| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 363 KB | Adobe PDF |
Orientador(es)
Resumo(s)
This paper assesses the impact of Federal Reserve“s public communications on asset pricing, particularly post-FOMC meeting statements. Price variations and their causes are studied, using both linear and non-linear models. The sentiment score implied in Fed statements is one of the variables, measured by FinBERT, a pre-trained NLP algorithm. I managed to prove that two of the three variables tested (unexpected component and
sentiment) significantly influence market movements and that the actual change in the Federal Funds Target Rate is not a valid predictor. Increased volatility was also verified statement releases, something that was found to be heightened post-pandemic.
Descrição
Palavras-chave
Federal Reserve Machine Learning Monetary Policy Econometrics Sentiment Analysis
