The stock market is certainly reacting faster than ever to the slightest hint of good or bad news. Whether this means the stock market is more efficient than ever, though, is another question entirely. The answer may depend on whether you think "efficient" is a synonym for "jumpy and hysterical."

As Exhibit A, we offer the recent trials and tribulations of the so-called Internet stocks, trials and tribulations that said less about the actual business of the Internet than about the manic behavior of Wall Street in a just-won't-quit bull market. This most recent hysteria saw investors pouring billions of dollars into Internet stocks before suddenly--and all at once--changing their collective mind and racing for the exit. In this they resembled nothing so much as the knights in Monty Python and the Holy Grail, who took comfort in screaming "Run away! Run away!" as they fled killer rabbits and the like.

The ride began rather oddly, when some of the most prominent makers of computer hardware--memory chips, microprocessors, and PCs--announced that their earnings for the first quarter would fall short of the all-knowing analysts' estimates. Intel, Compaq, and Motorola all said that slowing demand--both in Asia and at home--was hurting profits, and both Intel and Compaq emphasized the negative impact of sharp cuts in PC prices. The bad news crushed these stocks, but much of the money that left Intel and Compaq immediately flooded into companies such as Yahoo!, Amazon.com, and myriad other search engines, online retailers, and Internet access providers whose listings now dot NASDAQ.

In part, that move into Internet stocks is simply evidence of what happens when everyone believes so strongly in the stock market that keeping any money in cash is regarded as heresy. In other words, if you were a mutual fund manager, you couldn't very well sell Intel and then say, "Now I believe I will wait until I can find a company that seems fairly valued." If you did that, you might miss out on the latest rally. The money had to go somewhere, and when Net stocks became the Next Big Thing (for those few days), that's where the money went.

S till, there was a certain logic at work in making Net stocks the Next Big Thing. If PCs are getting cheaper and cheaper, then more people will buy them. If more people have PCs, then more people will eventually use the Internet (fewer than 20 percent of U.S. households currently have access). This in turn means that more people will search on Yahoo!, buy books from Amazon.com, and so on. But while the long-run implications of cheap PCs for the Internet are important, it's hard to see how Intel's profit concerns made Amazon.com $500 million more valuable last Tuesday morning than it was the Friday before. The same was true of all other Internet companies whose option-holding executives must have been thanking their personal deities for the Asian crisis. It was a classic case of the next five years of good news for a company being priced into its stock all at once. And then the worm turned.

Why it turned isn't really clear, but it definitely did. It was a very odd thing to watch, in fact. Sometime midmorning last Tuesday, the Net stocks peaked, then tumbled even more quickly than they had ascended. Amazon.com, Yahoo!, Excite, Lycos, EarthLink, MindSpring: All lost between 10 percent and 15 percent of their value in just a few hours. Apparently, the cheap-PC theory had a few holes in it.

Just as the shift of money into Internet stocks was partly the product of a virtuous circle created by institutions trying to keep their portfolios invested in stocks that had "momentum," so too was the shift out partly the product of a vicious circle created by institutions trying to dump stocks that suddenly didn't have momentum. The rout also had something to do with the fact that only a few Internet companies have ever reported profits and that even those that have profits trade at valuations to which, as they say, no traditional criteria can be applied. Everyone knows that you can't value these companies the way you'd value General Electric (perhaps because if you did these companies would all have stock prices that are one-eightieth of what they are now). But every so often the market does look at their bottom lines, and it finds that discretion is sometimes the better part of valor.

Still, the most dubious part of this little minidrama is not really investors' herdlike advance and retreat as much as it is the governing principle behind that herdlike behavior. That governing principle is that all these companies belong to something called "the Internet sector" and that it's the sector, and not the companies, that matters. This is, of course, the principle governing most institutional investing, which is why pharmaceutical stocks and airline stocks and semiconductor stocks tend to move as groups. It's a rare day when American Airlines' stock rises while other airline stocks fall. There is the kernel of a good idea at work here, namely that macroeconomic or external events--like, say, a fall in the price of gasoline--that help one company in an industry will also help others. But sector investing also tends to reward a successful corporation's competitors for that corporation's success, and in that sense seems not merely counterintuitive but downright perverse.

More to the point, in the case of the Internet, sector investing rests upon a basic misconception, namely that there are things we can meaningfully call "Internet stocks." In fact, there are at least five different types of Internet companies, each of which has very different future prospects. Online retailers, such as Amazon.com and music seller CDnow, are still selling into a tiny market but one that's growing explosively. Search engines, such as Yahoo! and Lycos, are evolving into what are called "portal" sites, hoping to become more like mini-AOLs than just search engines. Infrastructure companies, such as VeriSign, Verity, and Netscape, are creating the tools that make the Web work, but their main source of income will come from corporations. A few companies, most notably CNET, are trying to make money packaging content. And, finally, there are the Internet access providers, which is to say, America Online and everybody else.

To be sure, all these companies will benefit from having more people use the Internet. But not all will benefit equally. And even within those five groups the differences among individual companies are more important than the similarities. To put it a different way, Amazon.com, Yahoo!, and AOL, which have established tremendous brand recognition and clear business models, have more in common with each other than Amazon.com does with other online retailers or AOL with other access providers. And yet the strength of these companies--each, by the way, now valued at stratospheric heights--is helping carry those around them. The Internet will undoubtedly be a crucial part of the economy of the next century. But Wall Street would be better off remembering that approximately 12,000 auto companies were around in the early part of this century. Instead of mindlessly tossing billions at or taking billions from the Net as such, investors should be spending their time making sure that it's the future Fords and General Motors of cyberspace that are getting the capital they need.