On the application of credibility theory and GLMS
Abstract
Premiums are payable to an insurance company for a cover against a certain risk. Credibility
models are actuarial tools to distribute premiums fairly among a heterogeneous group of
policyholders. The problem is usually to devise a way of combining the experience of the group
with the experience of the individual risk the better to calculate the premium. Credibility theory
gives the solution to this problem. Most papers and researches on this subject are difficult to
follow through, especially without the basic knowledge of Credibility theory. They also involve
fairly recondite mathematics. This study describes the basic concept of Credibility theory and the
standard methods of finding the Credibility factor, which are the Limited Fluctuation, Greatest
Accuracy, and Bayesian. This is basically giving the areas of study there are in credibility
theory. Once we have the insurance claim experience with us, we need to fit mathematical
regression models to it. Usually, the data is assumed to be normal, and so we are restricted. The
solution to this problem is the use of Generalized Linear Models which is looked at herein in this
study. The topics at hand are generally broad and therefore for deeper comprehensive study the
reader is advised to look at the numerous original papers on the subject.
Citation
Muchina,Jotham N.;2013.On The Application Of Credibility Theory And GLMS.Publisher
University of Nairobi School of Mathematics