Analyzing Brazilian markets using the Global VAR & IIS Approach

• Autor(a) principal: Barbosa, Bruno Tebaldi de Queiroz
• Data de Publicação: 2022
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Repositório Institucional do FGV (FGV Repositório Digital)
• Texto Completo: https://hdl.handle.net/10438/32256
• topic: Macroeconomic; Global VAR; GVAR; Macroeconomia; Economia; Economia - Brasil; Modelos econométricos; Macroeconomia; Mercado de trabalho - Brasil

This thesis consists of three chapters. The chapters can be read separately, ie there is no predetermined reading order. However, the suggested order follows a linear evolution of the theme. This work expands the work of Barbosa (2017), in his study the author establishes a model for the Brazilian market taking into account the interdependencies between regions using the Global VAR (GVAR) methodology, and uses this model to estimate the elasticity of regional employment in relation to the country’s economic activity. In this study we expand the Barbosa (2017) model on several fronts. First, the study addresses the use of different weight matrices. Traditionally, the weight matrix used in the literature is based on trade weights and bilateral trade between two countries. Barbosa (2017) proposes a weight matrix based on connections between regions, this study in its turn expands these weight matrix allowing the weight matrix to be based not only on connections between regions but also on macroeconomic variables of each region such as GDP, GDP per capita and population. A second innovation is made with the proposal of a new econometric model. This new econometric model is built from the Global VAR model, which is expanded through a saturation with impulse indicators (henceforth called GVAR-IIS). It is worth mentioning that the hypothesis of weak exogeneity remains a requirement for the validity of the GVAR-IIS. For validation of weak exogeneity, the study applies not only the classical tests of weak exogeneity proposed by Granger and Lin (1995) but expands the tests by applying the concept of separability proposed by Hecq et al. (2002). The first part of the study analyzes the original model proposed by Barbosa (2017) using a weight matrix based on connections between cities. The database is expanded to include the period of forecasts that took place between 2016 and 2018. In this chapter we assess how the forecasts behave in the face of the scenario, we also assess the resilience of the regions and the heterogeneity of the responses. In the second part, the GVAR model proposed by Barbosa (2017) is saturated with impulse indicators, this new model is referred as GVAR-IIS. The chapter proposes an estimation procedure for the GVARIIS. The Chapter also presents an empirical exercise in which the new GVAR-IIS model is evaluated together with other models to assess its predictive power. To validate the hypotheses of weak exogeneity, the classical tests proposed by Granger and Lin (1995) are carried out, in which an innovation is the use of the concept of separability proposed by Hecq et al. (2002) to validate the hypothesis of weak exogeneity. Finally, in the third part, the GVAR-IIS model developed in the previous chapter is used together with a weight.