Duration-dependent Markov-switching model: an empirical study for the Brazilian business cycle

This paper uses a duration-dependent Markov-switching model to identify business cycles in the Brazilian economy and to test for the presence of duration dependence in periods of expansion and contraction. The model is estimated using the growth rate of quarterly GDP from 1980:II to 2016:II. In the empirical application we found evidence of significant asymmetry in growth rates and duration dependence in the business cycle transition probabilities. The parameter estimates indicated that as the recession ages, the probability of a transition into an expansion increases (positive duration dependence in recessions). On the other hand, as the expansions ages, the probability of a transition into a recession decreases (negative duration dependence in expansions). The smoothed probabilities of the model captured several periods of contraction during the last three decades, matching the recession dates of the Business Cycle Dating Committee (CODACE) from the Getúlio Vargas Foundation.