Essays in Macroeconomics and Financial Econometrics

Orientador(a): Carlos Viana de Carvalho

This dissertation consists of three essays on macroeconomics and financial econometrics. In the first essay, we evaluate each category within the Personal Consumption Expenditures (PCE) basket separately to construct a cyclically sensitive inflation index. Building on Stock and Watson’s Common Activity Index (CAI), we show that our category-by-category approach – augmented with a broader set of items – yields superior out-of-sample performance, when evaluating the correlation between our measure and the CAI. We also do an exercise using regional data that provides interesting insights into the dynamics of inflation for different items. In the second essay, we investigate sectoral heterogeneity in price stickiness by deriving both sectoral and aggregate Phillips Curves from a New Keynesian framework. Using PCE data for fifteen sectors, we demonstrate that ignoring heterogeneity biases aggregate slope estimates and that our sector-level price adjustment frequencies closely align with micro-data evidence, thereby bridging the gap between micro price behavior and macroeconomic models. Finally, the third essay examines the properties of realized volatility estimators using high-frequency data on nearly 11,000 U.S. stocks from 2003 to 2020. We compare a variety of measures – addressing microstructure noise and jumps – and find that simpler estimators (e.g., five-minute realized volatility) often  outperform more complex methods for illiquid stocks, highlighting important implications for risk management and volatility forecasting.