excellent
Unless syllabus terminology has changed, the formula in your post is wrong.
ie the first formula, to evaluate allocation effect of your ability to allocate you fix benchmark return to isolate the weighting impact only
= (portfolio sector weight - benchmark sector weight) * (benchmark return)
OK, firstly I’ve no idea what the curriculum/syllabus/etc. looks like now, and secondly apologies for the year it’s taken me to respond. This is what happens when I only log in once every few years!
Attrribution - how you explain where excess return came from
Two main attribution models are used → Brinson-Hood-Beebower and Brinson-Fachler. The difference is how the allocation term is handled (*see side-note below). BHB’s allocation term is (Wp - Wb) * (Rb); whereas BF’s allocation term is (Wp - Wb) * (Rb - RB)
Rb = Sector return of benchmark
RB = Overall return of benchmark
To me BF makes more sense; BHB penalizes you for overweighting a sector with a negative return - but BF only penalizes you if you overweighted a sector that underperformed the overall benchmark return.
*Side-note - back in the olden days when I took L3 they referenced both attribution models but didn’t really explain them; they were just presented as two different sets of formulas. In one of them, they’d collapsed the interaction term and selection term together, but in the other formula they didn’t, which caused a ton of confusion. In either attribution model (BHB or BF) you can collapse interaction/selection together, or display them as separate terms. From my perspective it’s better to collapse interaction into selection because it’s pretty hard to explain what interaction means to a client in simple terms.