Comment

This is an excellent paper that reexamines the effect of monetary policy surprises on macroeconomic activity using high-frequency methods of identification. Let me say up front that the paper is well motivated, extremely clear in its findings, and a pleasure to read. There are two main innovations in this paper relative to the existing literature. First, following others in the literature, the paper shows that high-frequency monetary policy surprises are predictable based on lagged information contained in asset prices. Controlling for this predictability provides much sharper estimates of the effect of monetary policy on macroeconomic outcomes. Second, the paper incorporates information gathered in Swanson and Jayawickrema (2021) to construct monetary policy surprises based on communications by the Federal Reserve chair that occur outside of the usual Federal OpenMarket Committee (FOMC) announcements. This greatly expands the sample available fromwhich to construct measures of monetary policy surprises and further increases accuracy. The paper also extends the modeling framework considered in the authors’ earlier work to motivate why monetary surprises may be predictable. In contrast tomodels that emphasize a “Fed information effect” regarding the state of the economy, their framework highlights the possibility that the Fed conveys information about its own policy rule. The model in the paper is stylized but conveys two essential points. First, in an environment where the private sector learns about the monetary policy rule, high-frequency monetary surprises are predictable and impulse response estimates will be biased to the extent that there