Analisa Struktur Dependensi Variabe Pembentukan Asuransi Pertanian Berbasis Indeks Cuaca dengan Multivariat Copula dan Vine Copula

Abstract

The purpose of this study is to analyze the structure of the dependency on variables for calculation of insurance based on weather indices such as crop prices, yields, and rainfall. The object of research observation was secondary data on the sub-district of Dlingo Bantul District. In analyzing the dependency of variables that can be used in agricultural insurance calculations, it can be seen that both using multivariate copula and vine copula have the same results. A multivariate copula that directly looks at dependency relationships between three variables. Whereas copula vine can see the size values ​​of the variable pair dependency for each edge in the copula vine tree. In more detail the best dependency for the grain price and rainfall variable is Copula Joe with the parameter θ = 1.76. correlation τ = 0.3. The best dependency between rainfall and yield is Frank Copula with parameters θ = 4.98 and correlation τ = 0.46. The best dependency between rainfall and yield is Frank copula with parameters θ = 2.42 and correlation τ = 0.25.