Once upon a time, there were two people. Let's call them "you" and "me". Each of us had a penny. One day, a stranger came along and sold me a lump of clay. It was fun to play with, and I thought I might be able to make a clay pot out of it eventually. So I bought it with my penny. The next day, you saw I was having fun playing with my clay, so you offered to buy half of it from me, with your penny. The day after that, I realized that I had made a whole penny of profit off of buying a penny's worth of clay, so I thought buying back some of the clay would be worth a try. You agreed to sell me half of what you had. On day four, you notice that there's a hitorical trend, where the clay market pays a good return. You can see that this justifies reinvesting at a fair price, eight cents per lump. Each day, one of us made a profit, while the other prudently invested the previous day's profit.
In a few weeks, we were both clay-market millionaires -- even though all we had was still that same penny and that same lump of clay. It didn't require any artificially low interest rates, any increase in the money supply, any repackaging of clay-backed securities, any buying on margin, any phony bond ratings. It didn't take any particularly remarkable characteristics of clay. All it took was a market with no reality check.
Of course, it did require that disconnect from reality: it's completely unreasonable to think that a lump of clay is worth a million dollars, particularly when you bought it for a penny the month before. But the disconnect isn't normally so obvious. Lots of things really do grow exponentially, at least for a while, so it's not at all unreasonable to think that the price of houses or gold might do so. And a small difference in exponential growth rates will soon lead to a big difference in the actual numbers. All it takes is a market with no reality check.
Markets are magic. That's not just a quasi-religious belief of the financial elite in this country. It's a theorem. Not a theory, but a theorem: a precise mathematical statement with a proof. The problem is, the theorem doesn't say what the ideological doctrine says it does. With a theorem, you don't get to conclude the conclusion unless the hypotheses are true. Markets are very good at aggregating certain kinds of information. But the information has to come from outside the market: in effect, from reality. Producers produce, consumers consume, and the market connects the two so that the stuff is neither accumulating in inventory nor getting depleted from inventory. It can even still work if middlemen get paid to store inventory at harvest time and release it gradually over the course of the year, or if speculators get paid to bear more than their share of financial risk -- as long as it comes out even over the whole cycle, with demand ultimately from consumers and supply ultimately from producers. But if the demand for something comes from investors buying it to sell later, while the supply comes from previous investors disinvesting, to the point where any actual producers and consumers are irrelevant, then the whole thing has come untethered from reality. And it's free to go floating skyward, expanding until reality finally does intrude and it bursts.