Here be dragons

Volatility is only a good measure of risk when shocks are distributed normally. This post argues that extreme value theory offers a better, if imperfect alternative. Even better would be to use more fundamental analysis.