enabled, and the company has moved aggressively in recent months to offer new features like real-time wireless access and customized marketing tools.

With its reputation for fanatically obsessive customer service--the company even names its conference rooms after corporate clients like Ford and Kellogg--Siebel has signed up a passel of new customers in recent quarters, including giants like Merck, Du Pont, and Reuters. Plus, Siebel has consistently managed to beat the Street's earnings forecasts. In the first quarter of 2000, its revenues hit $309 million, up from $141 million in 1999. Profit growth was just as hot, with income rising 113%, to $46 million.

So why isn't Siebel in our first, less risky group? For starters, it's a young company, and young tech companies can be volatile in this market. More important, Siebel stock is expensive: Even after Nasdaq's recent pullback, it's still trading at more than 100 times next year's earnings. With the company dominating its niche and focusing heavily on customer service, however, it's likely that earnings will grow by 50% annually over the next three years, says B of A's Austrian. He thinks Siebel could hit $175 by the middle of next year, a 50% gain from its current price of $117.

Our next pick, PMC-Sierra, isn't a conventional B2B play. Indeed, linking this maker of high-speed networking chips to the B2B phenomenon may seem like a stretch. It's not. PMC-Sierra's semiconductors will be critical if the Net is ever going to move data at the pace that corporate customers require.

PMC essentially creates the guts of the switches and other equipment that propel data around the Internet. Networking giants Cisco, Lucent, and Nortel account for roughly one-third of revenues, in fact. Stoked by demand from the red-hot networking industry, the company is on a tear: Revenues in the most recent quarter were double those of a year ago, while profits nearly tripled. CS First Boston analyst Charlie Glavin has been forced to raise his earnings estimates for PMC seven times in the past 18 months. "I use a pencil, not a pen, when I put together my models for PMC," says Glavin. "We don't expect growth to abate anytime soon."

As with Siebel Systems, PMC-Sierra's multiple is in the high-altitude range--now trading at 149 times 2001 earnings projections. Glavin, however, says that conservative EPS forecasts make the company seem a little more expensive than it really is. He has a 12-month price target of $250.

A month ago, we might have put Web hoster Exodus Communications in our high-wired group (see below). As it is, the company was one of our "riskiest" picks last year, though it did manage to eke out a 259% return since our recommendation in June. In recent weeks, though, Exodus has stumbled in the market, creating an opportunity for investors who may have balked at the company's recent three-digit price tag.

What makes Exodus worthy of a return to the winner's circle? As the country's leading Web hoster, Exodus operates sites for nearly 3,000 corporate clients, including Net giants like Yahoo and DoubleClick. Most companies know they're not experts at running Websites, so they bring in Exodus to do the virtual drudgework, making sure the site operates smoothly and reliably and is safe from hackers and other threats. Demand for such Internet outsourcing is exploding--Exodus signed up more than 500 new customers in the most recent quarter, and saw revenues jump a breathtaking 346% from the same quarter last year.

With the company's underlying business growing so fast, it should come as no surprise that competitors are hungrily eyeing this niche. Intel has plans to enter the Website-hosting market, and big telcos like MCI WorldCom and Qwest are already competing with Exodus. Fear of these potential rivals explains, in part, why Exodus now trades at $78, down from $180 in March. But despite the threat of new competition, says Lehman analyst Bill Garrahan, there's still plenty of room for Exodus to grow. "Even if Exodus' long-term growth rate slows, this company will still do very well," he says. "And with every passing day, it becomes harder for Exodus' new competitors to catch up."

THE BULL RIDERS

Veteran investors know that buying shares of smaller tech stocks requires a strong stomach. But if you want to bet on the two stocks in this section, you should have a strong wallet, too. These companies--Kana Communications and Internet Capital Group--are right on the hump of the B2B bronco. If they get thrown off, you could end up with little more than a capital loss for next year's taxes; if they hang tough, the returns could be huge.

It's hard to find a better example of this risk/reward profile than tiny Kana Communications. Kana's software enables companies to handle the thousands of e-mail inquiries that e-commerce inevitably generates. Got a question about that garbage can you ordered from Wal-Mart.com? Well, Kana's software makes sure the inquiry is routed to the right place and is answered promptly. And there certainly is a market for such programs; CSFB's Brent Thill notes that one-third of e-mails to corporations currently go unanswered. Another third of e-mailers have to wait at least three days before they get a reply. "It's an enormous market," says Thill. And Kana is the biggest player in the space. But, as we said, it's still small. Revenues in 1999 totaled $14.1 million, a significant increase over 1998's $2.3 million in sales but less than the revenues IBM takes in over the course of an hour. And the company isn't expected to turn a profit before mid-2001. Even so, with Kana stock trading at $39, down from $175 earlier this year, you've got a chance to buy a nifty tech play at a bargain price. Thill believes Kana could rise 75% over the next year as Wall Street realizes the company's potential.

Our last pick, Internet Capital Group, isn't so much a play on B2B as it is a play on the appetite for such stocks. Like CMGI in the B2C realm, Internet Capital Group is an incubator that invests in and develops individual B2B firms before eventually taking them public as IPOs. An early backer of B2B exchanges VerticalNet and Onvia.com, ICG scored when the two companies recently went public. And it now has investments in more than 40 private exchanges as well as a piece of some 20 B2B infrastructure companies. If these businesses prove as successful as the first six B2Bers it brought public, ICG investors will enjoy a nice long ride on the gravy train.

But we're getting ahead of ourselves. Because in the past month, demand for B2B IPOs has fallen even faster than the once soaring Nasdaq. ICG's stock is now at $41, down a stunning 61% from its high of $106 in early January. Analyst Tom Berquist of Goldman Sachs insists that with ICG at these levels, investors have already discounted for an extended correction in the segment. And if B2B shares turn around, ICG will be an immediate, not to mention huge, beneficiary. "The B2B economy could eventually be ten times as large as the B2C economy," says Berquist. "ICG gives investors a way to get on board now."

REPORTER ASSOCIATE Noshua Watson

FEEDBACK: nschwartz@fortunemail.com