Understanding deductibles

[The Spocker]

Am I the only one having a hard time grasping these? I believe my main problem is not understanding exactly what the terms mean.

for example, in question 7.6, what is the difference between average loss below deductible and E[X^d]? i thought they were the same.

Getting into lesson 8 and bonus's, I don't understand like payment per claim/loss and whatever terms they use. Are those terms what is actually paid out? Or is it just the claim amount?

Is there a table, or even a post that organizes these in an easy to read/memorize table? I think it would help me a lot more in figuring out these calculations.

[Abraham Weishaus]

To answer your second question, E[X^d] is an average that includes losses above the deductible (d for each) in the numerator and a count of all losses in the deductible. The average loss below the deductible only includes losses below the deductible in the numerator and a count of losses below the deductible in the denominator.

For your third and fourth questions: loss and payment are pretty clear. Claim is sometimes ambiguous, and requires context to figure out whether it means loss or payment.

[The Spocker]

E[X^d] is an average that includes losses above the deductible (d for each) in the numerator and all a count of all losses in the deductible.
So let's assume there is a $2,000 claim with a $500 deductible. Are you saying the numerator is $1500 and the denominator is $2000? Doesn't that make E[X^d] a ratio, which I know isn't correct.

[The Spocker]

Looking at question 8.5 in your manual, It says the average payment per loss is $2500. It asks you to solve the average loss, which ends up being $4086. So the average payment per loss is the actual amount the insurance company pays out (after deductible)? And the average loss is basically the average claim amount (not factoring in deduct)?

Then in 8.6, it says the average payment per paid claim is $2500. I don't understand the difference between that and the average payment per loss in question 8.5.

thanks for your help abe. I'm signing up for you mfe seminar too, so I can harass you in person :dsmile:

[Actuarialsuck]

Looking at question 8.5 in your manual, It says the average payment per loss is $2500. It asks you to solve the average loss, which ends up being $4086. So the average payment per loss is the actual amount the insurance company pays out (after deductible)? And the average loss is basically the average claim amount (not factoring in deduct)?

Then in 8.6, it says the average payment per paid claim is $2500. I don't understand the difference between that and the average payment per loss in question 8.5.

thanks for your help abe. I'm signing up for you mfe seminar too, so I can harass you in person :dsmile:

If you want to type out 8.5 and 8.6 we can work through it

[Abraham Weishaus]

So let's assume there is a $2,000 claim with a $500 deductible. Are you saying the numerator is $1500 and the denominator is $2000? Doesn't that make E[X^d] a ratio, which I know isn't correct.
I said the denominator is the count of losses. So it is 1.

[Abraham Weishaus]

Looking at question 8.5 in your manual, It says the average payment per loss is $2500. It asks you to solve the average loss, which ends up being $4086. So the average payment per loss is the actual amount the insurance company pays out (after deductible)? And the average loss is basically the average claim amount (not factoring in deduct)?

Then in 8.6, it says the average payment per paid claim is $2500. I don't understand the difference between that and the average payment per loss in question 8.5.


Not every loss is paid.

[Abraham Weishaus]

thanks for your help abe. I'm signing up for you mfe seminar too, so I can harass you in person :dsmile:
But only about MFE topics.

[The Spocker]

Not every loss is paid.

:wall:

thanks for your help abe. I think I'm going to come back to this topic later. For now, I need to keep moving forward.

[The Spocker]

So let's assume there is a $2,000 claim with a $500 deductible. Are you saying the numerator is $1500 and the denominator is $2000? Doesn't that make E[X^d] a ratio, which I know isn't correct.

after watching broverman's dvd, i see he is talking about loss limits. in his example, he gives and equation, then says the insurance policy pays amounts up to a limit of 1000 per claim. calculate the expected payment under this policy for 1 claim.

then he sets it up as E[X^1000].

sigh.

[Abraham Weishaus]

:aypi:How is that different from what I said?

[Actuarialsuck]

X \wedge x just means min(X,x) so if you have a policy limit of 1000, for the insurer, they will pay X if it's less than x i.e. if X < x then min(X,x) = X or if X is greater than x they will pay x i.e. if X > x then min(X,x) = x.

[Jim Daniel]

The textbook calls what Actuarialsuck described a "coverage limit". The book calls the maximum possible _payment_ created by a policy modification the "policy limit". A coverage limit causes there to be a policy limit; if there's no ordinary deductible and no co-insurance factor then the policy limit equals the coverage limit.

Jim Daniel

[The Spocker]

Thanks everyone. Do you feel I am in serious trouble if I am struggling with these concepts?

[badmaj5]

Thanks everyone. Do you feel I am in serious trouble if I am struggling with these concepts?

I wouldn't get too stressed out just yet. I don't think I really fully understand it all in super great detail yet either, but we still do have over 3 months. If we start getting down on ourselves already, it is going to make for a loooong 3 months. I'm just trying to get the basics down, see the notation and get through the material for the first time before the exam seminar. Then after that I'm going to really try and start mastering stuff like this that is tricky.

At least that is what I'm telling myself... :)

[Sparktz]

Thanks everyone. Do you feel I am in serious trouble if I am struggling with these concepts?

I am just trying to keep my head down and get through the material. I don't know if this is common, but I seem to struggle through things the first time I see them, but then later it starts to make sense.

[Dahlia]

Thanks everyone. Do you feel I am in serious trouble if I am struggling with these concepts?

All you can do once you get the basic concept is to do as many problems as you can. There are a ton of questions on these topics in the released SOA sample questions. Which manual are you using?

One of the first things to get clear imo is that most questions will ask either for:
A payment per loss
or
A payment per pmt (this involves dividing by some probability of being greater than the deductible).

[The Spocker]

All you can do once you get the basic concept is to do as many problems as you can. There are a ton of questions on these topics in the released SOA sample questions. Which manual are you using?

One of the first things to get clear imo is that most questions will ask either for:
A payment per loss
or
A payment per pmt (this involves dividing by some probability of being greater than the deductible).

I am using ASM.

I never saw these SOA sample questions you mentioned... can you please provide a link? All I see is the old exams, if that is what you mean.

[Dahlia]

C 09- 08: Soa 289 sample questions (mentioned in the Exam study note) (http://www.soa.org/files/pdf/edu-2008-spring-c-questions.pdf)

It's very early for you to start work on these but I remembered there were a chunk of deductible problems around #80 something.

[The Spocker]

C 09- 08: Soa 289 sample questions (mentioned in the Exam study note) (http://www.soa.org/files/pdf/edu-2008-spring-c-questions.pdf)

It's very early for you to start work on these but I remembered there were a chunk of deductible problems around #80 something.

thank you very much dahlia!

[silvergrey]

Sorry to kick this up again. I was also a little bit confused about the definition before. Right now, I think I am some kind of clear about "Payment Per Loss" and "Payment Per Payment", but who could explain to me what's the diffence of "Payment Per Claim" and them? I think it should be equal to one of them.


Thanks:oops:

[Vorian Atreides]

Sorry to kick this up again. I was also a little bit confused about the definition before. Right now, I think I am some kind of clear about "Payment Per Loss" and "Payment Per Payment", but who could explain to me what's the diffence of "Payment Per Claim" and them? I think it should be equal to one of them.


Thanks:oops:
In practice: consider that not every claim will be covered--this is not saying that someone will file a fraudulent claim, but that sometimes a claim is filed that turns out to not be covered. In this case, the count for the denominator will include this claim for "payment per claim" calculation, but it wouldn't be counted for the "payment per loss".

For the exam: I would associate "payments per claim" with "payments per loss" for calculations. However, read the problem carefully to ensure this is proper. I would not be surprised if one of the answers is the result of working out the problem under the "payments per payment" method and another being from calculating under the "payments per loss" method.

The best way to prepare: understand the two problems types and work on identifying which method is the correct method quickly.