Abstract
Weather derivatives are financial instruments whose payoffs are based on a specified weather event and are used to hedge the financial impact of weather fluctuations. In this article we focus on weather derivatives whose payoffs are based on precipitation and propose Tobit type models for predicting daily precipitation for evaluating such derivatives. Daily precipitation data take nonnegative values and are also zero-inflated. Tobit type models are suitable for such data. We illustrate practical advantage of our models with some data sets. We also apply a proposed model to an evaluation of the prices of some precipitation derivatives and their payoff functions.
