So for the micro-attribution portion of “pure sector allocation” the formula is (Wjp-Wjb)(Rjb-RB) and the “security selection” component is (Rjp-Rjb)(Wjb) Now in the international performance decomposition when comparing to an international benchmark a few chapters later, the formula for the “Market Allocation Contribution” is (Wjp-Wjb)(Rjbl) … notice no Rb (overall benchmark) deducted from the second term. Also, and more importantly to me, the Security Selection component is (Rjpl-Rjbl)(Wjp) … the second term is the weight of market j in the PORTFOLIO, whereas in the micro performance attribution calculation it is the weight of the benchmark. I am confused to why these are different. Particularly the second discrepancy, the weight of the portfolio being used rather than the weight of the benchmark. We are, after all, trying to figure out the returns to security selection only, so shouldn’t we use benchmark weight? I do realize these are two diff formulas, but they are essentially calculating the same thing, just one is compared to a global benchmark and one is not, so i don’t get the difference in the formulas. For now i’ve just outright memorized it, but intuitively it doesn’t make sense. SS17 kinda sucks.
Since nobody has answered either of my posts i figured i would re-post and consolidate to one. I would really appreciate some guidance on these topics. Also, while im at it, in the calculations for derivatives it makes you solve for “D1” by using actual days in a month, so for certain months it’s 30, some 31, etc. I’m assuming this will be given on the test, is that true? [END REVISION] ************************************************************************ So after finishing the books a few weeks ago and then having a “wtf did i just spend the last 3 months reading I remember nothing” moment… i began my second pass. A few questions while reviewing IPS: 1.)It is well established that you should have 3 months after tax salary set aside, however it seems like most of the return calculations do not deduct this out of the investable asset base. They do mention it in the liquidity section though. So how are we to know when to remove it or not? 2.)For IPS construction on the returns portion, how are we suppose to know when they want a number or when they just want the conceptuals unless they specify. For example question 11 in the CFA book (IPS section) doesn’t give a return number in the answer, just a process. It would be difficult anyhow given the transient expenses for college, how would you account for that if it wanted an actual return? Capitalize it over the 5 year term and deduct from the investable asset base? Hard to tell. 3.) Question 12 says Liquidity is being met in the IPS construction question, but it clearly is not. Expenses are 150, pension income is 65, and income from the “gift portfolio” is 40, so there is a 45 shortfall, this an error? In the actual return calculation later they get it right, but for purposes of describing the liquidity section i feel like they fkd up.
- Investable asset base doesnt take out liquid investments, so why would it take out the assets for the amounts set aside (i.e. you can put in a money market, but it is still an investable asset). 2. You should always be describing the process as though it is a calculation. If they give you enough information to do the actual calculation, do it. If they are leaving you to make big assumptions, they are likely not looking for a specific number, but I would still sort of walk through the methodology. For this one, it looks like they are telling you how much it cost per year. I would take th 50k out of the investable assets, and the 40k is the annual expense for those years. This would get slightly more complicated as it only last 4 years so you would have to do multiperiod calc (which is likely why they tell you the required return). For 11, they ask for the return objective only. For others, they specifically say to include the requirement. 3. The family portfolio is invested “entirely in very safe securities” and goes on to mention T-Bills etc. So, there is obviously no liquidity crisis. If I was answering this on the test, I would mention that there is a shortfall, but they have liquid assets that could cover it.
Watch the t-bills or cash equivalents across every kind of IPS and asset allocation question: Its a silent , hidden beast , take care to not ignore it .
Jmac, I think I agree with all your points. For point 1 I guess the safe route would be to specify that they need 3 months after tax salary in liquid assets (if the asset allocation doesn’t already show this – to janakisri’s point), and then you don’t deduct it because it’s still an investment (just in treasuries). Makes sense. Point 3 is a bit sketchy, the portfolio cannot meet the expenses without liquidating, so I still think they are wrong to say it’s met the liquidity needs. Your solution of mentioning ample liquidity is probably ok, however if they dip into principal they will need an even higher return requirement (although maybe not much, depending on the amount of principal used). Since you are apparently pretty on top of this stuff, see if you can answer my post above that as well if you have time, I would appreciate it.
Read my posts in these three threads, and let me know if you’re still confused. http://www.analystforum.com/phorums/read.php?13,965608,1123712#msg-1123712 http://www.analystforum.com/phorums/read.php?13,949683,949683#msg-949683 http://www.analystforum.com/phorums/read.php?13,981752,985458#msg-985458
“(I DID NOT collapse the interaction term into selection, which is what they did and this is why the two selection terms are different)” I’m a bit slow today, if you explain this, I think I will be done with this stuff. I’ve just memorized the formulas at this point though, so not sure what use it will be, but I always try as best I can to get the conceptuals just in case they throw me off with some funky wording.
Interaction isn’t exactly an intuitive concept. Selection measures how well you pick stocks, while allocation measures how well you pick sectors - and interaction measures the combination of the two, and it’s sort of hard to explain to clients who aren’t familiar with performance attribution, so some people choose not to include it in their reports. However, nothing is going to add up if you just drop the term entirely - so instead you can just combine selection and interaction together into a single term. Consequently, the selection term is different between reading 46 and reading 47.