So how did people find it? Not really how I expected, and not happy with a lot of my answers, but I thought it seemed a fair paper.
St9 why was it different to your expectations? I'm thinking of sitting this subject and would be grateful for more feedback on the exam itself and how you found the course. What did others think of the exam??
A few gripes, but overall I think those who pass the exam will be those who have the best grasp of ERM as a whole. 1. No really specialist questions - I thought you could answer most of the exam from just knowledge of core reading. I think reading Lam, Chapman & McNeil didn't help very much at all, and ditto the papers in core reading(on model uncertainty & rating agencies) didn't help. For those thinking of taking this subject, working through the course notes, the three books that come with the course notes, and all the Acted questions will take months - it took me longer than working through ALL the material for CA1! It wasn't really necessary for this exam. 2. The acted questions were useful, but perhaps a little too technical for the exam, which was more general knowledge. I felt the actual exam was similar (& a little easier) than the specimin paper. 3. A few non-standard questions which required thinking on your feet - this is normal for an ST though, and they weren't v. hard. 4. The way they examined numerical material was poor. We were told numeric stuff would be 20-30% of the exam (as in the specimin paper) but it was only 10%. And of that 10%, there were no calculations and I guess two or three of the marks were for writing down the formula for Kendall's Tau. Why give us a special formula sheet for ST9 and then set a question which asks you to state a statistical formula which isn't on the sheet? The ST9 formula sheet wasn't used at all in the exam. Having said all that, I thought the paper was a good one to sit, better than I expected, and I did alright on it. Hoping for a pass!
Oh yes..... there is way too much in that course. And it feels bigger than CA1. Maybe in a few sittings Acted will have a better idea what the Institute were actually after and produce a more modest set of notes for students as opposed to read through 2 books, 2 binders and as much of QRM as you can stomach.... (Which in some respects is a shame as the numerical material was far and away the most interesting part of the course for me.) And yeah.... when I say not what I expected, the almost utter lack of specialist knowledge questions was what I meant. The course notes helped to generate ideas for some of the later general knowledge questions of course, but I just wasn't expecting a more detailed CA1. (I was also slightly surprised there seemed to be nothing on the case studies.) Haven't done an ST before. I'll take your word for it that it's standard. Wasn't very impressed at the question on inflation swaps as I knew nothing of the wretched things before sitting the exam and had to guess what they were. And whilst I'm griping I was amused by the advice in the notes that there are no settled definitions for terms like "risk appetite", "risk tolerance", etc. and candidates should concentrate on understanding the ideas behind them rather than learning the definitions by heart. With hindsight, I'm embarrassed that I fell for this " helpful advice." I must admit I was expecting a question that required you to know the formula for the generator of the Gumbel copula or something equally dumb. Knowing the formula for Kendall's Tau was reasonable relative to that, but the point of memorising formula for an exam that you'll forever more look up if you ever use them still escapes me....... Still kicking myself for the hours of revision I "wasted" on multivariate processes. Yeah.... it wasn't too bad as you say in spite of my whines. Good luck. Not so sure myself how I've done, but might scrape through...
