I did not even get a confirmation from CAS that I registered.
Not yet. Getting registered was a real headache for me.
No email confirmation from CAS or authorization to test. The site was so slow on registration day.
1 Like
Has anyone worked 2016 #3? (fixed reference)
Iām curious about part d, where you need to find a portfolio with the same return as portfolio 1, but the same sharpe ratio as portfolio 2.
The solution is to invest equally in A, B, and C. Crystal Clear says itās obvious thatās the solution because var(p2)/var(p1) = 3. I donāt understand that reasoning at all. Itās certainly a reasonable thing to try, but in an exam, I donāt have time to try random things and hope.
What the problem (#3, not #1) is trying to illustrate is that portfolio 1 is not efficient. The investor could either achieve more return for the same amount of risk, or the same return for less risk. Part d is asking how to achieve the same return as portfolio one, but with the increased return per risk (Sharpe ratio) as portfolio 2.
Assets A, B, and C all have identical expected returns and standard deviations, the inefficiency in p1 comes from its lack of diversification. He can achieve the same expected return by being allocated 70:30 in risk free:risky. But within the risky portfolio he is missing out on diversification.
So I think a natural thought of how to maximize diversification is to do an even split. Or if instead of maximizing, you were interested in simply matching the diversification benefit of portfolio 2, then you match the allocation to risky assets in p2 (even split 10,10,10 instead of 30,30,30).
That is the way that I thought about it.
Thanks, your last paragraph finally clicked:
The Sharpe ratio of all portfolios on a CAL is the same. If we reallocate the risky part of portfolio 1 to have the same distribution of A/B/C as #2, then it essentially becomes a different point on a CAL, so it would then have the same Sharpe ratio.
Does anyone have copies of older exams (2012 and prior)? I havenāt found anything older than 2013.
Hi CuriousGeorge,
Attached are the 2011/2012 CAS exams that were available before the CAS website refresh.
Of course, remember that a lot of the problems are on old, off-syllabus papers, so many problems wonāt be as useful for studying as the 2016-2019 exams are.
CAS Exam 9 - 2011.pdf (900.6 KB)
CAS Exam 9 - 2012.pdf (1.1 MB)
3 Likes
Has anyone done guess and check for solving algebra questions on exams before? E.g. Find the risk margin needed to achieve the target RAROC rate, and instead of solving for the unknown variable, just having a table with a lot of incremental values and copying the formula down until you find it?
I just find it would be less room for error and potentially quicker, but donāt want to try this on my last exam and turns out they dock points.
Iāve done plug-and-chug for IRR problems, where you set up a worksheet so that by tinkering with the value of a cell, you can get the numbers to match whatever was needed. Iām pretty sure those solutions are accepted as long as the formulas are set up correctly.
1 Like
Yes I did it and passed with a seven.
Iām confused on the syllabus for learning objective D5. The compete text references says,
Robbin, Ira, āThe Underwriting Profit Provision,ā CAS Study Note, as updated in 1992. Excluding Sections V, VI, and IX and related exhibits
However, the Learning Objective knowledge statement lists:
a. Calendar Year Investment Offset procedure
b. Present Value Offset procedure
c. Calendar Year Return on Equity and Growth Model
d. Present Value of Income over Present Value of Equity method
e. Present Value Return on Cash Flow method
f. Risk-Adjusted Discounted Cash Flow method
g. Internal Rate of Return on Equity Flows method
Knowledge statements c, d, and g align with sections V, VI, and IX in the text. So are these covered or not?
Edit, should have taken a few more minutes to look into this. D5 has two readings associated with it. Robbin IRR covers c, d, and g, which appear to have been removed from Robbin UW from a syllabus perspective.
Yes, thatās right. A handful of years ago it was just the Robbin UW paper on the syllabus and the three models from those sections were based on the Robbin UW paper.
Since then, the CAS added the Robbin IRR paper and now those three models are covered in that paper, replacing the three sections from Robbin UW.
Does anyone know whether any formula sheet will be provided on exam day?
No formula sheet.
Is it just me, or are both answers in the examinerās guide wrong for 2016 Q29?
Part a asks for the probability that that the actual result will require more surplus than what was allocated.
Both solutions take the sum of risk loads and back into the value of z (1.646) that produces that risk load and come up with 5% as the answer.
But that isnāt the question asked! That is just the ruin probability assumption that will give the risk load provided.
To find the probability that the surplus is inadequate, we have to calculate the surplus with V = zS - R = 32578, and add that to the premium (I think this is E(L)+risk load, right? = 5525+4071 = 9596). Then we can find is the probability that the actual loss will exceed 42174. That would be events 1, 2, or 3, so 0.5%+1.0%+2.5% = 4.0%, not 5%.
Which is a fair amount of extra work required to answer the question actually asked.
If what was given was an aggregate loss distribution or if we could only experience one of the events stated then I would agree with you
From what I gather the table of losses and probabilities is more like an event loss table where we can have any of the above occur more than once and the events are independent from each other.
For instance, the formula variance = sum(p*(1-p)x^2) is an approximation formula for variance of the aggregate loss distribution, just like the covariance formula of sum p(1-p)x1x2. Otherwise the formula for the variance of the individual loss distribution would be
sum(px^2) - (sum (px))^2.
So based on the my understanding of the paper, the solution is correct since capital is for the aggregate losses.
At least this is my understanding of the paper. Someone feel free to correct me if Iām completely wrong here.
I took a look at 2016Q29 and made a different kind of mistake. I calculated the expected loss to be the sumproduct of the event probabilities and (X+Y) losses. So expected agg loss is 5525. Risk load is given as $4,071. So I took the question to be asking what is the probability that the losses exceed 5525+4071 = 9596.
All four of the events exceed this (by a lot), so the only way to not wipe out the expected loss and margin is to have no losses. = (1-p1)(1-p2)(1-p3)(1-p4) = .995.99*.975*.95 = .9124
So the probability of any one or more events occurring is .0876
Iām clearly misunderstanding something critical.
Hope to spend the remaining days filling in gaps in my knowledge and taking practice exams. Howās everyone here doing and feeling?
Home stretch! Just a few more days to go!
Doing practice examsā¦ consuming way too much caffeineā¦