Macro Again...the calcuation.

comp_sci_kid Wrote: ------------------------------------------------------- > Allocatio Actual Return Benchmark Return > > Domestic Equity 70% 6.50% 6.25% > Mgr A Large-Cap 60% 5.20% 4.90% > Mgr B Small-Cap 40% 7.50% 7.85% > > Domestic FI 30% 5.60% 5.25% > Mgr E Government 45% 4.50% 4.25% > Mgr F Corportate 55% 5.70% 5.80% > > 1) wa*wb(rb-ra) = 0.7*0.6*(4.90-6.50) > 2) wa*wb(rb-ra) = 0.7*0.4*(7.85-6.50) > 3) wa*wb(rb-ra) = 0.3*0.45*(4.25-5.60) > 4) wa*wb(rb-ra) = 0.3*0.55*(5.80-5.60) > sum them all up to get your attribution > > I dont know what the hell is Benchmark return for > domestic equity as you dont benchmark asset > classes, you benchmark particular managers CSk, thanks for that. see, looking at the formular from the book, it looks right. However, it just doesn’t make any Sh*t sense to me at all. BTW, I got lazy in my example, in the actual example (On one of the MOCK exam), each manager had a benchmark.

F ME! I’ll definitely have to review.

a bit of algebra as follow may help out the definitions of each pieces. assume there are “i” different asset categories, “j” different managers for each category, and “b” stands for banchmark. so, portfolio return is Rp = sum(i)sum(j){(wi)(wij)(Rij)} = sum(i)sum(j){(wi)(wij)[Rij - Rij(b)]} + sum(i)sum(j){(wi)(wij)[Rij(b)]} = sum(i)sum(j){(wi)(wij)[Rij - Rij(b)]} + sum(i)sum(j){(wi)(wij)[Rij(b) - Ri]} + sum(i)sum(j){(wi)(wij)(Ri)} = sum(i)sum(j){(wi)(wij)[Rij - Rij(b)]} + sum(i)sum(j){(wi)(wij)[Rij(b) - Ri]} + sum(i){(wi)(Ri)} = sum(i)sum(j){(wi)(wij)[Rij - Rij(b)]} + sum(i)sum(j){(wi)(wij)[Rij(b) - Ri]} + sum(i){(wi)(Ri - Rf)} + sum(i){(wi)(Rf)} = sum(i)sum(j){(wi)(wij)[Rij - Rij(b)]} + sum(i)sum(j){(wi)(wij)[Rij(b) - Ri]} + sum(i){(wi)(Ri - Rf)} + Rf there are four pieces in the last formula. the 1st piece is manager over his benchmark, the 2nd piece is benchmark over asset class, the 3rd is asset category over risk free rate, the last is the risk-free rate itself. clearly, ws was right in his computation.

rand0m Wrote: ------------------------------------------------------- > a bit of algebra as follow may help out the > definitions of each pieces. > > assume there are “i” different asset categories, > “j” different managers for each category, and “b” > stands for banchmark. > > so, portfolio return is > > Rp > = sum(i)sum(j){(wi)(wij)(Rij)} > = sum(i)sum(j){(wi)(wij)} + > sum(i)sum(j){(wi)(wij)} > = sum(i)sum(j){(wi)(wij)} + > sum(i)sum(j){(wi)(wij)} + > sum(i)sum(j){(wi)(wij)(Ri)} > = sum(i)sum(j){(wi)(wij)} + > sum(i)sum(j){(wi)(wij)} + sum(i){(wi)(Ri)} > = sum(i)sum(j){(wi)(wij)} + > sum(i)sum(j){(wi)(wij)} + sum(i){(wi)(Ri - Rf)} + > sum(i){(wi)(Rf)} > = sum(i)sum(j){(wi)(wij)} + > sum(i)sum(j){(wi)(wij)} + sum(i){(wi)(Ri - Rf)} + > Rf > > there are four pieces in the last formula. the > 1st piece is manager over his benchmark, the 2nd > piece is benchmark over asset class, the 3rd is > asset category over risk free rate, the last is > the risk-free rate itself. > > clearly, ws was right in his computation. almost right, he mixed up couple of numbers. :slight_smile:

Go WS! :slight_smile:

Allocatio Actual Return Benchmark Return Domestic Equity 70% 6.50% 6.25% Mgr A Large-Cap 60% 5.20% 4.90% Mgr B Small-Cap 40% 7.50% 7.85% Domestic FI 30% 5.60% 5.25% Mgr E Government 45% 4.50% 4.25% Mgr F Corportate 55% 5.70% 5.80% Risk-Free = Rf Asset Category = sum(i){(wi)(Ri - Rf)} Benchmark = sum(i)sum(j){(wi)(wij)[Rij(b) - Ri]} Manager = sum(i)sum(j){(wi)(wij)[Rij - Rij(b)]} Risk free = 3%. asset category = (0.7)(6.5-3)+(0.3)(5.6-3) = 3.23% Benchmark = (0.7*0.6)(4.9-6.5)+(07*0.4)(7.85-6.5)+(0.3*0.45)(4.25-5.6)+(0.3*0.55)(5.8-5.6) = -0.44% (why does this seem wrong) Investment manager = (0.7*0.6)(5.2-4.9)+(0.7*0.4)(7.5-7.85)+(0.3*0.45)(4.5-4.25)+(0.3*0.55)(5.7-5.8) = +0.045% Residual = +0.40% (why does this seem high)… Total = 6.23%

What are the chances that macro attribution will be on the exam in quantitative, rather than qualitative form? And I read that macro attribution was in the CFA mock exam (the full one). Is that right?

i would say unlikely…

how did you get risidual without knowing the actual return?

Willy, your Benchmark: you have 07 and it should be .7

bigwilly Wrote: ------------------------------------------------------- > Go WS! :slight_smile: Sh*t, I am with you, F*ck me, I will have to review this big time.

Sorry should be .7 doesnt change answer.

CSK I just took 70% of Equity returna nd 30% of FI return to get a wghtd average…

i think Willy’s explanation is spot on but i feel like the example itself isn’t bulletproof… For example, for the Benchmark why are we using 6.5 instead of 6.25? Maybe, I am not bulletproof…

bigwilly Wrote: ------------------------------------------------------- > CSK I just took 70% of Equity returna nd 30% of FI > return to get a wghtd average… that is not right, risidual is the difference between actual return and sum of all return attirbutions. it is return attributed to difference between policy weights and actual weights

^Not sh!t Sherlock :). I used the 70/30 wghtd average to get the Approximate return of the Fund since it wasnt provided. I need to fill in a blank :slight_smile:

The CFAI Reading #43 Problem #10 I thought was going to be a great problem to practice this on, but NO they dont really calculate anything!!!

SS#16 vs BigWilly tomorrow… I can’t wait! This might last into Friday but Friday is supposed to be SS#17 review…

I will be looking forward to your battle…

I won’t…It wants to fight me while I’m weak and beaten down…