---------- Forwarded Message ----------
> This will seem like a really dumb question, but how do you model t=
he
> aggregate (frequency x severity) distribution with @RISK, once you=
> have selected a frequency and a severity? I can see how to model =
with
> one distribution (very easy), but how do you put the two together?=
> =
I did this when I wrote the paper I presented at the Reinsurance Seminar.=
It had been a long time since I'd worked with simulations and I =
wanted to do it at a level where I could see what was going on- =
not just take the input out of a black box.
It was messy, but as an example, for 100 trials, I first generated
100 claim counts. Then I generated enough severities to cover the =
likely number of claims. (If the expected number per year were 15, =
I'd generate 2000 to be safe.)
Then I had an Excel spreadsheet which picked the next n severities =
from the 2500 for each trial. EG if the counts for the first =
3 trials were 10, 15, and 18, the spreadsheet grouped the =
first 10 severities, then the next 15, then the next 18...
On top of that, I was calculating excess layers before aggregating =
them. This process choked my laptop (I had to de-install Doom II)
but it worked and it gave me enough detail that I could verify that =
the computer was doing what I expected it to do.
Regina Berens
MBA, Inc.- Consultants in Casualty Actuarial Science
http://ourworld.compuserve.com/homepages/MBAInc
Visit the CAS Web Site at http://www.casact.org
===============================================
To subscribe or unsubscribe from CASNET:
Send an e-mail to caslists@lists.casact.org
Type in the body join casnet to subscribe
or leave casnet to unsubscribe.