Skip to Content

Dumaria Tampubolon


Name: Dumaria Tampubolon

Course: PhD;ACST

Department: Department of Actuarial Studies

Staff Supervisor: Prof. Piet de Jong

Associate Supervisor: Dr. Glen Barnett

Email Address: dtampubo@efs.mq.edu.au

Thesis Title

Uncertainties in the Estimation of the Outstanding Claims Liability in General Insurance (Awarded 2007)

Abstract

General insurance companies are required to set aside (reserve) enough of their premium income to cover future claim payments from past and current policies. If the forecasts of future claims are poor, the company may fail – a factor in the failure of HIH and several overseas companies. In any statistical modeling, there are several sources of uncertainties and it is important to gain information on, and furthermore, to measure these uncertainties.

The thesis addresses some of the important issues related to the uncertainties of an estimate of the outstanding claims liability for long-tailed business. There are three main problems discussed in the thesis: developing a framework to perform diagnostic checking on models belonging to a family of link ratio models; developing a measurement to quantify the sensitivity of an estimate of the outstanding claims liability to small perturbations in the data; and predicting the 75 th percentile of the outstanding claims distribution (given Hertig’s model). Working in the framework developed in the thesis, a claim analyst is able to gain information on the uncertainties related to: the model specification or selection; the parameter(s) of the employed model; and the process generating the data.