Kolmapäev, 3. märts 2021 kell 13.00 Join: Teams
LUCA ALFIERI (Tartu Ülikool), kaasautor Diana Gabrielyan
The Communication Reaction Function of the European Central Bank. An Analysis using Topic Model Indices
In this paper we analyse the communication reaction function of the European Central Bank (ECB) through indices built using topic modelling derived by the speeches of the central bank. These indices are used as dependent variables in models supported by recent literature on policy and communication reaction functions. Topics are extracted using Latent Dirichlet Allocation (LDA), one of the most exploited text mining algorithms. While, the ECB is currently reviewing its monetary policy strategy, scholars are incorporating new methods offered by modern text analysis to investigate the policy reaction function of the bank. Even if the communication reaction function is connected with the policy reaction function, studies on the former are still limited. We show how indices built through topic modelling can be used to study the communication reaction function of a central bank and analyze what variables are significant for every topic communicated by the ECB.