Ensaios em macroeconometria e finanças

• Ano de defesa: 2008
• Autor(a) principal: Gutierrez, Carlos Enrique Carrasco
• Orientador(a): Issler, João Victor
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Link de acesso: https://hdl.handle.net/10438/2719

This thesis is composed of four essays referent to the subjects of macroeconometrics and nance. In each essay, which corresponds to one chapter, the objective is to investigate and analyze advanced econometric techniques, applied to relevant macroeconomic and nancial questions. The first chapter studied the lag length selection in VAR models with common cyclical features as introduced in the early work of Engle and Kozicki (1993). In particular, the paper extends previous results by Vahid and Issler (2001) to the case of VAR models having the weak form of serial correlation common feature (WF henceforth) proposed by Hecq et al. (2006). The second chapter presents an empirical application of common features. The aim of this work is to analyze the business cycles of the Mercosurís member countries in order to investigate their degree of synchronization. The model estimation uses the Beveridge-Nelson-Stock-Watson multivariate trend-cycle decomposition, taking into account the presence of common features such as common trend and common cycle. In the third chapter is proposed a methodology to compare di§erent stochastic discount factor (SDF) proxies based on relevant market information. The starting point is the work of Fama and French, which evidenced that the asset returns of the U.S. economy could be explained by relative factors linked to characteristics of the firms. An empirical application of our setup is also provided. Finally, the last chapter presents the study of estimation and testing of di§erent representative classes of consumption-based asset pricing models (CCAPM). We use the factor model to construct a small set of portfolios and the generalized method of moments (GMM) to identify and estimate the parameters of the utility function. The main result is obtained by the CRRA preference. We show that when the factor portfolio is used, the implication of this model is not rejected.