Sharpe Ratios and sd’s to measure volatility etc. are not appropriate for most AI’s. Who knows which are OK? For example, commodities – nothing I could find in the text (pgs. 294-308) suggests that there’s a problem using SR’s and sd’s. (In contrast, CFAI is quite clear on pg 334-337 that these measures don’t work for hedge funds.) So I assumed that SR and sd’s were ok for commodities… until I got to the solution for Q. 5A. So I am supposed to assume that SR’s and sd’s don’t work for any AI’s?
The problem with most of alternative investments is that distribution of their returns are usually non-normal (they have skewness, mostly negative and high kurtosis). Normality is the main assumption for Sharpe’s Ratio and for all bell curve measures (i.e. standard deviation). When normality assumption is under question, these measure don’t work. Its not appropriate to generalize by saying that Sharpe Ratio and Std won’t work for any AI, but for most of them they usually don’t work.