Well, there are as many different explanations are there are schools of economics, so it really depends on your leanings. Personally, I think the Austrian School's business cycle model comes closest to describing the most recent cycle. Basically, the story goes like this.
Like all other economic goods, time has a price: the interest rate. On an unhampered market, the interest rate will bring the supply and demand for loanable funds into equilibrium. That is, people looking to borrow money (through mortgages, loans, bond issuances, stock issuances, etc.) will be able to get it from savers (households, banks, businesses, etc.) at a certain interest rate. This rate reflects the aggregate time preferences of the public: if people choose to save more, the interest rate will fall.
Problems develop when the central bank intervenes in the financial markets to conduct monetary policy. The Fed sets interest rate targets and manipulates the supply of money to achieve those goals. For example, if the "natural" rate of the investment markets is 5% and the Fed targets 3%, it will inject new money into the markets to bring the rate down to the target. Note that these additional funds were not supplied by savers; rather, they were created out of thin air by the central bank. They represent a claim to economic resources with no real backing, like printing extra tickets to a football game when the stadium isn't getting any bigger.
When the Fed pumps new money into the credit markets, businesses are typically the first recipients. They use this money to embark on new investment projects that were not feasible at the old, higher interest rate. Some money is also diverted into the financial markets and can produce booms in the stock, bond, commodity, and/or housing markets. Households will save less and consume more because the interest rate has dropped. We get increasing consumption and increasing business investment at the same time, with the difference between household saving and firm investment being covered by the central bank's monetary injections.
This process is unsustainable because, as I described above, the newly-created money is not backed by anything. As they receive new money from the central bank, businesses and consumers fight over scarce resources, which when combined with growth in the money supply, creates inflation. The central bank (supposedly) hates inflation, so eventually it will tighten its monetary policy and raise interest rates.
When this happens, the capital projects begun by firms are no longer profitable and have to be liquidated. Consumer spending also drops as the interest rate rises. Economic activity slows as firms write off their malinvestments, reduce capacity, and adjust to post-boom reality. This process happens fairly quickly, as the "house of cards" created by an easy-money environment collapses and people realize they've been had.
Certainly there are psychological considerations as well, but without growth in the money supply, funds spent on dot.coms and other buzzword investments have to be diverted from somewhere else. During the 1990s, virtually every sector of the economy expanded in concert - there were not enough shrinking industries to free up the resources being poured into the tech sector. They had to come from somewhere else. They did not come from household savings: the personal savings rate went from 7% in the early 1990s to about 2% in 2000.
Despite the fact that the US economy was at or below the natural rate of unemployment in the late 1990s, the Fed carried out a very loose monetary policy in those years. To quote from Dr. Roger Garrison:
After increasing at a rate of less than 2.5 percent during the first three years of the Clinton administration, MZM [a measure of the money supply] increased over the next three years (1996–1998) at an annualized rate of over 10 percent, rising during the last half of 1998 at a binge rate of almost 15 percent.
The party was in full swing. The Fed cut the fed funds rate 100 basis points between June 1998 and January 1999. The rate on 30-year Treasuries dropped from a high of over 7 percent to a low of 5 percent. Stock markets soared. The NASDAQ composite went from just over 1000 to over 5000, rising over 80 percent in 1999 alone.
The business plans for many of the start-ups involved negative cash flows for the first ten or 15 years while they “built market share.” To keep the atmosphere festive, they needed the host to keep filling the punch bowl. But fears of inflation led to Federal Reserve tightening in late 1999, which helped bring MZM growth back into the single digits (8.5 percent for the 1999–2000 period). As the punch bowl emptied, the hangover—and the dot-com bloodbath—began.
Garrison's article (http://www.fee.org/vnews.php?nid=235
) has additional information on the Austrian theory of the cycle and how it played out in the tech boom. If you'd like a more detailed (and somewhat more technical) analysis, I have a great journal article saved as a PDF file. E-mail me or post a reply if you'd like to see it.
[Edited 2004-05-26 20:29:13]
Keynes is dead and we are living in his long run.