Pemodelan Klaim Asuransi Kendaraan Bermotor dengan Regresi ZAIG

Abstract

In motor insurance pricing based on risk of policyholder, modeling claim is the most important step.
The modeling includes two main models there are model which relates to event of claims and model
the cost of claims submitted to insurance companies. Most studies modeling the cost of claims
involving only the amount of claims which are positive, i.e. when an accident happens and then the
policyholder filed a claim with the claims cost is greater than zero. In one period of insurance, there’re
policyholders who have not had an accident and there’re policyholders who had an accident but does
not have claim, in this case is said to the claims cost is zero. This paper investigate the
implementation ZAIG (Zero Adjusted Inverse Gaussian) regression on the model of automobile
insurance claims that involve the cost of claims that zero and positive use data supported by
Insurance Services Malaysia (ISM) Berhard. By regression ZAIG note that both the event and the
average of claim cost significantly affected by the premium.