Perda máxima aceitável para investimento de risco em commodity brasileira

Abstract

This study aimed to simulate the maximum acceptable loss for the risk of investing in a major agricultural commodities in Brazil, wheat. To fulfill this objective, the study used the modeling of time series with autoregressive processes moving average (ARMA), conditional heteroskedasticity (GARCH) and value at risk (V@R), applied on a historical series of asset prices in a ten year period. The investigation database was collected in the site CEPEA/ESALQ/USP and the motivation for choosing this historical series was due to the productive and economic importance of the State of Paraná throughout Brazil, the largest producer and commodity trader. The results discussed in this paper show that a level of statistical significance of 1%, wheat producer admits a loss of R$ 228.40 under the investment and 5% R$ 174.19. Overall, these data suggest that every tonne of wheat sold, the producer can lose on your investment up to R$ 228.40. These and other information covered in the survey may provide support strategic tools for decision‐making investment in commodities. The dynamics of prices expressed by the market volatility may reveal some active behavior pattern, which can be useful for the formation of hedging policies, in which various producers assume the risk of the effected investment, thus minimizing the impact of the same on producers.