A BVAR Analysis on Channels of Monetary Policy Transmission in Brazil

This article measures the responses of GDP and inflation to a positive shock of the variables that make up the channels of transmission of monetary policy. The results of impulse-response functions of the estimated Bayesian VAR (BVAR) were- an increase in the short-term interest rate (SELIC) leads to a long-term interest rate increasing and consequently a reduction in GDP. Free credit does not have a significant impact on Brazilian GDP, given the low free credit/GDP ratio (Bogdanski et al., 2000). A shock in inflation expectations result in a decreasing trajectory of GDP, a fact consistent with the Fisher effect (Mishkin, 2009); and a shock at SELIC reduces inflation in the first two months, there is no “price puzzle”. A credit shock does not cause significant pressures on inflation. The Inflation does not show a well-defined time path after a shock in asset prices. The decomposition of the variance of the forecast error, in turn, showed that- GDP, in the short term, has its forecast errors explained by its own shocks, 70% on average. However, in the medium term, their forecast errors are explained by their own shocks, around 35%, by inflation shocks, 34%, and by interest rate shocks, 20%. The other transmission channels do not have, in the short and medium terms, significant influence on GDP forecast errors, except the asset prices; and inflation forecast errors are explained, in the short term, mainly by their own shocks, 85% on average. In the medium term, inflation forecast errors are explained 68% by inflation itself, 6% by GDP and the others transmission channels participate individually, with approximately 6%. These results are robust when controlled for commodity prices.