Dynamic impact of macroeconomic news on long-term inflation expectations

Central banks increasingly explain their decisions with the need to keep inflation expectations well anchored. While there is a general agreement that well-anchored long-term inflation expectations should not respond to short-term oriented macroeconomic news, it is not obvious how to implement this anchoring criterion empirically. The predominant part of the literature regresses long-term inflation expectations on the surprise component contained in macroeconomic news announcements (MNAs). A significant reaction of expectations to MNA surprises implies de-anchored expectations and a lack of central bank credibility. This part of the literature restricts the attention to the instantaneous response of inflation expectations to MNA surprises. Thus, it cannot account for the complex dynamics of inflation expectations. In particular, if the effect of news on expectations is very persistent, the de-anchoring problem might be more severe than the immediate response of expectations seems to suggest.

A second strand of the literature builds on the fact that short-term inflation expectations react much more strongly to macro news than longer-term expectations. Therefore, spill-overs from short– to long-term expectations are seen as a sign of de-anchoring. Although these contributions use short-term inflation expectations as a variable summarizing macro news shocks, they do not account for the information content of observable MNA surprises. However, a weak link of estimated shocks with observable information, like MNA surprises, severely undermines the credibility of the identification strategy and, thus, of the whole empirical analysis. This paper tries to bridge the gap between both strands of the empirical literature.

To that aim, we use MNA surprises, such as consumer price index and consumer confidence reports reports, as proxy variables to identify unobservable macro news shocks in a structural vector autoregression (SVAR) framework. Using a proxy SVAR allows us to investigate the dynamic impact of macro news shocks on U.S. inflation expectations, where the identified shocks correlate with observable information on short-run economic developments by construction. We use Federal Reserve Board data from July 2009 to August 2016.

Our empirical results indicate a non-negligible degree of short-run de-anchoring of U.S. inflation expectations. Macro news shocks have a significant impact on long-term inflation expectations for up to 50 trading days. The response, however, fades out eventually, confirming that expectations are anchored in the longer run.