Well, let’s give this a shot.
1 Like
Et tu Brute?
What are you using to study?
There’s 2 others here who are going to sit for 9, I may wait and see what they do and then supplement. I did RF the last time I sat - seemed solid, I just needed to cram the material down until it was burned into my retinas - and CCE also seemed good, but I thought at times the article summaries were basically a re-write of the paper; if I’m going to read that much, I’ll just read the paper.
So … something? I’m going to slow-walk things for the next 6 weeks, get serious after the 1st of the year right as year-end stuff is going on. Probably not helping that as I switch roles, I’ll go from “rate review” to “rate review” instead of “rate review” to “off-quarter that’s used to do miscellaneous projects, that would be great for using to study.”
Starting a soft launch for studying this weekend. Hoping this exam will be one and done
How technical do you think the questions on the Mildenhall will be? The level of detail in this paper is staggering and I am struggling to remember it all.
Hard to say, and I have no insider knowledge, but I think the best you can do is start pretty broad and then drill deeper and deeper as time allows.
Will they require you to know the priority of all 13 classes of claimants from 8.2.2? Probably not, but you for sure need to know that policyholders come before stockholders.
In my view, the point of the book is to introduce “modern price allocation” Ch 14-15. A lot of the prior chapters are dedicated to justifying why spectral risk measures are the correct measure. What properties should a good risk measure have? Shortcomings of other measures?
I do think you’ll need to have a good handle on this, as you’ll need to be able to describe how the Bodoff method is different than the SRM method.
Carrying out the calculations for a chapter14-15 problem should be pretty mechanical. I forget all the steps, but you should be able to set up a table and bang them out without having to think about it. SORTBY() is a useful function.
Sort
Group
Calculate F, S, p, S*, p* (distorted probability and Survival), deltaX under limited assets
alpha, Beta
M, Q, E[L]
So in my mind the broadest possible interpretation of the text is “SRM = good”. And everything else is viewed in that context. So when they are talking about default, VaR, TVaR, etc., keep in mind, how is VaR used? What is a shortcoming, how does SRM address said shortcoming.
1 Like
Thank you @Fred for the detailed explanation.
I think context and mental framing makes the text more digestible. So maybe the Introduction chapter is worth reading even if it isn’t directly testable.
There is also a youtube video where an actuary interviews Mildenhall and Major for a couple hours. This is also a good way to hear what the main purpose of the book is, directly from the horse’s mouth, and in a conversational format.
Everybody’s different, but I’m always a proponent of reading the original source material, which I did in this case. It may not be the most time efficient approach, but it makes me uncomfortable to think that a summary article may skip something.
Leaving this here in case anyone else wants to watch it.
1 Like
Yep, I found this after I had already read the text and I still found it very helpful. Idk if it would have been more helpful if I had watched it up front or if I would have been too lost to follow what they were talking about.
What is the difference between CTE and TVaR? They both seem to be measuring the expected value given that a loss exceeds a certain threshold.
They are closely related, and in the continuous case they are equivalent. There are possible differences in the discrete case where there the p threshold coincides with a mass point.
In general VaRp(X) <= CTEp(X) <= WCEp(X) <= TVaRp(X)
That is probably about a quarter point’s worth of explanation, the text goes into more specifics and shows graphs to help illustrate.
On a related note, in examiner’s reports of prior exams you’ll see that they accepted two answers for TVaR of a discrete loss list. One for > p, one for >=p. I think they had to accept both answers in the past because different syllabus papers defined it differently. I don’t think that is the case anymore.
1 Like
How are we feeling about the exam?
Also, @Ted_Hoffman do you think you will be able to use a final 4 appearance to propel you to a successful sitting?
Sat yesterday - pretty brutal exam overall but I’m feeling good. Anyone else done?
Done today - didn’t feel like any questions were too difficult, but as with every upper exam I’ve taken, you have to move through these questions so fast. Barely finished with only a few minutes remaining.
Seriously, it seems like it’s expected that we’ve purchased third-party materials such as the RF cookbook, or else I have no idea how anyone would manage to finish in time without having a good portion of potential question responses templated and ready to go.