In general and vague terms (so as not to fall foul of the request not to divulge info about the exam for a few hours)... how was it for you guys? I felt like kicking myself in all honesty. It didn't feel like a massively hard paper... but with a little bit of help from mind-blanking and taking too long on the early questions the story's not as good as it should have been.
Thoughts Personally I don't think it covered the core material very broadly at all. There were some pretty big parts of CT6 that went either completely untested or were only touched on. Kind of annoying when you try to learn everything and see an exam that focusses on just a few areas. Oh well, best of luck to those that took it this time!
I would agree. I don't think the examiners are on the right track to give so many marks for obscure sections of the course. Last time it was ARCH models and this time we had the scary 14 mark Question 9 which I don't recall seeing in the course or any Q&A, Assignment X or past exam paper!!! The exam also focussed heavily on bashing numbers through formulae with very little opportunity for comment on results or discuss the underlying assumptions. Perhaps the Institute are looking to generate a lot of robot actuaries? Not to mention that I was frozen to the core in the Hammersmith Town Hall. Only 1 heater for such a big room?
I was under serious time pressure for the last question. To get to the formula for VAR = 550000p - 100000p^2, I was getting 500000p rather than 550000. I guess it's wishful thinking to think others may have got the same answer and it may have been a typo on the exam paper?!..........................I thought so. As for the multi-variate time series question, forget about it.
Freezing! Man, I was so cold!! If they can't read my writing its because I was shaking so much! Agree...where was chapters 1, 4, 5, 14? No proofs or assumptions or anything. None of the stuff I'm good at!
I thought the paper didnt cover a lot of the subject compared to previous sittings. Saying that i was quite happy with the paper...up to question 9 which completely threw me! I spent so much time thinking about Q9 (and then not completing it!) that i missed out precious marks on Q10. I reckon q9 and 10 will be the difference between people that pass and (just?) fail.
All in all, I agree with most of you that it was not the best of papers and only covered a few areas of the course. Not very many marks for comments or assumptions. I think I managed to do Q9 relatively well (I learnt how to do it the night before!), but, I was really clueless about question 10, even after taking a lot of time to think about it, I doubt I will get anything for q10. That leaves me with questions 1-9 obviously with silly errors here and there that will reduce overall number of marks I get. I wonder how this paper will compare to the Sept07 with a 44% pass rate.
Amen to that. I wasted 45 mins (!) trying to work out how to even get started on Q10 and I'm certain I will receive next to 0 marks for my efforts. I knew i was dealing with S being a comp binomial but couldn't get the parameters right. I realise it was bad time management on my part and i really wish i'd spent that time making sure I get full marks on other Qs. Does anyone have a vague idea what the pass mark tends to be for CT6?
Yep, the paper was quite manageable until the last 2 questions. That Poisson MGF was quite a surprise, but managed to get it right using that exponential expansion. Even after going through it in my reading, I weren't able to prove the Pearson and Deviance residuals were equal. Hopefully I get some part marks for that. For the inflation question, I initially made a huge mistake taking the inflation factor to be 10% (I got so used to using this in the assignments!), but thank God for making look at the question again whilst doing it. The last question was terrible. I knew it was compound poisson but still couldn't get it. I just used that result to get to expectation. Didn't manage to complete the variance part and the rest of that question. As for the Eigenvalues, I forgot whether they needed to be < or > than 1, so I obtained them, and made sure both were either < 1 or > 1 and used that for the next part. Solving for c was a nightmare and it got messy, so I didn't complete that either. Hopefully the pass mark, is not more than 60% for this paper...I'm already feeling gutted now that my team is out of the Champions league , and possibly out of the EPL, especially considering that they are up against a firing ManUtd on Sunday (can't see those guys losing/drawing any more games now).
They generally say the pass mark for a CT subject is around 55, plus or minus a few marks depending on the difficulty of the paper. Having said that, I'm sure how stringently they mark them differs at times, i.e. if it's a tough paper, don't mark down too much for lack of working and vice versa.
Yeah, that makes sense. Agree with one of the earlier posts that Q9/10 could be the difference between a pass or fail...that being said, I now feel a lot less confident than before...
Oh well, so much for learning the accept-reject method, inverse-transform, the pdf of reinsurers claims where he makes a payment...for nothing
11 marks off syllabus I have sat the CT6 paper twice now, once with differential equations in September 07 which was off syllabus. And now this time Q9 part (ii) and (iii) were off syllabus. It asked us to prove that the multivariate time series they gave us were stationary. Once you had written it in Matrix form you were asked to calculate the “eigenvalue.” Neither of these are in the notes or syllabus (as confirmed by ACTED) and is another obscure and unfair question. The Institute has not given us an option to demonstrate what we know. Many of the students I have spoken to will be writing to the head man in charge Trevor Watkins *(Secretary to the Education and CPD Board) as well as the actuary magazine. I recommend all students do.
I'll have to check my notes at home, but I'm pretty sure the Acted notes advise you to be aware of what an eigenvalue is in relation to multivariate time series? Won't dispute that it's a bit unfair to set a huge question on what is covered by about a page in the core reading..
I agree that they do mention them in the notes however in tutorials we were advised that we would not have to calculate or prove them. But they are mentioned.
I reckon the moderators will increase the weighting on method-marks for Q9...hope it makes a difference because I didn't answer around 10-15marks of the paper
