Caracterização da distribuição e análise espacial da produtividade de milho no estado de Sergipe : efeitos para o seguro agrícola.

Abstract

Agricultural activity is subject to factors such as climatic weather, pests, diseases of difficult man control, etc. The idea of insurance in the rural context is to indemnify the producer whenever his productivity is below the productivity guaranteed by the insurer. Brazilian insurers do not have a consolidated actuarial theory, making it an impediment to premium calculations and expected payments. Therefore, this study proposed to calculate and price the agricultural insurance applying Generalized Linear Models, by grouping the data, identified by spatial clusters, statistically significant. The data set used was the municipal corn productivity, obtained from the site of the IBGE Automatic Recovery System (SIDRA), from 2003 to 2015, for the state of Sergipe. HH: High-High and LL: Low-Low clusters were identified. Both types indicate the occurrence of positive spatial autocorrelation, that is, it was observed that there is evidence that on average the corn yields in some municipalities are more significant than in other municipalities in the state, as was also verified by the coefficient of variation that has been indicative that the state of Sergipe is included in the category of high agricultural risk. In addition to being able to conclude that through these proposed methodologies have been indications that the indemnity that insurers may be paying higher values that should be collected, which consequently ends up discouraging the development of the market in the state, since the premium charged also must be above what should be charged.