Abstract
Recently, the number of studies about economic uncertainty has increased, in part due to the new techniques that allow the construction of appropriate proxies for uncertainty, fundamentally unobservable, specially the web-scrapping technique that allows to extract updated online information, and which has been frequently used in the construction of these indicators. Based on two of this uncertainty indicators, we investigate the dynamics and transition of uncertainty in Brazil using quantitative autoregressive representations. The results reveal asymmetric dynamic along different conditional quantiles, corroborated by the analysis of dispersion, amplitude and densities. Furthermore, it is suggested that there is a low or even null probability of migration from a high uncertainty condition to a low level and vice versa.
Keywords
Uncertainty; Quantile regression; Asymmetry