Algumas evidências de não linearidade e caos determinístico nos ciclos econômicos brasileiros entre 1975-2009

• Ano de defesa: 2010
• Autor(a) principal: Lopes, Luckas Sabioni
• Orientador(a): Toyoshima, Sílvia Harumi
• Banca de defesa: Faria, Mercio Botelho , Bueno, Newton Paulo , Carvalho, Luciano Dias de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Economia
• Departamento: Desenvolvimento econômico e Políticas públicas
• País: BR
• Palavras-chave em Português: Caos; Não-linearidade; Séries macroeconômicas; Brasil
• Palavras-chave em Inglês: Chaos; Nonlinearity; Macroeconomic series; Brazil
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::CRESCIMENTO, FLUTUACOES E PLANEJAMENTO ECONOMICO::CRESCIMENTO E DESENVOLVIMENTO ECONOMICO
• Link de acesso: http://locus.ufv.br/handle/123456789/3251

This research had as main objective to verify the possible occurrence of chaos in economic series of Brazilian industrial output and production of capital goods, considered proxies for the total production and investment, respectively, between 1975 and 2009. In this period, Brazil experienced large swings in such aggregates, especially between the years 1980 to mid-1990s, mainly due to economic crisis generated by failures in control of inflation that reached levels increasing over the years. Such instability was transmitted by all sectors of the economy, shattering expectations and forecasting ability of economic agents, entering, thereby, high uncertainties in the system. The tests, based on filtering of the type ARMA - GARCH-M on the series of cyclic aggregates mentioned to then analyze the standardized residuals of the models estimated by BDS statistics indicated sensitive nonlinearities in both cases, however, only the proxy variable for investment can detect some level of deterministic chaos to the 5% level of significance in two-tailed test, is the assumption of normal distribution, is the analysis of empirical distributions generated by bootstrap simulations. It follows therefore that it is possible that between 1975 and 2009 have been periods in which they operated in the chaotic dynamics in the Brazilian investment, which suggests policies that serve to reduce the instability of this variable in the long term and ensure sustained growth in future ie, measures more structural than cyclical.