(FORTUNE Magazine) – NOW THAT THE GREAT BUBBLE OF THE 1990s has receded into the past, we tend to dismiss it as a bizarre anomaly, a wacky blip in which everyone "went nuts" and fell victim to irrationality, stupidity, and greed. But that's history, the story goes. We've returned to our senses. We're smart and rational now. We'll never make those mistakes again. Alas, the truth is more complicated.

Consider the Google phenomenon. The Internet search giant went public in August 2004 at $85 a share, a level many thought ridiculous. Two months later, with investors having drastically underestimated the company's earning power, the stock had doubled from the offering price. By May of this year it had tripled. Now some analysts expect it to hit $350 or more. In Google-land it feels like the late '90s again.

This issue of FORTUNE devotes some 50 pages to the subject of decision-making. For investors, the fundamental question about Google--should I buy now?--involves many of the same considerations that market decision-makers faced last decade. The stock trades at a scary 100-plus times the previous 12 months' earnings, but the company's growth potential is so huge that it might end up being worth several multiples of that. What's more, what a company is worth is only tangentially related to what its stock is going to do over the next few quarters, and for many investors, the next few quarters are all that matter. So even if you're not wrestling with the to-buy-or-not-to-buy-Google question, you can bet your growth fund manager is. What's the answer?

Before proceeding, I must acknowledge the obvious. During the bubble era, I was a high-profile, top-ranked, Wall Street analyst. In 1998, I famously put a $400 price target on Amazon.com (which the stock obligingly zoomed past within weeks). In 2002, however, after I left Wall Street, I suffered the disgrace of being party to a regulatory conflict-of-interest complaint and getting charged with civil securities fraud. Without admitting or denying the allegations, I participated in the global settlement, paid a fine, and accepted a lifetime ban from the industry. Because I have been blamed for causing the bubble, anything I say on the topic runs the risk of sounding self-serving.

Having said this, I would respectfully suggest that we are sounder of mind in the midst of bubbles than we like to think--just as we are sane to consider buying Google at $290 or a house in today's frenzied real-estate market. In fact, when we examine, without benefit of hindsight, some specific decisions that money managers and individual investors made in the late '90s, many look downright rational.

In the middle of a bubble (which until it bursts, don't forget, is a boom), individuals, executives, politicians, central bankers, journalists, and others are confronted with hard choices. Sometimes their decisions are aggressive: I want to make money. Sometimes they are defensive: I want to keep my job. Sometimes they are emotional: I just don't want to be wrong anymore. What these decisions are not is "insane."

One of the more enduring myths about bubbles is that participants are so besotted that they never imagine they might be in one. To debunk this, we need only glance at today's real estate market. An expert a week opines that ballooning property prices are a disaster in the making (which they probably are). But for a variety of reasons--we need to live somewhere, the skeptics might be wrong, and the market keeps going up--we continue to participate. And so it was with the stock market. In April 1999, a full year before the crash, Barron's asked professional money managers whether they thought the market was experiencing a speculative bubble. A resounding 72% said yes. And in the following 12 months, as the Nasdaq doubled and the Dow rose nearly 20%, many of those managers kept right on buying.

Why did those who thought it was a bubble stay in the game? In part because most equity investors have to own something--they can't just sit on the sidelines. And in part because, for portfolio managers, life is more complicated than buying "undervalued" stocks and selling "overvalued" ones. Timing matters too. And relative performance--how your fund did last year or last quarter compared with its peers--is more important than making or losing money.

Moreover, bubbles by their very nature have a way of distorting perspective. During the late '90s, as in the 1920s, 1960s, and other such periods, persuasive theories seemed to explain the market's inexorable rise. Legions of respected economists, analysts, and observers spoke of tech- fueled miracles of the "new economy." Falling interest rates, productivity gains, low inflation, broadening stock ownership, and increasing returns on capital suggested that things might, in fact, be different this time. With the market rising year in, year out, even the most entrenched bears had to wonder whether they had missed a seismic shift in the economy.

Being human, we learn more from experience than we do from history. One of the investment mantras of the 1990s, for example, was "Dips are buying opportunities." From 1982 to 1999, this was a fact. Beginning with the crash of 1987, those who panicked and sold when the market stumbled were repeatedly burned when stocks bounced back and jetted off to new highs.

But by 1995, after a dozen years of market gains, the ten-year average P/E ratio of the S&P 500 had risen solidly above 20, a level it had achieved only three times in the 20th century, near the major market peaks of the early 1900s, the late 1920s, and the late 1960s. If undervaluation were the only rational reason to buy, every purchase thereafter was "irrational" (just as every purchase today is, by the way: Yale's Robert Shiller puts today's smoothed S&P 500 P/E at about 25, more than 50% above the past century's average of 16). Yet in the five years after 1995, the S&P 500 and Dow nearly tripled, and the Nasdaq quintupled.

Portfolio managers who lag behind peers for a quarter or two are viewed as doing poorly, and those who lag for a year or more often find themselves out of a job. Those who missed the post-1995 bonanza, therefore, got shelled. One of them was Jeff Vinik, who ran Fidelity's vast Magellan fund. At the end of 1995, Vinik tried to hedge an overvalued stock market by steering some of Magellan's billions into Treasury bonds. Magellan continued to make money, but with stocks rising faster than bonds, it also began to lag the market--a fact immediately noticed by financial commentators from coast to coast. Two quarters later, after Fidelity had been pummeled for Vinik's ostensibly cautious move, the star portfolio manager was gone.

Vinik's high-profile departure put the fund world on notice: The biggest career risk was not losing money, but missing gains. Over the next few years a parade of crippled legends hammered this point home, including Fidelity's George Vanderheiden, Oakmark's Robert Sanborn, and others. When Chuck Clough, Merrill Lynch's veteran strategist, quit in 1999 because he was sick of the grind, the Wall Street Journal implied that his departure was in fact due to his having committed Wall Street's "cardinal sin: being bearish, and wrong." (Actually, Clough had just been early--which, on Wall Street, might as well be the same thing.)

This created a perplexing environment for decision-makers. If the boom was a bubble, it could presumably burst--so it made sense not to bet the farm. On the other hand, it could last for years, and it wouldn't help to be "right" on the unemployment line. Eventually, this quandary--to play or not to play--touched even the most iconic investors in the world. WARREN BUFFETTED, one headline in early 2000 read. Why? Because in 1999, a year in which the Nasdaq nearly doubled, the Oracle of Omaha had posted the seemingly comical return of 0.5%.

As usual, Buffett would have the last laugh, but he was one of the few who did. (He was also one of the few who had the job security of running his own company.) Most others who refused to play simply lost. For example, Julian Robertson at Tiger Management had delivered legendary returns for more than a decade by following a value strategy. In 1998, however, with growth stocks blasting off, Tiger's performance wilted, and impatient investors yanked their money away from the firm and threw it at "better" funds. In early 2000, after two brutal years of losses and redemptions, Tiger's assets under management had shrunk from an estimated $20 billion to about $6 billion, and Robertson pulled the plug. Tired of beating his head against the wall but unwilling to modify his strategy, he retired.

Thousands of lower-profile professionals faced a similar dilemma. Consider the case of a star fund manager whom I'll call the Pragmatist. The Pragmatist made his name pursuing a strategy based on return on invested capital. In the early '90s, on the heels of the personal-computer bust, technology companies had been starved of capital and had grown extremely efficient. So the Pragmatist bought their stocks and rode them to riches. Several years later, however, with billions of dollars flowing into the tech sector, returns on capital inevitably declined, and the Pragmatist decided it was time to look elsewhere. By mid-1999 he had trimmed his technology exposure from 50% to 10%--which meant that, with tech stocks still skyrocketing, his fund started to lag. Irate investors bolted. After one bad quarter the Pragmatist's boss gave him a warning. After two, he got an ultimatum: Buy technology, or you're fired.

The Pragmatist evaluated his options: (1) Quit, (2) get fired, or (3) comply. Quitting wouldn't save his investors: The firm would simply hire a gunslinger who would go to a 50% tech weighting and tee them up to get killed. Getting fired, meanwhile, would lead to the same result--and cut the Pragmatist out of his bonus. Option 3, buying technology stocks in the face of an expected crash, seemed crazy on its face, but it was the sanest choice. So the Pragmatist took his fund back to 25% technology--a conservative weighting in those days--and watched as the market tanked and his portfolio got creamed. Then, understandably, he left the business.

For individual investors the incentives to play were nearly as powerful. By the end of the decade those who had moved to cash or clung to "safe" stocks had watched as neighbors got rich. The market had become a fantastic wealth-creation machine. And in the middle of everything was an exciting new technology, the Internet, with the new jobs, fortunes, and success stories that have forever made up the American dream.

In addition to being an analyst, I was one of those investors. For years I had watched small stakes in AOL and Yahoo burgeon to more than ten times my original investment. I was happy about that and happy to be taking part in all the miraculous events that were going on.

Still, I hadn't failed to notice that others were "taking part" more than I was. Everyone I knew seemed to be a dynastically wealthy technology investor or entrepreneur--or about to become one. These folks had created companies, products, fortunes, and jobs; I had written some research reports. I never lost sight of how lucky I was. But in an era of unprecedented progress and opportunity, my success seemed sort of, well, pedestrian.

Such was my state of mind in early 2000, when some Internet venture capital and private-equity opportunities came along. If I wasn't going to join the Internet economy, I thought, I could at least invest in it. I had no illusions about the downside (total), and I didn't exaggerate the upside (possibly enormous, probably modest, nothing guaranteed). I knew that valuations were extreme and that the whole phenomenon might be a bubble, but I also knew that the same could have been said anytime in the previous five years. In the end, recognizing that the pain of loss would be only slightly greater than the pain of missing further gains, I decided I would rather play and lose than not play at all.

Do I regret this decision, which incinerated about 90% of a painfully large investment? Yes. Was it greedy? I suppose. Was it stupid? In hindsight, yes, but no more so than many high-risk bets. Was it "insane"? I don't think so. Under similar circumstances, I might make the same decision again.

In today's market, a stock like Google triggers analogous decision-making considerations. But the most amazing aspect of the Google phenomenon is not the stock price, which is expensive but not outrageous (40 times this year's free cash flow, in my estimation, vs. about 30 for Yahoo and 20 for a mature media company like Viacom). The most amazing aspect is the company's performance. Only seven years old, Google is generating about a third as much cash as media behemoth Time Warner (FORTUNE's parent company), and its single online property is sucking advertising dollars away from almost every traditional media company. For most individuals, the risks of owning any high-octane stock like this one are too high to justify, but this is not the case for portfolio managers, especially those benchmarked against indices that include it. Which is why, until Google stops exceeding expectations, they'll keep on buying.

Feedback? investing@fortunemail.com