Pricing weather index insurance based on artificial controlled experiment: a case study of cold temperature for early rice in Jiangxi, China

Abstract

The growth of early rice is often threatened by a phenomenon known as Grain Buds Cold, a period of anomalously cold temperatures during the booting and flowering stage. As a high yield loss due to Grain Buds Cold will lead to increasing insurance premiums, quantifying the impact of weather on crop yield is crucial to the design of weather index insurance. In this study, we propose a new approach to the estimation of premium rates of Grain Buds Cold weather index insurance. A 2-year artificial controlled experiment was utilized to develop logarithmic and linear yield loss models. Additionally, incorporating 51 years of meteorological data, an information diffusion model was used to calculate the probability of different durations of Grain Buds Cold, ranging from 3 to 20 days. The results show that the pure premium rates determined by a logarithmic yield loss model exhibit lower risk and greater efficiency than those determined by a linear yield loss model. The premium rates of Grain Buds Cold weather index insurance were found to fluctuate between 7.085 and 10.151% at the county level in Jiangxi Province, while the premium rates based on the linear yield loss model were higher (ranging from 7.787 to 11.672%). Compared with common statistical methods, the artificial controlled experiment presented below provides a more robust, reliable and accurate way of analyzing the relationship between yield and a single meteorological factor. At the same time, the minimal data requirements of this experimental approach indicate that this method could be very important in regions lacking historical yield and climate data. Estimating weather index insurance accurately will help farmers address extreme cold weather risk under changing climatic conditions.