Wall Street made online trading sound like an instant way to mint money: Discover Brokerage, the online arm of the venerable firm of Morgan Stanley, ran a TV commercial in which a scruffy tow-truck driver picks up a prosperous-looking executive. Spotting a photo of a tropical beachfront posted on the dashboard, the executive asks, “Vacation?” “Actually,” replies the driver, “that’s my home.” Taken aback, the suit says, “Looks like an island.” With quiet triumph, the driver answers, “Technically, it’s a country.”

The propaganda went further. Online trading would take no work and require no thought. A television ad from Ameritrade, the online broker, showed two housewives just back from jogging; one logs on to her computer, clicks the mouse a few times, and exults, “I think I just made about $1,700!”

In a TV commercial for the Waterhouse brokerage firm, someone asked basketball coach Phil Jackson, “You know anything about the trade?” His answer: “I’m going to make it right now.” (How many games would Jackson’s NBA teams have won if he had brought that philosophy to courtside? Somehow, knowing nothing about the other team, but saying, “I’m ready to play them right now,” doesn’t sound like a championship formula.)

By 1999 at least six million people were trading online—and roughly a tenth of them were “day trading,” using the Internet to buy and sell stocks at lightning speed. Everyone from showbiz diva Barbra Streisand to Nicholas Birbas, a 25-year-old former waiter in Queens, New York, was flinging stocks around like live coals. “Before,” scoffed Birbas, “I was investing for the long term and I found out that it was not smart.” Now, Birbas traded stocks up to 10 times a day and expected to earn $100,000 in a year. “I can’t stand to see red in my profit-or-loss column,” Streisand shuddered in an interview with Fortune. “I’m Taurus the bull, so I react to red. If I see red, I sell my stocks quickly.”

By pouring continuous data about stocks into bars and barbershops, kitchens and cafés, taxicabs and truck stops, financial websites and financial TV turned the stock market into a nonstop national video game.

On December 20, 1999, Juno Online Services unveiled a trailblazing business plan: to lose as much money as possible, on purpose. Juno announced that it would henceforth offer all its retail services for free—no charge for e-mail, no charge for Internet access—and that it would spend millions of dollars more on advertising over the next year.

On this declaration of corporate hara-kiri, Juno’s stock roared up from $16.375 to $66.75 in two days. Why bother learning whether a business was profitable, or what goods or services a company produced, or who its management was, or even what the company’s name was? All you needed to know about stocks was the catchy code of their ticker symbols: CBLT, INKT, PCLN, TGLO, VRSN, WBVN. That way you could buy them even faster, without the pesky two-second delay of looking them up on an Internet search engine. In late 1998, the stock of a tiny, rarely traded building-maintenance company, Temco Services, nearly tripled in a matter of minutes on record-high volume. Why? In a bizarre form of financial dyslexia, thousands of traders bought Temco after mistaking its ticker symbol, TMCO, for that of Ticketmaster Online (TMCS), an Internet darling whose stock began trading publicly for the first time that day.

Oscar Wilde joked that a cynic “knows the price of everything, and the value of nothing.” Under that definition, the stock market is always cynical, but by the late 1990s it would have shocked Oscar himself. A single half-baked opinion on price could double a company’s stock even as its value went entirely unexamined. In late 1998, Henry Blodget, an analyst at CIBC Oppenheimer, warned that “as with all Internet stocks, a valuation is clearly more art than science.” Then, citing only the possibility of future growth, he jacked up his “price target” on Amazon.com from $150 to $400 in one fell swoop. Amazon.com shot up 19% that day and—despite Blodget’s protest that his price target was a one-year forecast—soared past $400 in just three weeks. A year later, PaineWebber analyst Walter Piecyk predicted that Qualcomm stock would hit $1,000 a share over the next 12 months. The stock—already up 1,842% that year—soared another 31% that day, hitting $659 a share.