ECBs communicationa machine learning approach

Resumo

Throughout the years, central banks have become more transparent in terms of monetary policy making. It has been shown in recent years that anticipating hikes and cuts through forward guidance can have a huge impact in market expectations making it easier for policy makers to achieve their inflation targets. In order to increase their transparency, central banks have been using a plethora of communication channels, such as monetary policy statements and speeches, which give insights about their thoughts and present their views on future monetary policy. Even though communication has become clearer, it is important to note that a committee is composed of many members, each with his/her personal point of view regarding monetary policy. Being able to better understand the drivers for each member and for the committee as a whole is really important when trying to understand their decisions. In this paper, we will delve into this matter by extracting information from committee members speeches using two different approaches: (i) topic analysis, to identify important subjects through time, and (ii) sentiment analysis, to extract hawk and dove sentiment. As a basis for our work, we will use the ECB (European Central Bank) which, unlike other central banks, is responsible for overseeing multiple countries, each with its own reality, making it a rather interesting study case.

Ficha do documento

Tipo

Dissertação

Ano

2023

Instituição

Fundação Getulio Vargas

Fonte

Repositório da FGV

Idioma

Inglês

Acesso

Acesso aberto

Identificador

oai:repositorio.fgv.br:10438/34807

Temas

EconomiaGovernança

Palavras-chave

NLPECBMachine learningLanguage modelsEconomiaPolítica monetária – Europa – Estudo de casosBancos centraisPrevisão econômicaEuropean Central Bank