Marine insurance claims analysis using the weibull and log-normal models : compensation for oil spill pollution due to tanker accidents

Abstract

In marine insurance, the identification of the distribution of the claim size processes from given observations is of major importance. We evaluate the claim amounts due to accidental oil spill volume from tankers. Previous studies on oil spills were mainly concentrating on the relation between spill amounts and some accident factors. The trend of the spill claims is a major concern to the sustainability of insurance funds, tanker shipping and the whole oil industry. By analysing the real claim data series of IOPC (International Oil Pollution Compensation Funds), we identify the trends and reveal findings which enable policy-makers and insurers for the sufficient financial protection from oil spill damage. This study introduces the Weibull model and the log-normal model to the framework for oil spill claims. The Weibull model and the log-normal model are obtained with Markov chain Monte Carlo (MCMC) simulation method. The statistical properties of oil spill claims are determined in the time series of oil spill claims. Moreover, the relations among claim properties of different settings are examined. The log-normal of 3 groups of gross tonnage is found the best fit according to R-squared statistics and the mean squared errors.