Re: brosius -loss development using credibility

foxr1@nationwide.com
Wed, 19 May 1999 08:23:38 -0400

I don't have time to look into this in detail, but I remember that Brosius is
one of those authors who doesn't use any rounding in his calculations. This
makes it difficult to follow the examples unless you enter all the source
numbers and calculations into a spreadsheet.

I'm of the opinion that anyone writing a syllabus paper (or any paper, for that
matter) should round interim values before using them in future calculations.
Better yet, make use of Excel's "Precision as Displayed" option. It rounds any
numbers to the decimal place shown in the cell. I've set it as the default on
my PC. If I want a more precise calculation, I simply show more decimal places.
Anyone looking at one of my printed exhibits would be able to follow the
calculations with a calculator.

I'm interested in what other opinions are on this subject.

bill g <billg799@yahoo.com> on 05/19/99 08:02:04 AM

Sent by: bill g <billg799@yahoo.com>

To: study group 7 <Studygroup7@lists.casact.org>
cc: (bcc: Robert C Fox/Nationwide/NWIE)
Subject: brosius -loss development using credibility

page 18 brosius middle of paper. reported loss ratio 36 month column.

avg(x)=.216;avg(y)=.405;avg(x^2)=.062;avg(x*y)=.106 so

b = (avg(x*y)-avg(x)*avg(y))/((avg(x^2)-(avg(x))^2); so

b= (.106-.215*.405)/(.062-(.216)^2) = 1.23; but the paper says b=1.162
does anyone know where 1.162 was derived from???

thanks for your help

bill g

billg799@yahoo.com

_____________________________________________________________
Do You Yahoo!?
Free instant messaging and more at http://messenger.yahoo.com