« What's so seductive about the efficient markets hypothesis is that it applies nine years out of ten. A lot of the time it works. But when it stops working, you blow up. »