Comovimentos de política monetária e ciclos de negócios: uma abordagem através de modelos de fatores

• Ano de defesa: 2018
• Autor(a) principal: Campos, Rodolfo Herald da Costa
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Modelos de Fatores Dinâmicos; Comovimentos; Taxas de Juros de Curto Prazo; Produção Industrial; Sincronização de Ciclos de Negócios
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/30239

The present thesis is composed of three essays that seek to investigate international monetary policy comovements and common fluctuations and synchronization of business cycles in the industrial production of the Brazilian states. The main objective of the first essay is to characterize the common fluctuations between short-term interest rates from a sample of 28 countries plus the interest rate of the Euro Area in the period between 1996 and 2015. Dynamic factor models following the structure proposed by Kose et al (2003) and Neely and Rapach (2011) are adjusted where the path of estimated latent factors, which represent the global and regional common fluctuations between countries' interest rates, are capable to capture the main economic events that occurred during the analyzed period. The analysis of variance decomposition from the estimated factors suggests that the estimated common factors of the model are able to explain, on average, a share of 28% of the total variance in the first differences of the interest rates of the countries considered in the period under analysis. Linear and nonlinear Taylor’s Rule including latent factors as proxy for monetary policy movements are estimated and suggest that common global and regional fluctuations between countries' interest rates have proved to be relevant to the Bacen monetary policy in period analyzed. The second essay seeks to study the common fluctuations in industrial production in 13 Brazilian states between 2003 and 2016. Dynamic factor models are also employed in the analysis, where the path of the national common factor can be interpreted as a coincident indicator for fluctuations in Brazilian industrial activity. In addition, a variance decomposition analysis also reveals that, on average, 51% of the variability in the growth rates of the industrial production of the states is explained by the national common factor. Finally, the DMA method proposed by Raftery et al (2010) is used to identify which economic variables explain the dynamics of the national common factor and suggests that variables related to the labor market in the industrial sector, oil prices, monetary policy, supply credit and exchange rate can be considered as potential determinants of fluctuations in industrial production in the period under analysis. The third essay tries to measure the degree of synchronization between the industrial cycles of Brazilian states considering the methodology proposed by Leiva-Leon (2016) and using the same database of the second test. Results from the synchronization tests suggest important heterogeneities regarding the degree of bilateral synchronization of the industrial cycles between the considered states, even between states of the same geographic region. Dynamic panel models in which the dependent variable is the measure of bilateral dissimilarity are estimated and suggest that the differences between the degree of sectorial specialization is an important determinant of the dissimilarities between the industrial business cycles of the states analyzed. Differences between state expenditure structures are also important in explaining dissimilarities in business cycles of the states.