(Extreme) Weather index-based insurances: data, models, and other aspects we need to think about

Farming systems are vulnerable and exposed to the extreme weather events such as drought, frost, and heat. Insurance schemes based on weather indexes provided by weather stations or satellite data may represent an innovative and suitable solution to cope with climate risks. However, their uptake remains low due to both high premiums that farmers are not willing to pay and basis risk issues. The literature is vast and growing, calling for a reflection on pros and cons of different indexes. The weather-index based schemes have been investigated more than the indemnity insurances, with prevalent emphasis on rainfall-based, temperature-based and vegetation indices. We review the literature on index-based insurance to animate the debate on the planning of risk management policies, with a focus on the Italian situation to discuss on current data availability and future needs. The challenge is to identify an effective weather index and calibrate the damage threshold depends on the availability of data and models, as well as on the (scarce) knowledge of the critical aspects driving the weather-yields relationships, a major driver of distrust in currently offered index-based insurance schemes. A large heterogeneity is observed on the methods dimension, the regressions and correlations methods are the most widely approaches to design weather insurances. Rainfall and temperature-based indices are the most widely used as they are arguably the most important factors that affect crop yields. The debate on the identification of the best climate econometrics models to assess variation in yields due to changes in weather conditions is still vivid. Clearly, the main limitation is the availability of highly disaggregated publicly data. The understanding of these dynamics allows to identify the best index and model that can explain the relationship between weather extremes and yield losses, i.e., the working principle of weather index-based insurances.