Modeling the Price of Catastrophe Swap with Stochastic Occurrence Severity

developing a model for catastrophe swap pricing based on the stochastic models and numerical model solving.

Vrance and Pielke (2009) database was used in this descriptive and retrospective study. Ito has been followed to determine the swap price changes, and the Black–Scholes modeling method was used to reach the catastrophe swap model. A partial integral differential equation was extracted and transformed to ordinary differential equations using Semi-discretization. The Finite difference method and the Euler method were used to solve the catastrophe swap pricing model. The parameters have been estimated and implemented numerically using Bjork's (2009) statistical inference method and finally, the model was implemented using MATLAB software.

A new two-factor damage model was introduced. In other words, instead of c in the Anger model, ce to the power Lambda is used and Lambda is considered to be stochastic at any given moment. Therefore, from a view of mathematical probability, intensity value is not constant and fallows a Geometric Brownian Motion process, which is correlated with the damage. A new model for catastrophe swap pricing has also been introduced, which has two integral and differential parts.

The price of a catastrophe swap securities is inversely correlated with the growth of the damage and the increase in the severity of the damage. Besides, the price trend for damage less than the threshold, has a regular trend and these changes are proportional to the changes in the damage and intensity. JEL Classification:  G12, G13, G22