R35 - Brinson Model vs BHB Model

Hi guys! Reviewing the Brinson Model (3.3) and the BHB Model (3.1) it is not clear to me when apply one or the other…they are very similar but the application should be differente…any help?



Yes I agree. Quite confusing and funny…

My view is avoid using (Wb - Wp) * Rb … this is not good at component level and will lead to wrong interpretation. it ll give possitive allocation effect even when the manager has performed lower than the benchmark once the Rb (benchmark component return) is possitive.

I believe the second formula Wb - Wp) * (Rb - Overall Rb) is an improvement on this because that shows possitive allocation effect only when the component return (Rb) that is over weighted has a return that is higher than the overall benchmark ave. return (Overall Rb)

you can put figures and test… cheers

Overall I agree in your reasoning…but I have two point:

  1. Following the (Wb - Wp) * (Rb - Overall Rb) approach also the selection effect changes because is considered together with the interaction effect, while in the other one there are 3 separate effects…then a new questiona arise: when they ask to calculate the selection effect is the selection only or the allocation too?

  2. I thing there is something missing in our reasoning…it’s so strange to have 2 chapter for the same thing: actually they give an explanation but it is not cleal the practical application

Thanks again!