Option pricing of weather derivatives based on a stochastic daily rainfall model with Analogue Year component

Abstract

In this study, we analyzed option pricing of rainfall derivatives based on stochastic daily rainfall model. We used Markov Chain Analogue Year model (MCAY) in order to describe occurrence process of daily rainfall. We have included the Analogue Year (AY) component in the Markov Chain (MC), which is a new component incorporated in this study and pricing rainfall derivatives. The inclusion of AY in the MC, provides excellent description of the occurrence process of daily rainfall. The amount of daily rainfall on wet days is obtained using Mixed Exponential distribution because it has the advantage of a better representation of extreme events. Combining the occurrence and amount model, we obtained Markov Chain Analogue Year Mixed Exponential model (MCAYMEM). Daily rainfall data from 2005 to 2017 were taken from Ethiopia National Meteorology Agency (ENMA) in order to assess the model performance. Based on the results of the daily rainfall models, we calculated an option price for different months. The price calculated using MCAYMEM gave an excellent result compared to the price calculated using MCMEM. This accuracy is mainly because of AY component included in the MC in the modeling of occurrence process.

Pricing Gamma Based Temperature Derivatives

“…Alaton et al (2002), applied martingale methods to value weather call options; however, this approach contradicts the Black-Scholes assumption of market completeness due to the non-tradability of temperature. This study uses the method of discounting expected payoff, incorporating temperature model information and probability density (Berhane et al, 2020).…”

Rainfall study based on ARIMA-RBF combined model

“…Low rainfall can lead to drought, while high rainfall can affect the growth of crops, leading to pests and diseases, and have an impact on people's lives. Therefore, the research and prediction of rainfall has been the focus of domestic and foreign scholars, for example, Poornima et al [1] used the FFNN model (Feed Forward Neural Network) in ANN model (Artificial Neural Network) to predict rainfall under various conditions; Kusiak et al [2] used random forest model, neural network model, categorical regression tree model, support vector machine model, and k-nearest neighbor 5 algorithm to predict rainfall data from radar echoes; Berhane et al [3] used Markov chains to predict monthly rainfall in a region and used the predicted values to price rainfall index options for analysis. All these studies have promoted the development of rainfall prediction analysis, but there are still some problems, such as limited applicable data types and insufficient fit to the data.…”