• We are pleased to announce that the winner of our Feedback Prize Draw for the Winter 2024-25 session and winning £150 of gift vouchers is Zhao Liang Tay. Congratulations to Zhao Liang. If you fancy winning £150 worth of gift vouchers (from a major UK store) for the Summer 2025 exam sitting for just a few minutes of your time throughout the session, please see our website at https://www.acted.co.uk/further-info.html?pat=feedback#feedback-prize for more information on how you can make sure your name is included in the draw at the end of the session.
  • Please be advised that the SP1, SP5 and SP7 X1 deadline is the 14th July and not the 17th June as first stated. Please accept out apologies for any confusion caused.

Oct 2009 Exam Discussion

I used the normal approximation..

Could anybody else confirm which approach is correct/appropriate..??

I recalled that I used exponential distribution which was given in Tables. Since exponential distribution has memoryless propery it was simplified to taking the integral of exponential distribution from 0 to 30.

But i did made a mistake of specifying the prob of ruin rather than prob of not ruin
 
I agree with evergreen.. The exam was a bit more difficult in comparison to last three sittings.. Proving the time series identity (Q1), writing the linear predictor (Q4), Calculating the mean & variance through percentile method (Q5), Proving E(1/X) (Q6), Stating in the form of credibility form (Q6), Calculating the given time series data (Q7), Proving the equilibrium states in terms of k (Q8) are the major things which made this exam more difficult in comparison to last three sessions..

I agree with both of you. I did got stuck upon various question particularly time series questions, Q1 and Q8, and the question of credibility theory and loss distributions. The question to find the estimated parameters of lognormal distribution using method of percentiles was also tough.

I notice questions were more trickier than last few sittings and moreover were time constrainted
 
Any idea regarding the passing mark this sitting..??

I recalled that I used exponential distribution which was given in Tables. Since exponential distribution has memoryless propery it was simplified to taking the integral of exponential distribution from 0 to 30.

But i did made a mistake of specifying the prob of ruin rather than prob of not ruin

I used normal approximation however, I would appreciate if someone from Acted comment on the right technique of solving the problem
 
What did you guys do in the last question's last part? I tried to solve the problem twice but I am not confident.
 
What did you guys do in the last question's last part? I tried to solve the problem twice but I am not confident.

I think it stated that the insurer paid the amount of each claim that followed an exponential distribution. In addition to that, 50 had to be paid for each claim above 200.
 
I agree, it was somewhere around 350.

What was the form of "k" in terms of q? Any clue on that? I think I was able to bring the policyholders' distribution in terms of k but then did something wrong with the calculation of average premium stuff. My answer was quite higher as compared to the 178 value of avg premium for low risk policyholders.

the transition matrix was like :

q 1-q 0

q 0 1-q

q2 q(1-q) 1-q
 
I think it stated that the insurer paid the amount of each claim that followed an exponential distribution. In addition to that, 50 had to be paid for each claim above 200.

The expected value was the product of lambda and expected value of Y.
Expected value of Y = Expected value of X + 50 (integral of f(x) from 200 to infinity)
 
Q No 6 : Cumulative claims were given and reserve was required to be calculated by using:

(a) Chain Ladder Method (6)

(b) Bornhueter Ferguson Method (6)

Any comments on whether to include the incurred claim values while calculating reserve?
 
Q No 4 : It was a GLM question.

(a) Calulate the MLE of theta(ij) (3)

(b) Express it in terms of exponential family (2)

(c) A predictor function was given as alpha + Beta * X and it was required to recommend any other suitable function. (3)

I got the answer to part (a) as 1/(1+y)

Could someone please confirm this..!
 
Last edited by a moderator:
Any comments on whether to include the incurred claim values while calculating reserve?

I seriously doubt that this question is ambiguous as the given claims were on incurred basis while the paid claims were not given on the other hand..

I just read in one of the older threads that we can write to the institute regarding these kind of ambiguities on Student Consultative Committee but that link is not working..
 
Q No 8 : This was a time series question and various sample values were given.

(a) Calculate the values of a1, meu and sigma (7)

(b) Apply turning points test (3)

Could someone please confirm if the probability was coming too high for turning point test..??
 
The expected value was the product of lambda and expected value of Y.
Expected value of Y = Expected value of X + 50 (integral of f(x) from 200 to infinity)

I applied the same method for E however calculating V was a bit tricky
 
My matrix was the same also. I couldn't get the average premium to equal their value of 178 either - is there anyone that could?!

I don't think we were required to confirm the value of that premium for low risk policyholders, nonetheless, it would have verified the preceding calculations. What was the expression for "k" any idea? I think it was something like :

1/(1-q+2q2-q3)
 
Back
Top