| Index insurance has been promoted as a cost-effective risk management alternative for agricultural producers in developing countries. In this paper, we ask whether spatially separated weather variables commonly used in index insurance design, such as rainfall at different weather stations within a defined geographical area, are more highly correlated at the tails. As a case study, we assess the degree of tail dependence exhibited by Iowa June county-level rainfalls using copulas. We search among various candidate bivariate copulas and, using goodness-of-fit for copulas, attempt to identify the copula structures that best explain the nature of dependence among rainfalls in adjacent counties. Our results provide strong evidence that lower tail dependence... |
