I propose a Bayesian quantile VAR to identify and assess the impact of uncertainty and certainty shocks, unifying Bloom's (2009) two identification steps into one. I find that an uncertainty shock widens the conditional distribution of future real economic activity growth, in line with a risk shock. Conversely, a certainty shock (a shock strongly decreasing uncertainty) narrows the conditional distribution of future real activity growth. In addition to the difference in signs, I show that the two shocks are different shocks. Each shock impacts the real economy uniquely. I support this with the underlying events: For instance, uncertainty shocks relate to events such as Black Monday and 9/11, but also to fears about future negative economic outcomes. In contrast, certainty shocks often link to phases of irrational exuberance. Commonly, no distinction is made between uncertainty and certainty shocks. I show that uncertainty shocks become more important if distinguished from certainty shocks.
Bayesian quantile VAR uncertainty shocks tail risks irrational exuberance