Curva de juros e indicadores macroeconômicos no Brasil: uma análise de domínio tempo-frequência

Abstract

In this dissertation, we sought to understand the relationship over time between the components of the Taylor rule (interest rates, inflation, and output). For the interest rate component, a yield curve was estimated using the Nelson model, and components of level, curvature, and slope were extracted. As a proxy for output, we used the IBC-Br index, with data extracted from the Central Bank of Brazil for the period from January 2004 to October 2023. The second step of the methodology involved time-frequency domain analysis, examining the relationship of variables over time with the frequency of their cycles. Through these two methodologies, it was possible to analyze the cyclical relationship between inflation and the output gap with the components of the yield curve. The results revealed that, for the most part, both the output gap and inflation precede the yield curve. Moreover, the analysis predominantly identified short-term cycles in the relationships between these variables. These findings suggest a complex dynamic and potential causal interactions among the studied economic variables, underscoring the importance of time series analysis in the financial and macroeconomic context.