Thesis and Dissertation

SM Cuiabano (2011) Sobre a taxa de câmbio : ensaios para países de renda baixa, emergentes e para o Brasil (supervisor: Ellery Junior, Roberto de Goes)

The goal of this thesis is to analyze the macroeconomic fundamentals that balance the real and nominal exchange rate in emerging countries (Latin America and Asia) and in low income countries (East Africa). The notion of equilibrium being examined is the one from the econometrics advances applied to macroeconomics: the relationship of non-stationary variables, in the long term, can be stationary when evaluated together, generating a cointegrating equation. Once we have the cointegration equation, it is possible to capture the long-term elasticities among the exchange rate and the macroeconomic fundamentals in order to compare the current exchange rate with the estimated equilibrium exchange rate. The first paper examines the real exchange rate equilibrium for the countries of the East African Community (Burundi, Kenya, Rwanda, Tanzania and Uganda). The second article discusses the appropriateness of the monetary nominal exchange model in Brazil and tests the predictability of this model compared to a random walk. The third article provides an analysis of the monetary model in some countries of Latin America and Asia, using the idea of equilibrium to evaluate the "imbalance" between the exchange rates of the two regions. This thesis aims to bring up the importance of the macroeconomic fundamentals in the behavior of the exchange rate and also aims to contribute to the end of a dark period in international macroeconomics showing the power of forecast of the monetary model compared to a random walk in Brazil.

SM Cuiabano (2007) Determinação da taxa de câmbio: Aplicação do modelo de cagan para o Brasil. (supervisor: Divino, José Angelo Costa do Amor)

This work aims to test a variant of the monetarist exchange rate determination model in Brazil, as available in Obstfeld and Rogoff (1996). It initiated with a simple model of Cagan s demand money model and applied the hypotheses of purchase power parity and uncovered interest parity (UIP). Then we proceeded cointegration tests - Johansen and Engle-Granger - to verify the long run behavior of the variables in the Brazilian case. By the adjustment speeds given by the Johansen and the model of error-correction we verify the variables relation in a exchange rate model determination in a short term. At the presence of endogeneity in the variable, it was estimated an instrument model by the Generalized Method of Moments (GMM). The data corroborated the models signals, with exception of the signal for international interest rate which may indicate the non existence of UIP for the Brazilian case. We considered that the model can be improved by the insertion of variables from the balance of payments model and other international variables.