# selection effect

i thought the formula is the sum of Wi ~ iRi-Rbj, i means security i, b means benchmark but notes volume2 #morning exam 3 #Q7 says : Wb ~ iRi-Rbj, why???

Its teh Weight of the Benchmark Sector times the Different between teh Portfolio Sector Return - Benchmark Sector Return.

the notes book use weight of portfolio !!??

CFAI uses the weight of the benchmark! and that is all you need to know.

is this a Global portfolio? If so I think you do use portfolio, but if not then use Bench.

I don’t have V2. Is this security selection effect from Micro Attribution or Global Attribution? The two use different formulas…

Always default to CFAI. They use the benchmark weights. I have double-checked the formulas in the notes with CFAI texts and have found several discrepancies.

Hmm… don’t know if I’ve seen a question on this relating to global portfolio. why would it be any different? Aren’t you just substituting geographic weights in place of the normally used sector weights? Time to review this area…

i know the question you’re talking about - in that exam schweser was not very clear on what they were looking for as there is pure sector allocation and thne there is allocation/selection effet. schweser just asked for allocation which wasn’t clear. Expect the actually exam to use the full wording so as to not have this confusion occur.

right in global portfolio it is Wj * (Pj - Ij), but for regular macro it is Wbj * (Rpj - Rbj) where Wj = weight of security inportfolio Pj = capital gain on security Ij = return on international index Wbj = weight of the benchmark Rpj = weight of security in the portfolio Rbj = weight of security in the benchmark

Need to go look at some of these questions again… This is one of the last areas I’m reviewing before finishing up with Ethics/GIPS and then going on to the practice exams.

yep, i mean global…

soexcel Wrote: ------------------------------------------------------- > yep, i mean global… My bad. For global security selection, use portfolio weight.

dont try to memorize it, make sure you can open all parenthesis and minimize it to this: rp*wp-rb*wb. I cant stress that enough On that note for Global you use wp(rp-rb), as you incoroporate asset allocation/security selection into pure security selection

csk, for macro when you open all parenthesis, for individual security/sector total you should get Rpj*Wpj - Rbj*Wbj - Rb*(Wpj-Wbj) or Wpj* (Rpj - Rb) - Wbj*(Rbj - Rb)

I cant wait to review this again! What is the SS16 or so… I’m only on SS9. What an easy SS, problem wise…

volkovv Wrote: ------------------------------------------------------- > csk, for macro when you open all parenthesis, for > individual security/sector total you should get > > Rpj*Wpj - Rbj*Wbj - Rb*(Wpj-Wbj) > > or > > Wpj* (Rpj - Rb) - Wbj*(Rbj - Rb) you can take Rb out and weights sum to 1 so this will cancell out

agreed, this will work for total portfolio return, but when you are calculating sector or security specific return you need to include Rb (Wpj-Wbj) for the total portfolio return is 0, so Rb*(Wpj-Wbj) = 0, and the equation I quoted transforms into Rp*Wp - Rb*Wb, same as you quoted

volkovv Wrote: ------------------------------------------------------- > agreed, this will work for total portfolio return, > but when you are calculating sector or security > specific return you need to include Rb > > (Wpj-Wbj) for the total portfolio return is 0, so > Rb*(Wpj-Wbj) = 0, and the equation I quoted > transforms into Rp*Wp - Rb*Wb, same as you quoted yep exactly. You always need to end up with Rp*Wp - Rb*Wb