Transmissão de preços do trigo entre países do Mercosul e Estados Unidos no período de 1995-2005

• Ano de defesa: 2008
• Autor(a) principal: Lopes, Taize de Andrade Machado
• Orientador(a): Freitas, Clailton Ataídes de
• Banca de defesa: Alves, Tiago Wickstrom , Feistel, Paulo Ricardo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria
• Programa de Pós-Graduação: Programa de Pós-Graduação em Integração Latino-Americana
• Departamento: Direito
• País: BR
• Palavras-chave em Português: Trigo; Integração de preços; Brasil; Argentina; Uruguai; Estados Unidos
• Palavras-chave em Inglês: Wheat; Prices integration; Brazil; Uruguay; United States
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::DIREITO
• Link de acesso: http://repositorio.ufsm.br/handle/1/9698

The aim of this dissertation is at answering the question if there was price transmission to the wheat market at Argentina, Brazil, Uruguay and the United States, or more specifically, if these markets are spatially integrated. To verify if existed price co-integration among the wheat markets from Argentina, Brazil, Uruguay and the North American market it was used the DF test, DFA, Johansen and Granger test. The results demonstrate that the wheat prices in the United States effectively participate of the long-term equilibrium, meaning that Argentina and Uruguay not have significant for the formation of prices in Brazil. The function impulse-response demonstrated that after a price chock, the wheat prices in Brazil takes nearly 13 months to get back to an equilibrium state, that is, the instabilities are corrected slowly. The decomposition of the error variance indicates that the prevision errors are expressively explained by the wheat prices in the United States. In regarding to the shortterm relations, the first difference in wheat prices to Brazil and Argentina have a significant coefficient at a 95% confidence level, that is, it responds positively to the chokes. The results for the United States and Uruguay did not demonstrate to be significant in a 5% significance level. The coefficient obtained from the error significant term demonstrates that the discrepancy of about 39% among the explaining variables and the dependent variable it is being corrected from the prior period to the actual period each month.