This study examined the mechanism of prices from major rice producing states in Brazil (Rio Grande do Sul, Mato Grosso and Santa Catarina), as endogenous variables, and volume of public stocks in Brazil, as exogenous variables for the period July 2004 until December 2010. The analysis includes the use of VAR-VEC methodology to identify the number of lags used and check for co-integration among variables through the test of Johansen Co-integration. We tested the existence of causality between variables by the method of Granger causality. We applied the variance decomposition of forecast error and impulse-response function with Cholesky decomposition to analyze the relationship between prices and volume of the states in public stocks. With the proposed methodology, one can observe that the price of rice in the state of Rio Grande do Sul to the price in the state of Santa Catarina and the volume stored in stockpiles and is influenced by the price of rice in the state of Mato Grosso.
price of rice; stocks public autoregressive models with error correction; Granger causality
