Skip to content
Prev 165563 / 398506 Next

Percent damage distribution

Good idea. At this point I do not need to be so precise as to work with a
mixture distribution, but I will keep this in mind.
I am trying to model loss severity. One common simplified approach is to
sample from e.g. a gamma or lognormal distribution to determine the dollar
value of each loss. My problem with this approach is that I have the
individual insured amounts, so a $100,000 loss which could result from
sampling from a lognormal distribution does not seem reasonable if the
insured amount is $25,000, to put an example. That is why I thought of a
damage distribution instead.
I am not sure what you mean by using a plot analysis or re-sampling the
data. 

I posted back to Ben yesterday and the post was not accepted yet, so it
probably does not show in the thread, but there I stated I was going to use
a beta distribution, so my problem is solved by now. If you want, we may
continue this conversation privately.

Many thanks.
On Thu, 25 Dec 2008, diegol wrote:

            
-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595
______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.




-----
~~~~~~~~~~~~~~~~~~~~~~~~~~
Diego Mazzeo
Actuarial Science Student
Facultad de Ciencias Econ?micas
Universidad de Buenos Aires
Buenos Aires, Argentina