Agricultural risk management using NDVI pasture index-based insurance for livestock producers in south west Buenos Aires province

Purpose – Based on the Normalized Difference Vegetation Index (NDVI)-based insurance developed by the Ministry of Agriculture, Livestock and Fisheries of Argentina (MAGyP) with technical assistance of the World Bank (WB), the purpose of this paper is to evaluate the out-of-sample performance of the NDVI-based insurance in Bahía Blanca Department of the south west of Buenos Aires (SWBA) Province in Argentina, by calculating the technical premium with the methodology developed by MAGyP-WB and NDVI information up to 2007, and analyzing the results that would have been obtained in 2008 and 2009. Design/methodology/approach – With the available NDVI information (1982-2009), the authors uses the out-of-sample method to analyze the rating of the contract and the reasonability of the premium payment through a comparison of the frequency and severity of payouts in the NDVI coverage with the losses suffered by droughts in SWBA. Specifically, it has been taken the data until 2006 to set the Triggers and Exits values and to calculate the premium rates, and the payout and loss ratio in 2007 was then analyzed. A similar analysis was done for years 2007-2008 and 2008-2009. Findings – According with the rating methodology described in this paper, payouts determined by the NDVI-based insurance fit with falls in forage production and reduced meat production yields, which confers reasonability to this tool as a coverage option for cattle and fodder producers. Definite technical premiums based on 1982-2007 period, capture the occurrence of severe drought events, defining a risk profile that is consistent with the used information. Adding more observations to the sample, this profile is redefined, showing the sensitivity of the results to the quality and quantity of data used in the analysis. Research limitations/implications – There is a subjective assessment in the determination of the sample and the weighting of this information. The election of the period to be considered, how to incorporate the changes in the patterns of climate behavior in the medium- and long-term and the expected effects in the NDVI, etc. will impact on the values of resulting technical premiums. In the specific case of this analysis, the fact of not having considered in the sample two extreme years (2008 and 2009) has a concrete implication in the obtained premiums. In this paper, only the Bahía Blanca Department was considered. Rainfall pattern and extensive grazing of natural grassland in the SWBA area might vary from Department to Department, so the results and payouts might change according to these circumstances. Practical implications – The situation shown in the findings could be seen as an underestimation of the risk to which the producer is exposed and should be taken into account for further researches and insurance products to be designed. When the considered data series shows atypical or extreme values and these are incorporated into the sample, significant changes can be registered in the triggers and premiums. Originality/value – This paper analyzes how a risk profile is redefined depending on the sample considered and shows the sensitivity of the results to used data when using index-based insurance.