This paper investigates the international spillover effects of nonverbal communication by central banks, focusing on the facial expressions of Federal Reserve Chairs during Federal Open Market Committee press conferences. By applying machine learning techniques to video data from 70 press conferences spanning 2011 to 2023, we construct a facial sentiment index to assess its impact on equity returns in 50 economies. We find that more positive facial sentiment from Fed Chairs significantly boosts international stock markets, with stronger effects observed in economies with higher financial openness, higher cultural similarity with the U.S., non-English official language, or real-time access to the press conference. Channel analyses suggest that these effects are driven by lower risk premium, reduced uncertainty about future monetary policy changes, and improved expectations of future economic fundamentals. Our findings highlight the substantial global impact of nonverbal cues in central bank communication.

zlynne@ceibs.edu