Finding Inflation Uncertainty Factors: A Sparse Stochastic Volatility Approach

Abstract Inflation uncertainty is an undeniable factor in economic decision-making. In this study, we investigate predictive factors that possess information on inflation uncertainty among many observable macroeconomic variables and uncertainty indices. To do this, we estimate several inflation prediction models popular in the literature, allowing for stochastic volatility with predetermined variables. We apply the Dirac spike-and-slab prior to the volatility-explaining variables to detect relevant macroeconomic determinants of the inflation volatility process. Contrary to prior studies suggesting that the inflation level is essentially a unique factor explaining inflation uncertainty, our findings reveal that survey inflation expectations and the capacity utilization rate are significantly more relevant. These results remain robust to different models of inflation.