Q4c, the answer says we should use sortino ratio to measure the performance, but i think it is not ok, because sortino using only downside deviation, but manager always outperform every year, thus we should use upward deviation.
any idea why we use sortino?
It says Manager is constantly outperforming - means most of time he generates return more than the benchmark if we use std deviation - it would be positively skewed??
Sortino measures return per unit downside risk . Manager has very little downside risk. So denominator of Sortino is lower than average so manager will look good in Sortino and not so good using sharpe
Jana:
The three performance appraisal measures used in the ques were Sortino, SR, RoMAD -----SR & RoMAD has the same ranking, whereas Sortino has other.
Ans chosen Sortino
What do you mean “SR and RoMAD have the same ranking?”
SR = RP/St Dev
RoMAD = return over Max Drawdown
They do not use the same inputs, are you thinking of SR and M^2? they both may show no skill where Sortino and Jensen’s may.
In answer to the original question , if the manager is almost always outperforming , then the way to distinguish him from another manager who is not performing as consistently is to use Sortino.
the manager who constantly outperforms will have a low downside deviation.
The Sortino ratio will be larger for him than the manager who has up and down years .
So it will show him in better light than the inconsistent performing manager
Fin it was in reference with the question. As per SR & RoMAD - Brooks outperformed Davis. Davis outperformed Brooks only on Sortino ratio
Ans: Most appropriate measure is Sortino since Davis has beaten his benchmark every year, his volatility is on the upside, std dev & max draw down overstates his true risk??
why is sortino an acceptable choice? roys safety first is not an acceptable choice because it is a measure of shortfall risk. not excess return (i emailed schweser about this and thats what they said). so how is sortino any different–the only difference in the formulas is the standard deviation used.