Effects of monetary policy news on financial assets: Evidence from Brazil on a bivariate VAR-GARCH model (2006–17)

Abstract

The impact of news releases related to the inflation targeting regime on the financial market is analyzed by estimating a bivariate VAR GARCH-BEKK-in-mean model. We use daily data, from January 2006 to May 2017, of stock prices index (IBOVESPA), exchange rate (BRL/USD) and interbank deposit rate (DI360). We developed a positive and negative news index to measure the impact of news releases based on Caporale et al. (2016) and Caporale et al. (2018). Although the literature on the subject is vast, this paper fills relevant gaps in three ways. First, we investigate the bidirectional relationship between monetary policy related news releases and the behavior of asset prices before and after the 2008 crisis in Brazil. Second, we consider the relationship between the second moments of the variables of interest, using the conditional volatility as a proxy for uncertainty. Third, we provide a time series approach to measure the effect of macroeconomic related news releases on financial asset returns. The results indicate there are mean spread effects from news for the exchange rate and the Brazilian stock index: (i) the GARCH-in-mean parameter is statistically significant for positive and the difference of news for the DI360; (ii) monetary policy and external shocks are statiscally significant as expected with exception of the external shocks for the Brazilian stock index; and (iii) there are volatility spillovers and changes of this volatility after the crisis for stock index and DI360.