Philip Elmer-Dewitt of Fortune (if you'd like some reading on the subject) does a rather amusing quarterly piece in which he compares analyst predictions of Apple's performance with observed reality. If it happened once or twice, that would be understandable; such predictions are complex and difficult to make. But the track record is so consistently and so universally bad that I find it difficult to believe there's a truly rational process behind it; sooner or later, folks should be noticing the disparity between prediction and reality and trying to fix their process.

Apple's stock price seems to behave similarly; if any company not named "Apple, Inc." announces a new product which receives good reviews, generates high pre-sale demand and looks to have a high profit margin, that company's stock price quite predictably rises. But if the company happens to be named "Apple, Inc." the stock price almost invariably drops (go look at charts for AAPL on and around the days of major product announcements and launches). Perhaps you're aware of some deep principle which explains why this is rational behavior on the part of investors and traders, but I certainly am not.