Maybe this is more of a rant than a question. Eoc q. 10 in reading 26 gives two efficient frontiers, one above the other. It tells us that the lower one is the frontier with correctly-estimated parameters. It asks: Could the higher one be the “conventional” mean-variance efficient frontier? The answer says no. Because the lower one is “optimal”, how could a frontier with misestimated parameters be any better? I say bollucks. Is that how you spell bollucks? The “conventional” frontier will have all sorts of misestimates and will overweight asset classes that did well historically. The result will be drawing a frontier that is unrealistically high, just like theyve drawn in the question. Of course, these results wont be truly achievable, but the conventional-frontier-drawer wont know that. Thats the frontier hes going to estimate...the whole point of black litterman and other approaches is that the conventional frontier isnt necessarily accurate. So, the book is wrong…right? Who`s with me?
I did not understand the logic also. I moved on.
I’m confused by Eoc q. 10 in reading 26. Why can’t the conventional efficient frontier be above the efficient frontier without estimation error? What if overestimating the return for a given risk?
Up. I just hit this question and can’t figure out the answer either. Can anyone shed some light?
at any point on the optimal efficient frontier - you have the highest return for a given level of risk (and by that same token the lowest risk *std dev* for a given level of return).
Given that - if the both are efficient frontiers - the two cannot exist simultaneously. For a given level of risk (draw a vertical line) you cannot have two levels of return (one on the upper one - higher return and another on the lower one (with a lower level of return)). If the lower one had been achieved - the higher level CAN NEVER BE ACHIEVED - whatever be the estimation error etc.
does this make sense?