Measuring the neutral real interest rate in Brazil: a semi-structural open economy framework

This paper applies a multivariate Kalman filter to estimate the neutral real interest rate (NRIR) in Brazil, from 2002Q1 to 2017Q3. Our state-space representation combines semi-structural macroeconomic models with the Hodrick–Prescott (HP) filter approach. This framework differs from previous literature in two aspects: (1) instead of relying in the usual closed-economy case, we explore a reduced-form open economy’s set up, where the openness makes the NRIR to depend on world potential output and the openness degree of the economy, in addition to domestic potential output, the time preference rate and the relative risk aversion; (2) instead of imposing for potential output structures that follow random walks or have autoregressive patterns, our set of state equations for this variable relies on a production function. This work contributes to the literature in two ways: (1) we provide a simple and complementary empirical tool to evaluate the monetary policy stance in open economies that explicitly take into account the foreign output dynamics; (2) an empirical advantage of our model is to provide a single stage conditional estimation of four latent variables that are very useful for policy analyses: NRIR, potential output, non-accelerating inflation rate of unemployment and non-accelerating inflation rate of capacity utilization.