Do central bank cycles drive stock returns? New evidence from the US, UK, and Japan

Abstract

Cieslak et al. (2019) show that the equity premium in the US since 1994 is earned entirely in even weeks of the Federal Open Market Committee meeting cycle and that these same even weeks also drive international stock returns. Updating their data, I find that their US result does not hold out-of-sample and show that with an extended sample, the result loses its robustness as early as 2004. As further evidence, I show that their proposed mechanism also does not hold from 2004 onwards. Examining the data prior to 2004, I show that there are important outliers that appear to be driving the result. Finally, I construct central bank cycles for the Bank of England and the Bank of Japan and show, when accounting for potential pre-announcement effects, their international result also no longer holds.