I remember reading an interview (probably in one of Jack Schwager's Market Wizards books) with a guy who had been a Wall Street analyst. He kept the news reports of the day sorted into "good news" and "bad news" piles on his desk. When a journalist called to ask why something had gone up that day, he'd cite whatever was on top of the good news pile. When asked why something had gone down, he'd cite whatever was on top of the bad news pile.