The informative value of central banks talks: a topic model application to sentiment analysis

Abstract

Nowadays communication is acknowledged as a central bank tool to guide markets expectations. Speeches vary in topics, which are not discriminated ex-ante by text analysis. In this paper we develop a topic-weighted central bank sentiment index as a combination of machine learning and text analysis techniques to investigate large datasets. First, we grid search the best Latent Dirichlet Allocation (LDA) model to uncover the latent topics in central banks' speeches and releases of the last twenty years. Then, we build a topic-weighted Central Bank Sentiment Index (CBSIw) for the Bank of Canada (BoC), the Bank of England (BoE), the European Central Bank (ECB) and the Federal Reserve (Fed). We find that the CBSIw’s main common drivers are monetary policy, macroprudential policy and payments and settlements. We also detect topics related to new challenges, for example innovation and climate change. Moreover, we find that the CBSIw decreases after the Great Recession as well as during the COVID-19 crisis. Finally, we employ a probit regression to further assess the predictive power of our monetary policy topic-specific index. We find that the indicator helps predicting future changes in policy rate, corroborating the evidence that central banks communication signals future monetary policy decisions.