Hi Craig,

Again, thank you for the links. Even if the efficient frontier were incontrovertibly proven empirically it still does not get rid of the 'ex ante' or 'a priori' problem. In fact, it actually demands another layer of estimation to the problem of anticipating the future correctly. Also, there is nothing about the efficient frontier that tells you the correct time frame for which you should model your portfolio. Some suggest using 'unbiased estimators,' meaning if you have a 37-month investment time horizon then you should use 37 month data to build your estimates. But the efficient frontier still requires choices to be made by people, and at least to me personally as a former investment manager, actually adds to my estimation problem, rather than subtracting from it.

With smiles,

Jason