Not quite. I’m saying that the space between efficiency and predictability is very small and that claims of deep insights that don’t do as well as essentially nothing are not demonstrations of positive forecasting skill. [I actually spell that out quite explicitly at the end of the article.]