I thought it was a reasonably fair paper. There were lots of standard questions. However, last night when I was in bed I was thinking about my answers (sad I know) and I could only recall 9 questions. I really hope that I haven't missed one out because I didn't do them in order (I tried them in order of what I thought was easiest). I was sure I checked at the end but now I'm not so sure as I remember the other questions so well. I barely slept last night because of this and I have CT2 tomorrow! If I remember correctly the paper went like this: 1. No arbitrage pricing question 2. Definitions 3. Compare the 2 scenarios 4. Bond price with CG test 5. Spot rates 6. Durations 7. ? 8. NPV 9. Exponentials/payment streams 10. Stochastic models If someone could tell me what q I'm missing here I'd be most grateful! I realise I can't do anything about it now but I can't stop thinking about it!
I think youre right about question 7. I do seem to remember it asking what the real rate of return assuming the inflation rate is 3%. I thought the exam was reasonable. Was a bit stumped with the lognormal transformation (last question). For the first part of that question am I right in assuming you can calculate the expected value using E(Sn) = (1+j)^n And Var(Sn) = ((1+j)^2+s^2)^n -(1+j)^2n or do you have to convert it into a lognormal function? Does anyone know when the exam paper (and solution) will be made available on the Actuarial website?
I agree that it was straightforrward for the most part, but fiddly, which lead me to run out of time on q 10 1 was starightforward vanilla question 2 ditto 3 ditto 4 some may be caught out by the bond not being redeemed at par 5 some will be caught out by not sttaing "no abritrage" as an assumption 6 some may have stumbled on the duration being 12.5? 7 fine... but confusing wording and I calcuated it in not the most straightforward way. 8 some may be confused by what I hope was a payment stream increasing compound yearly, but payable continiously. 9 was starightforward vanilla quesdtion 10 Ok, but ran out of time grrrr.
I think for that particular question you did need to convert to Lognormal, then multiply variance and mean by 10 and then find expectation and variance given these parameters. There is an almost identical question in the revision guides which employs this technique.
Lognormal Ques 10 That's how I did it, though I made an error with a number I used, so hopefully this is carried forward and doesn't make a drastic difference to the marks!!
Duration Question? 1 was starightforward vanilla question 2 ditto 3 ditto 4 some may be caught out by the bond not being redeemed at par 5 some will be caught out by not sttaing "no abritrage" as an assumption 6 some may have stumbled on the duration being 12.5? 7 fine... but confusing wording and I calcuated it in not the most straightforward way. 8 some may be confused by what I hope was a payment stream increasing compound yearly, but payable continiously. 9 was starightforward vanilla quesdtion 10 Ok, but ran out of time grrrr.[/QUOTE] Think I got 10.5 for the duration. Did anyone else calculate the Ia and a with an arithmetic progression?? (Since the interest of 8% i think, cancelled out the compound increase on 8% per year)???
Think I got 10.5 for the duration. Did anyone else calculate the Ia and a with an arithmetic progression?? (Since the interest of 8% i think, cancelled out the compound increase on 8% per year)???[/QUOTE] It may well have been 10.5! I used the method you state.. But I seem to remember the term being 25 years... increase 8%, valuation rate 8%, so answer would be 12.5, as all payment have same PV??
It may well have been 10.5! I used the method you state.. But I seem to remember the term being 25 years... increase 8%, valuation rate 8%, so answer would be 12.5, as all payment have same PV??[/QUOTE] Sorry.. is definitely 10.5! with 20 years..
I believe since it stated in the question that interest rates were i.i.d then you can use the simple 1+j, etc... You needed to know mew and sigma squared and mulitple them by ten for the later part however, when you were asked to find the probability of the acccumulation being more than whaterver it was (£115 i think)
q10 [QUOTE=maybe_baby;8397]I believe since it stated in the question that interest rates were i.i.d then you can use the simple 1+j, etc... You needed to know mew and sigma squared and mulitple them by ten for the later part however, when you were asked to find the probability of the acccumulation being more than whaterver it was (£115 i think)[/QUOTE] I think both methods probably would work to get the mean and varience (ie change to LogN dist or calculate mew and sigma, then pop into the 1+j etc equation.) But for the second part you def have to convert into LogN dist to calculate the Probability. Anyone else agree??
I think both methods probably would work to get the mean and varience (ie change to LogN dist or calculate mew and sigma, then pop into the 1+j etc equation.) But for the second part you def have to convert into LogN dist to calculate the Probability. Anyone else agree??[/QUOTE] I think that's correct, certainly the fact that 1+i is lognormally distributed and independent would mean that converting to lognormal distribution is a perfectly sound method to find the expectation
