Best Paper Award für Maximilian Düsterhöft

Für Ihren Beitrag "Effects of central bank communication on monetary policy transmissions - A computational linguistics approach" wurden Nora Lamersdorf (Goethe-Universität Frankfurt) und Maximilian Düsterhöft (TU Dresden, Professur für VWL, insb. Managerial Economics) mit dem Best Paper Award (2^nd Place) der "1^st Conference on Behavioral Research in Finance, Governance, and Accounting" ausgezeichnet.

Der Preis wurde am 8. Oktober durch das Organisationskomitee der Konferenz um um Frau Prof. Dr. Christina E. Bannier und Herrn Prof. Dr. Peter Tilmann verliehen.

Rückblick zur Veranstaltung:
http://www.uni-giessen.de/fbz/zentren/ggs/forschung/sektionen/bsfa/Conf/past

Abstract:
Does disagreement among policy makers reduce the effectiveness of monetary policy? Focusing on 1746 speeches of FOMC members with a voting right, the present paper attempts to answer this question. Using methods from machine learning and computational linguistics, we select relevant speeches and analyze the tone conveyed both in speeches as a whole and in specific topics, using sentiment analysis combined with statistical topic modeling. We use the obtained tone measures to define periods of high and low tone divergence in US central bank communication and compare the reactions of US government bonds and its expectations and term premium components to monetary policy shocks across high and low divergence states. While we find already significant differences across these states when we use speeches as a whole to distinguish high and low tone divergence, the results become even more pronounced when we focus on single topics. We find that when the communication regarding specific topics becomes more heterogeneous, it seems that monetary policy indeed loses its effectiveness, measured as its ability to steer intermediate and long-term interest rates.