Joint Claims Frequency and Severity Distribution Modelling : A Copulas Approach

This paper uses a copula model to identify the dependency structure for motor insurance data. The study identified the poisson and lognormal distributions as the best marginal distributions for the claim frequency and severity respectively. Furthermore, six(6) different copula models were fitted for the joint distribution of the claim frequency and severity. The gumbel copula was observed to be the best copula model for accounting for the dependency structure between the claim frequency and severity for the given motor insurance data