Este trabalho busca comparar o poder preditivo de diferentes variáveis nasrecessões datadas para a economia brasileira no período de 1997 a 2017. O modeloeconométrico utilizado é um probit simples que fornece a probabilidade de a economia brasileira entrar em recessão de um a oito trimestres à frente. Com isso, é possívelconstruir o pseudo R2, o qual permite comparar a performance de diferentes variáveispara cada horizonte de previsão. Dentre os resultados encontrados, destacam-se, paraprevisões de curto prazo, a utilização de variáveis que captam a situação fiscal do país,o índice de preços internacional das commodities e variáveis financeiras, como o retorno do índice Ibovespa e o risco-país. Entretanto, o trabalho identifica um gap naliteratura referente a variáveis que possam prever recessões para horizontes de tempomais distantes.
This paper studies the Brazilian business cycles in the period of 1900 to 2012. Since the quarterly series of the real GDP only starts in 1980 we had to build the series for the period of 1900 to 1979, using the structural time series model with temporal disaggregation for the first period. After that, a Markov Switching Model is estimated in order to build a chronology of the business cycles. The chosen model has two separate regimes with different regimes for expansion and recession, and the dating carried out in this paper is compared with other studies on the theme, and characterizations for the growth phases that can support studies of the economic history in Brazil are proposed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.