Abstract

With minimal moral hazard and adverse selection, weather index insurance promises financial resilience to farmers struck by harsh weather conditions through swift compensation at affordable premium. Despite these advantages, the very nature of indexing gives rise to production basis risk as the selected weather indexes do not sufficiently correspond to actual damages. To address this problem, we develop a stochastic yield model, built upon a stochastic soil moisture model driven by marked Poisson rainfall. Our analysis shows that even under similar temperature and rainfall amount yields can differ significantly; this was empirically supported by a 2‐year field experiment in which rain‐fed maize was grown under very similar total rainfall. Here, the year with more intense, less‐frequent rainfall produces a better yield—a rare counter evidence to most climate change projections. Through a stochastic yield model, we demonstrate the crucial roles of rainfall intensity and frequency in determining the yield. Importantly, the model allows us to compute rainfall pattern‐related basis risk inherent in cumulative rain index insurance. The model results and a case study herein clearly show that total rainfall is a poor indicator of yield, imposing unnecessary production basis risk on farmers and false‐positive payouts on insurers. Incorporating rainfall intensity and frequency in the design of rain index insurance can offer farmers better protection, while maintaining the attractive features of the weather index insurance and thus fulfilling its promise of financial resilience.