Well I thought the first two questions were ok and then it went downhill from there. I just can't believe this is the second time I've had to sit the damn thing and will have to take it again! Did anyone out there actually do question 5?
I attempted it at the end...left myself an hour to do it. I finished part (i) with ten minutes to go! Nightmare question
Like you Muppet I left it until the end and had no chance of completing it. I struggled through part i) and had started part ii) but got nowhere near calculating a profit figure. On a separate question - what ideas did everyone come up with for add ons for cycle insurance? I got a few, but not enough for an 8 mark list question.
Well I tried to think of motor insurance... and said things like medical cover, courtesy replacement if being repaired, breakdown!!, something to do with no claims discount, additional riders, replacement parts couldn't get much more. anyone add anything else?
haha - anyone working for cycleguard was certainly at an advantage for that paper! Seriously though - I'm glad I wasnt the only one who hated question 5! It was a nightmare. I did it last (I think everyone did) and came into it with an hour to go brimming with confidence having thought the rest of the paper had gone great. Then Q5 happened. And I think in one hour I only attempted maybe 5 marks worth of answer, by: trying part 1 one way... ...getting through about a tenth of it in ten minutes and realising it was going to take literally hours... ...so trying it another way.... Getting about a qaurter of the way through before realising it was riddled with mistakes... ...so trying part ii)... ...which relied on part i... ...and then not understanding the claims either! So just worked out all the expenses! Then realised part iii) was impossible without part ii) DOH. So maximum of 2 marks for assumptions in part i), 2 marks for some botched attempts at the part i) numbers and 2 marks (generous) for working out the expenses in part ii). Maybe 1 mark for (in desperation) writing out the basic structure of a set of accounts without any numbers in!!! 7 absolute max. What was to my mind most unfair was the links between the parts. If you couldnt do i) you were never going to manage ii) and clearly then iii) was impossible! HOWEVER, Q 5 aside, I thought the paper was very fair indeed. What this does mean is anyone who did manage even half marks on Q5 has probably passed. And theres going to be some serious bunching of marks around the 50%-60% range, with very few brilliant or terrible scores! I thought it was a little sneaky to through in an aggregate deductible (which I believe isnt in the core reading or the appendix!) - but luckily I had come accross that in a past paper. The excess question was very easy (unless I missed something). The bikes question was unusual - but mostly very "sensible guessable"...except the 8 marks for special features. Given the unusual-ness of the product I was guessing 1/2 mark for each rather than 1/4 - but even then you would need 16 for full marks! I probably managed 10 or so - but they were really scraping the barrel by the end! (cover including payment for tube tickets if the roads are closed leading to not being able to cycle to work!!!) Anyone have a crack at guessing the pass mark? If it's 60% I may have just sneaked it. If 65% then almost certainly not.
Out of interest how did you approach Q5? I treated each period where the premium was constant as a underwriting cohort. Then I worked out the average inception date (halfway between the dates the premium changed) and then worked out the proportions which fell in the calendar year (assuming yearly contracts). Is this what everyone else did? I feel like I missed something obvious. Was there some way to do it quicker? btw I also used about 1hr to do part (i) and didn't make any significant progress on (ii).
That's what I tried to do. But after about 20 minutes of doing that I realised that it was going to take forever (I've only managed about half of contract A) and then I also noticed I'd made a recurring mistake throughout it. So that's when I stopped and tried (in vain) to do part ii instead!
Thanks Greg. It seemed an odd question for the examiners to ask. I honestly can't see how anyone could have made it through the question in a reasonable amount of time. I think I spent about 10 mins investigating another (innappropriate) method then spent 50 mins thinking through and turning the handle on the earning calculation. Even allowing for someone hitting the right method straight away I just don't see how you could get through the question. Do the examiners intentionally set questions they don't expect you to finish just to see how you cope?
I've always thought STs were about 45%. And SAs 40% Not sure why. But having looked at the paper online again I can see that I haven't even got that many marks!!! This question has prob been asked many times, but the Profession doesn't reduce the pass mark if there are lots of fails does it? I thought they just failed more people? Thanks
Other than question 5 I was not quite sure about question 2 part iii) on the September 2008 exam. I know they where asking for an interval. The way the question was phrased suggested that you should take 7% of the standard deviation of S as the margin i.e. 7% each side of the mean. I did not feel this was correct but that is how I answered as that is what was being asked for. I think the question was trying to get you to say that n was large and that the standardized version of S ( subtract the mean then divide by the standard deviation) was in fact standard normal. As such the confidence interval would have been based on a multiple of the z value from the normal distribution and the standard deviation of S. I actually teach this stuff in university and would like to know what happens when a question is not phrased correctly? Thanks
I think within limits yes you're right Catchy. But there does have do some form of limit in place. They wouldnt go as far as to fail 90% of candidates (perhaps except for, say, SA1 where the sample size is so small). Or indeed pass 90% for a very easy exam. I believe (possbily wrongly) that the idea is that they pass everyone who "meets a reasonable standard" - and if they set the mark at a mark such as 50% and virtually no-one hit it I think it would be pretty harsh to suggest that "no-one hit that standard" - so they would revise upwards. But I think the marks may have to be pretty dire before they resort to that. You may have done better than you think catchy. Everyone always remembers the hard exam questions most. And especially if they're also the ones you tackle last (e.g. Q5 here!) I suspect a lot of people scored well on the rest of the paper - but have since forgotten that cos all they can remember is the nightmare that was Q5! I'm sympathetic to anyone who decided it may be good to "get the hard question out of the way first" though (sometimes a tactic I do use!) - as I think that would have put you off the rest of the exam big time!
I think (this was about 2 weeks ago and my memory is hazy) the question included the word "contingency reserve" and my gut instinct was no-one would set a contingency reserve BELOW a best estimate. So I put the answer to part ii) as the lower limit and then calculated an upper limit (I think in the way you suggest - though dont quite remember the specifics). However, I put a note to the examiner in capital letters saying this was my logic. So if they wanted a two way interval they may give me some credit for my "1 way interval" logic. If you're not sure it's always worth putting a comment down as to what you ASSUME the question wants - therrby getting at least some marks if you dont do it the way they want but you've explained why not and can understand why. Incidentally, you'll get more responses if you post in the ST3 section in the EXAMS part of the forums, rather than the gneral subject info part. G
