Dictionary-based sentiment analysis of monetary policy communication: on the applicability of lexicons

Abstract

In this study, we discuss central bank communication sentiments assessed using different dictionaries. Policy communication by monetary authorities is considered an essential additional policy tool. Various text-mining techniques could evaluate its sentiment. Dictionary methods are a prominent example here because they offer simplicity and are a starting point for applying more elaborate methods. We decided to test four lexicons: generic, economic and financial and two domain-specific lexicons to conclude on their applicability to assess the sentiments of monetary releases and provide recommendations for their further application. This study surveys four lexicons qualitatively and conducts a few exercises: lexicon content comparison, performance tests for highly positive and negative messages, and statistical tests of dictionary alignment and correlation. The sample covers small open economies implementing inflation targeting. The study concludes that all investigated dictionaries could be used for detecting central bank intentions: they do not return contradictory results. The choice among them for a particular study depends on the study’s setup and goals—as presented in our recommendations