The Internet may prove to be the greatest human invention of all time. Investing in Internet companies in 2000, however, may prove to have been one of history's greatest follies.

Yet 2000 was a heady year for Internet investment. Guides such as Greg Kyle's 100 Best Internet Stocks to Own showed you "how to get in on this once-in-a-lifetime opportunity." Kyle predicted that there would be 430 million Internet users by 2003, and that by 2005, "consumers will spend $150 billion shopping online."

In fact, those estimates proved conservative. By 2003, nearly 600 million people were online. In 2005, shoppers spent more than $175 billion on the World Wide Web.

Time to cash in
But even though Internet usage blew away expectations, you would have been a big loser if you'd invested in Kyle's 100 best Internet stocks. How much of a loser?

In the fall of 2007, we went spent hours computing the returns figures. Spoiler alert: The results are painful.

Had you invested $1,000 in each of Kyle's 100 Internet names back on April 20, 2000, and held them through September 2007, your $100,000 investment would have turned into -- drumroll, please -- $37,814. That's a total return of negative 62%, and again, that return is through the fall of '07 -- before the current bear market.

You were more likely to pick a company that would go bankrupt (18) than you were to pick a company that simply increased in price (13)!

To the moon!
Even the success stories struggled with their valuations. Shareholders who bought in April 2000 made a fair profit (52%) on F5 Networks. Shareholders of Wit Capital Group, which became Soundview Technology, found 42% gains when Schwab acquired their company.

There were, of course, some amazing returns. You would have done quite well buying Verio, which was acquired in 2000. Verio shareholders scored a cool 117% on that deal in about a month.

But even the big winners can't change the fact that 18% of Kyle's companies went bankrupt. And many of the companies that survived, including Ariba (Nasdaq: ARBA), Check Point Software (Nasdaq: CHKP), and S1 (Nasdaq: SONE), had their share prices cut in half -- or worse.

What went wrong, and why
Most of the companies profiled in the book were profitless -- and burning through capital at a rapid rate. Indeed, many of the companies shouldn't have been worth a dime ... let alone billions of dollars.

See, Internet companies at the turn of the century were expected to generate massive cash profits. They didn't. A stock's value is nothing more than an estimate of its ability to generate cash profits over time. Before long, "market share," "network effects," "eyeballs," and "B2B business models" were exposed as Northern California euphemisms for "no cash."

The value of valuation
That's why valuation is such a critical component of investing. As the Internet mess illustrates, taking a top-down investing approach -- starting with the best, fastest-growing industry -- will usually lead to failure. Show us that industry and we'll find you a stock operating therein that's going down in flames.

That's why we advocate a bottom-up investing approach. Start at the company level and work up from there.

It's also why there are no no-brainers in investing. Just to repeat: Although the Internet has been even more successful than Kyle imagined, the stocks he profiled were mostly disasters.

China = the new Internet
When an earlier version of this article was published, we made the case that the lesson of the Internet was as timely as ever -- and not because of the burst housing bubble. Why was it timely? China.

After all, the Chinese government was concerned enough about a bubble to triple the tax on stock trades last summer. According to The New York Times, that move was "aimed at braking what many business executives and economists inside and outside China now see as a stock market bubble."

The Chinese stock index was up 130% in 2006, and another 97% in 2007. According to data from Forbes, Chinese stocks, as measured by the Shanghai and Shenzhen 300 Index, were trading for 52 times earnings that fall -- at a time when the S&P 500 was going for a P/E of 17. And according to Bloomberg, "Domestic [Chinese] investors opened about 49 million trading accounts [in 2007], nine times the total for 2006."

The more things change ...
Not even a decade after the Internet bubble, investors assumed that picking the right place to invest trumped picking the right companies to invest in. The lesson was just as painful this time around -- Chinese stocks fell more than 60% in 2008.

While "buying China" was a sucker's bet back then, things are a bit different now -- even after a recent rally in the Chinese market. For starters, multiples in China have come way down since 2007 -- as evidenced by the likes of Yanzhou Coal (NYSE: YZC), Yingli Green Energy (NYSE: YGE), and Cogo Group (Nasdaq: COGO).

Although these stocks may still look expensive compared with their American peers, the recent choppiness in the Chinese market means that you can buy into the country's great growth story near the best prices in recent memory -- which is a major reason why our team at Motley Fool Global Gains recently completed a research trip to Asia.

If you'd like to see our team's top ideas from that trip, we offer a 30-day free trial without obligation to subscribe. Click here for details on this special offer.

This article was first published Sept. 28, 2007. It has been updated.

Neither Brian Richards nor Tim Hanson owns shares of any company mentioned. Schwab is a Motley Fool Stock Advisor selection. Check Point Software is a Rule Breakers pick. The Fool's disclosure policy does a lot of spelunking.