April 2009

So how did everyone find it?

 

I was just posting the same thing!

I preferred paper 1 to paper 2. I thought both papers were fair unlike some previous years. Nevertheless, I didn't do very well.

I have some questions
-P1 Q6 The economics q - was this bw? "Explain why the gvt did this" Did you explain how the cut affects the different factors as well as the rationale behind the cut?

-P1 Q4 The risk control q - could we talk about all risk measures or only ART?

-P1 Q5 The fast food restaurants - what kind of things did you put for the factors to consider? I wasn't entirely sure what the question was asking.

-P2 Q5(i) How did you calculate the multiples?

 

Hi I agree paper 1 was better than paper 2 - q5 on paper 2 - I was unsure how to calc factors but at the end did 1 / probability.

 

P2 - Q on rating factors

This was probably one of the more easier questions - but I did not answer the question properly. So I just put the usual rating factors and said that if .... then higher mortality else lower mortality.

eg: age: higher age - higher mortality.

Given that the question has asked us to consider rating factors for a policy that offers lower prem for lower mort risks,
- Do you think this will give me any marks at all?
- I would imagine that rating factors would be similar but the explanation for them would be different. So maybe I could get a few 1/4th's for the rating factors?

Any thoughts?

 

I am not sure, there was a similar question (discuss rating factors) in 2008 so you could read the solution to that to try and guage how your answer compares.

I didn't do well on that question at all.

 

I thought the paper 2 was very difficult... they certainly weren't giving away any marks that's for sure. It was so abstract... I'd hazard to guess how I even did.

 

I took the CA1 exam for the second time this April. I failed back in september 2008. I think that CA1 is very Bad. I prepared very well for this sitting but i am still not sure if i have passed or not.

Does anyone know about the pass mark for CA1?

 

eek I thought both papers were very easy and I had way too many points to get down most of the time and needed 2 extra booklets for each paper. I think I might of slipped up a little depending on how much detail they want when they ask for discuss vs state/list.

I even left the second exam 30 mins early.

For the betting question, you should of started with the fair value pay off multiple - 1/P(event) being paid out on the event, then discount it back basically.

Then there are expense and other assumptions. You could of probably got half marks or a third for just assumptions.

I don't get what was abstart? I feel there was a very strong mapping to the core reading.

 

Wow, Fiend. I can't believe you wrote three booklets for each paper.

With regards to the betting q, why was it 1/P(winning)? I just solved 1=xpv where x=the multiple, p=the probability and v= the discount factor.

 

Not everything you read on the internet is true Even it is was, the examiners prefer concise answers.

 

This is true but I am being very honest. It worried me how easy it was and if I was missing something - I am not saying I think I passed, just that I thought I did well on the questions.

I didn't write down anything I didn't think was completely relevant to the question, so it was as concise as it could be - if they can read my hand writing of course

For the odds question, it is 1/p(event) because if you do the arithmetic then you see the multiple are in proportion to the probability of events.

Also, for example taking a simple event like flipping coins, p(heads) = 0.5, i.e. 1/2. You expect to get a head for every 2 flips, hence the zero sum multiple is a payout of 2.

I am suprised you remembered the question numbers!

For risk control talked about underwriting, experience monitoring, diversification, art, claims control and management control - These are all valid alternatives to controling risk that are not reinsurance I think?

For the gov bonds talked about everything why, the thinking behind it, possible results and what they might of done in conjunction for the parituclar reasons.

For factors to consider just put a lot of stuff down, that was relevant, from the general commercial and economic environment and also capital projects section.

 

Thanks, Fiend. I am very annoyed 'cos I just did the calc using 0.4 (the first probability) and I realised that's what I did in the first place and then I changed it. Did the q actually specify that the multiple was in proportion to the probability? It must have....

I have a few more q's I was hoping you could answer. Sorry, it's just no one I know did CA1 this sitting and I haven't been able to talk about it with anyone.

Paper 1
The loan from the benefit scheme to the manufacturing company - did you put down the canons of lending and other things specific to the scheme

For the risk control measures I did the same as you, although I probably wasn't specific enough when it came to ART.

What did you put as the advantages compared to reinsurance? I used the list in the CR for the benefits of Art together with comments on profits and contingencies loaded in the reinsurance premiums.

Paper 2
The risks avoided by taking out different policies for the mortgage - how did you write 7 marks for the three products?

 

Bookie Question

If anyone wants to know about the bookie question, there are many ways of setting odds. The (1/P) method is a prudent approach which is essentially setting the E(NPV) to 0 for each of them, A to E. The reason it is prudent is because it doesen't take credit for any of the money you might expect to make on the other losing parties when you pay out on the winning party. This calculation effectively assumes that everyone invests in the winning party, and the multiple is derived to set the NPV = 0.

A more rational/realistic assumption would be that the premiums would be allocated in line with the probabilities (although you could validly assume otherwise), giving a E(NPV) equation involving all 5 multiples, i.e. there's as many possible combinations of multilpliers as you like. You could increase the payout on one and decrease the payout on another so that the E(NPV)=0 condition is still satisfied.

A bookie probably wouldn't be happy with just E(NPV)=0. It will need expense and profit loadings. It wouldn't even be happy with E(NPV)=K say. The variance is pretty important too! The bookie could choose the 5 multipliers so that the variance of their NPV is also minimised. Although they may not be too fussed provided they are comfortable that maximum losses are capped in some way. They will have a certain risk appetite and will use whatever criteria they see fit to run the business.

Other constraints on the multipliers are related to competition with other bookies....

I wouldn't worry about this Q, I made an **** of it in the exam and wasted time on it. I think a lot of people did.

I thought the exams were probably fair relative to past exams. I thought paper 1 SEEMED more straight forward than paper 2. I had time issues in both papers as well... Would be nice to pass but I can't say that I'm very confident.

There's no telling with these exams, just have to wait and see!

Good luck everyone

 

The methods outlined are the same, just the arithmetic slightly difference, it all comes down to what assumptions you used and if you were consistent with them. Then talking about overrounds etc and business mix to maximise profits.

 

The same? One takes business mix into account and one assumes the worst possible case, resulting in a lower multiplier.

Yes - you're right. You could take any valid approach provided you could stick to your assumptions, as stated in the question.

 

They both take business mix into account... The most reasonable starting point for these types of calculation in theory and practise is to calculate the fair value or theoretical value and then when this is calculated work out what you plan to charge.

You wouldn't start any other exam pricing question with get to premium charged before working out the theoretically correct premium.

 

You're winding me up!

 

A little but the point is still valid

 

The Crux of it is, a bookie/provider/whoever can pay out whatever they like and they will definitely take credit for "assumed mix of business" (in order to be competitive), meaning they may pay a multiplier which is higher than the "1/P"-derived one, before considering profit/expense loadings.

 

Conall MagFhogartai & Fiend
you two guys seem to have thoroughly enjoyed both papers!!
am just wondering whether the Examiners put a red herring in Qn 5 by saying the multiple can also run into decimals and gave examples I think close to 4.33321 say. 1/p on 40% gives you 2.5 times and for 10% will give you exactly 10 times.
Was this to mislead the students..infact the example they gave of another competitor seemed to confirm that mutliples could run into decimals....
how about the interst rate...was that to put us off or what?

 

sure those decimals were to fool students.. btw "i+Risk Premium" was supposed to be used to discount the payout i guess...