9: Mango and Walters/Morin

[Ike Broslofski]

Umm yeahhh....
Does anybody have any problems on these papers?
If yes, where did you get them?

Thanks!

[Maine-iac]

Only those devised by my study group . . .

[whisper]

Since these are new papers, the people writing the exams won't have much experience writing questions based on them either. Also, there will be no reason to come up with "unique" questions to avoid repeating past questions. Read the paper and the learning objectives, and you should be able to develop a good number of questions that is likely to appear on the exam.

[Cynical Realist]

The learning objectives will be very helpful... :lol:

Hey, I have those study group questions - I hope they help!

[jmls]

Hi y'all,

Feldblum wrote some possible questions for Mango. I guess he figures that they might ask one to list properties of the various methods (Shapley vs. covariance share etc.), and/or give an example of how the risk margins are allocated under the different method. Also they might ask what subadditivity or superadditivity means.

As far as Walters/Morin, your guess is as good as mine (or better!) I'm guessing either a "list" question or a "calculation" question :) :)

[Ike Broslofski]

Thanks everyone for your suggestions... somebody i work with went/is at the neas seminar so I will be able to see what questions Sholom has come up with.
I realize that I could try to make up my own questions, but I've got enough issues right now.
One of the main things with Mango is that his Exhibits 3 and 4 Covar Share Columns don't seem to line up with his formulas that he mentions in Section 10 (formulas 10.2 and 10.3)... the formulas in the section had a p * (1-p) included in the formula for Xi and Yi, but the table in the back does not calc that way for Xi and Yi (it does not include the p * (1-p) ).
Has anybody else seen this? If yes, do you understand it?

[MNBridge]

Here is a question I came up with.

If you find it useful I ask that you post another question here that you think may be useful to others.

Here it is is let me know what you think.
EXTRA CREDIT -- PART D -- Find the Risk Loads using the Covatriance Share Method.


Insurer A has one Risk X. The company is adding Risk Y.
E(X) = 10,000
Var(X) = 100,000,000
CoVar(X,Y) = 20,000,000
E(Y) = 5,000
Var(Y) = 50,000,000

The Company prices policies at @ 20% return on marginal surplus.

The company prices to a Z = 2.

*Note assume to risk load for the variance method = risk load for the surplus method / St Dev of Portfolio. (Not sure if we need to know that)

PART 2

The company is now at renewal season.and has elected to assign risk loads to risks X & Y.
(C) What is the risk load needed for each Risk? Using the Shapley Method

[MNBridge]

Thanks everyone for your suggestions... somebody i work with went/is at the neas seminar so I will be able to see what questions Sholom has come up with.
I realize that I could try to make up my own questions, but I've got enough issues right now.
One of the main things with Mango is that his Exhibits 3 and 4 Covar Share Columns don't seem to line up with his formulas that he mentions in Section 10 (formulas 10.2 and 10.3)... the formulas in the section had a p * (1-p) included in the formula for Xi and Yi, but the table in the back does not calc that way for Xi and Yi (it does not include the p * (1-p) ).
Has anybody else seen this? If yes, do you understand it?

Now to answer your question.

I'm not sure I understand what you are referring to here.

What I get out of table 3 and 4 is that it is showing that during renewal the st dev method and covariance methods do not work.
i.e. they are order specific.

It shows that the St Dev Build up has a Risk load that is to high and the Covariance wethod is to low (Don't quote that or use it without checking -- I'm not totally clear on this)

But I do think that is the goal of Table 3 & 4. I could be wrong.

[MNBridge]

THIS POST IS THE ANSWER TO MY QUESTION.

If someone understands part D I'd love to see it.


(A) Marginal Surplus Method
(B) Marginal Variance Method

Part (A) X Y Porfolio
Change in St Dev 10,000.0 3,784 13,784
Risk Load Multiplier* 0.333 0.333 0.333
Risk Load $3,333 $1,261 $4,595

*Risk Load Multiplier =y * z / ( 1 + y) = .2 * 2 / (1 + .2) = 1/3

Part (B) X Y Porfolio
Change in Variance 100,000,000.0 90,000,000 190,000,000
Risk Load Multiplier* 0.0000242 0.0000242 0.0000242
Risk Load $2,418 $2,176 $4,595

*Risk Load Multiplier =Risk Load multiplier for Surplus method / Portfolio St Dev = 1/3 / 13,784

The company is now at renewal season.and has elected to assign risk loads to risks X & Y.
(C) What is the risk load needed for each Risk? Using the Shapley Method

Part ( C)
Change in variance X Y Portfolio
Shapley Method 100,000,000 50,000,000 190,000,000
add 1 Cov to each 120,000,000 70,000,000 190,000,000
Risk Load Multiplier 0.0000242 0.0000242 0.0000242
Risk Load 2,902 1,693 4,595

[cas_student]

For Part D I get $3065 for x and $1533 for y. (10,000)/(15,000)*2*20,000,000*0.0000242+100000000*.0000242 =3065 and (5000/1500)*2*20,000,000*0.0000242+50000000*.0000242=153 3.

Please let me know if this matches your answer.

[MNBridge]

For Part D I seriously meant I didn't understand the weighting.

Can you match to the numbers in the paper?

I will be meeting with some people at lunch to see if we follow what you did.

Thanks and I will get back to you.

[Dooby Scoo]

Thanks, this example was very helpful.