SAMBA: stochastic analytical model with a Bayesian approach

Resumo

We develop and estimate a DSGE model for the Brazilian economy as part of the macroe-conomic modeling framework of the Central Bank of Brazil. The model combines the building blocks of standard DSGE models (e.g., price and wage rigidities and adjustment costs) with the following features that better describe the Brazilian economy: (i) a fiscal authority pursuing an explicit target for the primary surplus; (ii) administered or regulated prices as part of the consumer price index; (iii) external nance for imports,which amplies the eects of changes in external nancial conditions on the economy;and (iv) imported goods used in the production of intermediate goods. It also includes the presence of nancially constrained households. We estimate the model with Bayesian techniques, using data since 1999, when in ation targeting was implemented in Brazil.Evaluation based on impulse response functions, moment conditions, variance error decomposition, and initial forecasting exercises suggests that the model can be a useful toolfor policy analysis and forecasting.

Ficha do documento

Tipo

Artigo científico

Ano

2015

Instituição

Sociedade Brasileira de Econometria

Fonte

Repositório da FGV

Idioma

Inglês

Acesso

Acesso aberto

DOI

10.12660/bre.v35n22015.57573

Identificador

oai:repositorio.fgv.br:10438/26218

Temas

EconomiaFinanças Públicas

Palavras-chave

DSGE modelBayesian estimationInflation targetingBrazilEconomiaEquilíbrio econômico - Modelos matemáticosAnálise estocásticaInflação - Brasil