I’m bumping this. Yes still a bit confused. if local mkt Sharpe was given in addition to GIM Sharpe, would you use it?
No, because the CFAI textbook does not.
I had the exact same problem with the whole approach. Semented markets have rho = 1 (perfect correlation?) What?
I assume Singer and Tehraar are academics who don’t actually need to test their theories once the paper is published.
In short, logically yes you should use the local market’s Sharpe but in the exam, I will use the GIM Sharpe.
I have found from time to time that CFA material is confusing given my background isn’t finance or economics.
in the same topic in this years material, it says in a segmented market, the correlation between asset I and the market is 1. Why is this? It is the correlation between the market and an asset inside the market. It is not the correlation between the market and itself.
Bumping this… This assumption in S-T model does not make sense to me either.
It seems to say that we are looking from the inside the market perspective and in a segmented market GIM is the local market, whereas in fully integrated case it is the real global portfolio. But then it certainly does not make sense to assume that the global market risk premium and Sharpe are the same in both cases. They are two very different markets with the same name.
just take it as a simplifying assumption (as a footnote on the text says it is so) and move on…
This thread got confusing fast. I don’t see them asking anything beyond handing you the input values and asking for the risk premium, so I wouldn’t worry about this as long as you know the formula.
Here in 2021 and had the same question about global and segmented Sharpe ratios confusingly being the same. For anyone looking to understand the logic Singer-Terhaar presented, the book says (Vol. 3, pg 236), “pragmatic approach to specifying Sharpe ratios…is to assume that compensation for non-diversifiable risk is the same in every market. That is, assume all Sharpe ratios equal to the global Sharpe ratio”
Thanks Berry and every one who’s chimed in.
Interestingly, I’ve done a Boston mock (not sure it was 2022 or 2023) and the question provided the local Sharpe Ratio. I used the GIM Sharpe Ratio for the Full Segmentation and got the answer wrong (correct answer used local SR). Urgh.