GRASP Numerical Problems

[ajstudies]

I have been working through some of these (the ones that are posted on her site) and so far so good. Even if I couldn't remember a formula, I could at least understand the solution. On the disability reserve problem, I assumed the benefits were paid at the end of the month and the solution assumed they were paid at the middle. (In actuality, the insurer would only pay whatever benefit was due at a particular time. They wouldn't pay the entire March benefit on March 15. At least, my company wouldn't.)

I haven't started on the larger practice problem set that came separately. I am hoping for some practice problems from the GAAP text b/c I still don't have my head around that one yet.

[MarsLasar]

I've been doing these as well. They are a welcome break from memorizing.

[NerdyEve]

Can anyone explain to me how the modified claim Amount is calculated on p.749 of the extra problems posted online (it's page 17 of 22 of the document).

I think there are a few rounding errors in the problem because the probabilities don't add up to 1. But I also don't understand how the modified claim amount is $100 and $600 for the $250 and $750 claim level respectively.

Anyone can help?

Thanks

[Rainson]

I think there are a few rounding errors in the problem because the probabilities don't add up to 1. But I also don't understand how the modified claim amount is $100 and $600 for the $250 and $750 claim level respectively.

Thanks

This is an error. I will correct the problem and repost it to the webiste. I will post a note here when the corrected version is posted.

[Rainson]

The corrected version is on the website now.

[K1]

Thanks- maybe this is a really dumb question- but how do you read that matrix at the top? A first I thought it was transition probabilities of Person 2 one year to the next- then I thought it should have been labeled Person 1 and Person 2 but either way none of the rows,columns or in aggregate add up to 1 so I'm completely confused. I guess I should re-read that paper. Any help is appreciated.

[NerdyEve]

I think you're doing it right, but the probabilities have all been rounded to the second decimal so they don't add up to 1
You don't need to multiply any of the probabilities, you just need to add them up.

[ajstudies]

Problem on p. 752. I just flat don't understand this one. Can anyone explain it? I don't understand how they got from Benefit Cost of $2,575 to a "Cumulative pay as you go Amt" of $2,800. I really don't understand what a "Cumulative pay as you go Amt" is. Nor why $225 is added to the $1,000 transition obligation.


Problem on p. 753.

Why are the probabilities give for ages 25, 35, and 50 instead of 20, 30, and 50? It looks like the p's for 25 & 35 are reversed. Why are the same P's used whether you are counting years to full eligibility or years to retirement? Is it b/c you have to count every one who COULD retire, even if they end up dying/terminating before retirement?


p. 754

This looks like the Projected Unit Credit Method, but I have

NC(x) = B(y) [ N(y) / D(x) ] / [y-e]
AAL(x) = NC (x) * [x - e]

which is NOT the same as AAL(x) = B(x) [ N(y) / D(x) ] unless x = y

Do I have my formulas wrong or is this just another method, or what?

[saahu5]

p. 754

This looks like the Projected Unit Credit Method, but I have

NC(x) = B(y) [ N(y) / D(x) ] / [y-e]
AAL(x) = NC (x) * [x - e]

which is NOT the same as AAL(x) = B(x) [ N(y) / D(x) ] unless x = y

Do I have my formulas wrong or is this just another method, or what?

It's Traditional Unit Credit because the benefit is a set amount per month ($10 per year of service) rather than a portion of salary.

[ajstudies]

It's Traditional Unit Credit because the benefit is a set amount per month ($10 per year of service) rather than a portion of salary.

Since the problem doesn't specify, would you have to use a particular method? Or would you just use the method you feel is appropriate, and state what that method is?