economists have a certain tendency to disregard tail events, which happen much more often than they think, as they don't follow the normal distribution. there's been a plethora of six sigma market movements in the past few years, and six sigma by definition should happen once every 3 million observations or so...