StudyGroup7 List Archive: Berquist/Sherman Fall 1993 Question #58

------=_NextPart_000_0008_01BDE58D.F0320000
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

This was a real head scratcher for me as well - until my 1st review of =
my flash cards when the logic (?) of this question came to me. Here's =
how I see it:

The restated paid losses are simply just a linear interpolation using =
the disposed ratios to calculate the interpolation factors.

For example: to calculate the restated 1989 paid losses at 24 months =
development, interpolate between the 1989 @ 24 & 36 months using the =
factors (.94-.80)/(.94-.78) =3D .875 and (.80-.78)/(.94-.78) =3D .125 =
---> the "restated" paid loss @ 24 months is then 1720x.875 + 2222x.125 =
=3D 1,783.

Note that when the diagonal value (.80 in this case) is LESS than the =
disposed value above it (ie. .80<.82 when using 1990 @ 24 months), you =
must interpolate between the 1990 paid loss at 12 months and 24 months.
ie. the "restated" value for 1990 @ 24 months is 1296 x =
(.82-.80)/(.82-.56) + 2246 x (.80-.56)/(.82-.56) =3D 2,173.

Hopefully this will help you. I agree that this paper lacks a more =
detailed explanation on the methods - the exhibits obviously made sense =
to the authors though!

------=_NextPart_000_0008_01BDE58D.F0320000
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

This was a real head scratcher for =me as well -=20until my 1st review of my flash cards when the logic (?) of this =question came=20to me. Here's how I see it:

The restated paid losses are simply =just a=20linear interpolation using the disposed ratios to calculate the =interpolation=20factors.

For example: to calculate the =restated 1989 paid=20losses at 24 months development, interpolate between the 1989 @ 24 & =36=20months using the factors (.94-.80)/(.94-.78) =3D .875 and =20(.80-.78)/(.94-.78) =3D .125 ---> the=20"restated" paid loss @ 24 months is then 1720x.875 + 2222x.125 ==3D=201,783.

Note that when the diagonal value =(.80 in this=20case) is LESS than the disposed value above it (ie. .80<.82 when =using 1990 @=2024 months), you must interpolate between the 1990 paid loss at 12 months =and 24=20months.

ie. the "restated" value =for 1990 @ 24=20months is 1296 x (.82-.80)/(.82-.56) + 2246 x=20(.80-.56)/(.82-.56) =3D 2,173.

Hopefully this will help you. =I agree that=20this paper lacks a more detailed explanation on the methods - the =exhibits=20obviously made sense to the authors though!