Joanna Jędrzejewska

Abstract

While central bank communication has evolved toward greater transparency, the nonverbal behavioral signals of policymakers remain largely unexplored. This study analyzes European Central Bank press conferences from 2010–2025 using automated facial analysis on 10,748,094 video frames to extract facial action units, blink rates, and eye aperture measures. The objective is to assess whether these cues reveal real-time cognitive processing and underlying uncertainty. Each conference is split into scripted policy statements and unscripted Q&A, enabling phase-specific analysis. Behavioral metrics are linked to macro-financial indicators, including GDP growth, unemployment, inflation, the Composite Indicator of Systemic Stress, exchange rates, FX correlations, sovereign spreads, and consumer expectations. Results show that facial and ocular patterns systematically co-move with macroeconomic conditions and financial stress, indicating that nonverbal signals carry economically meaningful information. Relationships are more complex during Q&A, reflecting the demands of spontaneous interaction and emotional regulation. The study advances understanding of monetary policy communication, suggesting that incorporating nonverbal behavior can improve interpretations of central bank transparency, policy transmission, and the behavioral underpinnings of financial markets.