Surowiecki has an interesting take on the obvious up on his New Yorker blog. And I don't mean that in a bad way: it seems economists could usefully spend much of their time reviewing the obvious and why it doesn't fit our overly simplistic economic models.
He notes the fact that when stock prices go down, demand doesn't necessarily increase, it often decreases. The supply and demand curves change based on price changes. He cites Warren Buffet as one person who is mystified by this.
I took a bankruptcy class with Altman (of Altman's-Z fame) once a long time ago. I was surprised to see, when he presented the non-proprietary version of his bankruptcy prediction formula, that one of the terms was the stock price. This felt like cheating to me. But noone can argue against the fact that the price contains information and is predictive.
So it's not surprising that the value of a financial instrument is dependent not only on it's price but on it's change in price. Supply and demand curves should change as the product changes, and what you buy when you buy a share of stock is continuously changing.
Surowiecki concludes by saying
...Investors should be much happier buying when stocks are down... but it seems clear that this is not how most investors are psychologically built. Instead, we like to buy when stocks are rising, and we feel the need to sell when they’re falling. The impulse to do this is very hard to resist, and it is one of the biggest reasons why people, whether they’re investing in individual stocks or mutual funds, find it so hard to make money in the market.Ah, yes, everybody in the market is irrational. That's a useful insight. Why not, instead, when the data does not agree with your theory, accept that it's the theory that's wrong, not the data.