Evaluating effectiveness of rainfall index insurance

Purpose - The purpose of this paper is to examine the reduction of crop yield uncertainty using rainfall index insurances. The insurance payouts are determined by a transparent rainfall index rather than actual crop yield of any producer, thereby circumventing problems of adverse selection and moral hazard. The authors consider insurances on rainfall indexes of various months and derive an optimal insurance portfolio that minimizes the income variance for a crop producer. Design/methodology/approach - Various regression models are considered to relate crop yield to monthly mean temperature and monthly cumulative precipitation. A bootstrapping method is used to simulate weather indexes and corn yield in a future year with the correlation between precipitation and temperature incorporated. Based on the simulated scenarios, the optimal insurance portfolio that minimizes the income variance for a crop producer is obtained. In addition, the impact of correlation between temperature and precipitation, availability of temperature index insurance and geographical basis risk on the effectiveness of rainfall index insurance is examined. Findings - The authors illustrate the approach with the corn yield in Illinois east crop reporting district and weather data of a city in the same district. The analysis shows that corn yield in this district is negatively influenced by excessive precipitation in May and drought in June–August. Rainfall index insurance portfolio can reduce the income variance by up to 51.84 percent. Failing to incorporate the correlation between temperature and precipitation decreases variance reduction by 11.6 percent. The presence of geographical basis risk decreases variance reduction by a striking 24.11 percent. Allowing for the purchase of both rainfall and temperature index insurances increases variance reduction by 13.67 percent. Originality/value - By including precipitation shortfall into explanatory variables, the extended crop yield model explains more fluctuation in crop yield than existing models. The authors use a bootstrapping method instead of complex parametric models to simulate weather indexes and crop yield for a future year and assess the effectiveness of rainfall index insurance. The optimal insurance portfolio obtained provides insights on the practical development of rainfall insurance for corn producers, from the selection of triggering index to the demand of the insurance.