News Media, Common Information, and Sectoral Comovement

This paper investigates whether information complementarities can explain the strong patterns of sectoral comovement observed empirically. It tests the theoretical model by Veldkamp and Wolfers (2007), which suggests that fi rms' output decisions are based on aggregate information rather than sector-specifi c information, because the former is less costly. Employing the connectedness index by Diebold and Yilmaz (2009, 2012) as a new measure of sectoral comovement and using data on media coverage of economic news in Germany, we find that a higher volume of economy-wide news indeed signifi cantly increases the comovement of sectoral business expectations. This common shock to expectations is reflected in a delayed increase of sectoral output comovement. Although fi rms tend to be more susceptible to bad news, the tone of media coverage only plays a minor role.