Cyber-casino: Internet stocks. (initial public offerings by Internet companies)

WALL STREET regulars were already perplexed by the mania over Internet stocks: shares of firms with almost no track record, being bought by investors with no experience, at prices that appear to make no sense. But the announcement last week that Wired Ventures, which publishes Wired, a breathless but popular high-tech magazine, plans to go public at an Internet- style price was too much. The analysts may not know how to value mysterious things like digital cash; but they do know that no three-year-old monthly is worth $450m (Wired's prospectus estimate). Is this a sign that Net stocks have lost touch with reality?

Perhaps. But although the sceptics may be right about Wired's value, the reasons that they give are the wrong ones. In the real world, this is indeed just a magazine; but on the Internet it is potentially a great deal more. And by understanding why, investors can glimpse the factors that are driving the roller-coaster performance of Internet shares in general (see chart below).

[TABULAR DATA OMITTED]

Because much of Wired's value is tied to the success of Hot-Wired, its on- line subsidiary, it might be treated as a sort of on-line service, not unlike America Online, which trades at price-earnings multiples five times those of traditional publishing firms. On that basis, Wired's future revenues would depend largely on the size of the Internet itself, which has been doubling about once a year for as long as anybody can remember.

The potential for massive rewards, therefore, clearly exists. But to determine just how massive requires investors to solve a difficult problem. Even if they are willing to try their luck at guessing how fast the Internet will grow, they must also try to figure out how much of that increased traffic will translate into revenues. "Will people want to shop on-line?", for example, is a $100 billion unanswerable question. Even in biotechnology, where company valuations are notoriously unreliable, the potential demand for a given drug is reasonably easy to guess: just count up the sick people. So what should investors do? One approach is simply to give up trying to value Net stocks altogether. But this does not eliminate the need to decide: one can either buy the shares or not. Serious investors, therefore, must force themselves to make some assumptions, even if they turn out to be unrealistic.

One way to make those guesses more intelligently is to distinguish between different categories of Internet firms. They fall into three broad types: infrastructure, software and services, and media.

Internet infrastructure seems like a relatively sure bet. Whatever direction Internet commerce takes, the one thing that is certain is that it will all have to pass through the wires, computers and routing equipment of the network itself. The companies that make this hardware, led by firms such as Cisco Systems and Sun Microsystems, are therefore relatively sound investments. But they are also the least likely to stumble across an extraordinarily rich seam of hypergrowth: Morgan Stanley, an investment bank, reckons that revenues in the sector will rise by 30-40% a year.

Software and services are trickier. This sector includes Netscape, firms that provide financial services and so forth, and is bound to be a good market overall (50-60% annual growth, by Morgan Stanley's reckoning). But the usual difficulties in picking winning software technologies and companies are compounded by the speed at which the market is changing. A seemingly insurmountable technology lead can vanish overnight, bringing down an analyst's rosy valuation model with it.

Finally, there is media. The problem here lies in picking firms with sustainable advantages. Firms such as Wired that publish their own content risk losing the loyalty of the fickle Web surfer in a sea of competing sites. And those that earn money compiling other firms' information can be imitated easily (witness the profusion of search, indexing and news services).

Many Net-stock investors care little for these niceties. They simply see a huge new market opening and want a piece of it. Those who bought early and held on have made a bundle so far. Thirteen of the 15 firms in our chart are trading above the price at which their shares were initially sold to the public.

[TABULAR DATA OMITTED]

Perhaps this is a bubble, bound to burst when Internet stocks lose favour. But there is one reassuring sign: like normal stocks, they tend to rise on good news and fall on bad. Even if the price is wrong, therefore, at least it is always moving in the right direction. What more could investors ask?