Assessing Brazilian macroeconomic dynamics using a Markov-switching DSGE model

Abstract

ResumoO objetivo deste trabalho é avaliar o comportamento dos principais parâmetros da economia brasileira através da estimação de um modelo DSGE (Dynamic Stochastic General Equilibrium) de economia aberta usando métodos bayesianos e permitindo mudanças de regime markovianas de determinados parâmetros. Utilizando o modelo DSGE desenvolvido por Justiniano e Preston (2010) e o método de solução do modelo Markov Switching DSGE (MS-DSGE) proposto por Farmer et al. (2008), este trabalho encontrou superioridade nos ajustes dos dados dos modelos que incorporaram mudanças markovianas, rejeitando a hipótese de parâmetros constantes em modelos DSGE para a economia brasileira. Palavras-chave: Modelo DSGE. Markov Switching. MS-DSGE. AbstractThe goal of this paper is to evaluate the behavior of the main parameters of the Brazilian economy through the estimation of an open-economy dynamic stochastic general equilibrium (DSGE) model using Bayesian methods and allowing for Markov switching of certain parameters. Using the DSGE model developed by Justiniano & Preston (2010) and the solution method of the Markov switching DSGE (MS-DSGE) model proposed by Farmer et al. (2008), this paper found a superior fit in the data of Markov switching models, rejecting the hypothesis of constant parameters in DSGE models for the Brazilian economy.

Exchange Rate Pass-Through in Brazil: А Markov Switching DSGE Estimation for the Inflation Targeting Period

“…The response of inflation to the output gap is estimated at λ = 0.0722, which is in line with Bayesian estimations of more complex new Keynesian models such as Gonçalves et al (2016) who obtain λ = 0.0654. Our result, however, is not statistically significant.…”

“…A positive value indicates that the Central Bank responds to output gap deviation. Our estimate is smaller than those of Palma and Portugal (2014) and Gonçalves et al (2016) possibly due to differences in the time series used to measure interest rates.…”

“…With that in mind, we find that the IS curve demonstrates a strong response by output to the interest rate 'in excess' , with parameter ϕ = 0.64740. Our estimated parameter is much higher than the calibration of Baele et al (2015), of ϕ = 0.1, or the estimation of Gonçalves et al (2016) who obtain ϕ = 0.4063, as both of these works use only the nominal interest rate as input to their models. Our findings indicate that aggregate demand reacts strongly to the interest rate 'in excess' of the long-term real rate, which, in turn, suggests an efficient channel for monetary policy transmission in Brazil.…”

“…Nearly all the estimated DSGE models for the Brazilian economy have assumed constant parameters (da Silveira, 2008;Furlani et al, 2010;De Castro et al, 2015;Palma and Portugal, 2014). An exception is Gonçalves et al (2016), who investigated regime switching in the Taylor rule. On the other hand, a rich strand of research evaluates changes in the structural parameters of the Brazilian economy by applying conventional regime switching models such as Hamilton (1989).…”

“…One of the advantages of the representation method adopted by Baele et al (2015) is its simplicity. Liu and Mumtaz (2010) and Gonçalves et al (2016), for example, follow the more complex state-space representation and solution method of Farmer et al (2011). We follow the method proposed by Baele et al (2015), which is based on Farmer et al (2011) and Cho (2014), to characterise the stability and determinacy of the rational expectations equilibrium of the model.…”