Clustering of heal risk and determination of aggregate loss distribution by Fast Fourier Transform in agricultural insurance

It is a great importance for insurance companies to analyze the catastrophic risk that posess low probability and high severity by nature. With drastic changes in climatical conditions it became an obligation to form the risk maps and to compute the aggreagate loss for agricultural insurance that exposed to catastrophic risks in order to increase production yield and decrase economic loss. In this study, we aim at obtaining loss map for hail which has the biggest impact on production and computing the loss distribution for wheat caused by hailstorms. Firstly by using damage frequency, spatial scan statistics is employed for clustering purposes to build the hail loss map for Turkey. Then, each clusters are evaluated in Panjer Recursive, Monte Carlo Simulation, Normal Distribution Approimation and Fast Fourier Transformation algorithms to build the aggregate loss distribution.