Banca de DEFESA: Alexandre Henrique Lucchetti

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
STUDENT : Alexandre Henrique Lucchetti
DATE: 07/03/2025
TIME: 09:00
LOCAL: Vídeo Conferência
TITLE:

Natural Language Processing in Economics and Finance: Literature Review and Applications to Monetary Policy Communication

KEY WORDS:

Natural Language Processing; Monetary Policy Communication; Monetary Policy Coordination; Yield Curve; Sentiment Analysis.

PAGES: 163
BIG AREA: Ciências Sociais Aplicadas
AREA: Economia
SUMMARY:

This work consists of three papers on Natural Language Processing (NLP) applied to economics and finance. The first paper surveys the main methods, requisites and applications of NLP in the field. Structured to help shorten the path economic researchers have to trail to get introduced to these methods, our work covers the following topics: (i) useful NLP tasks to economics; (ii) NLP models; (iii) economic and financial textual data for NLP; and (iv) NLP applications in economics and finance. We further contribute by resorting to bibliometric tools to help us visualize the literature map of this field, also providing valuable insights to our survey and the remainder of our work. We finally indicate that there is much room to apply natural language processing to economic issues, but alert that, more than ever, researchers must be careful not to stray away from questions motivated by hypotheses closely tied to economic theories. The second paper provides a forward-looking measure of how central bankers implicitly coordinate their actions, as measured from public manifestations in the form of their speeches’ transcripts. In order to do that, we resort to the the central bankers’ speeches database made available by the Bank for International Settlements and build a network of similarities that connects central banks, adapting for this context the method proposed by Cajueiro et al. (2021). Our results show that our network successfully captures the long-term global importance of central banks overseeing the G10 currencies, with word-level point to evidence of their orthodox approach to monetary policy. We further explore this framework on a dynamic setting, with findings that indicate that coordination tends to increase in times of economic stress, as in the years of the Great Financial Crisis and the period after the Covid-19 pandemic. An evolution analysis on word occurrences then shows that our proposed measure is driven by mentions to policy instruments and economic views proffered by policymakers. The third paper proposes a framework for estimating expectation-embedded multi-dimensional sentiment from monetary policy communication, combining economic fundamentals and state-of-the-art deep learning neural networks. The economic basis of our approach is set by the Litterman and Scheinkman (1991) yield curve decomposition model, with its level, slope and curvature factors accounting for the three dimensions of our sentiment gauge. For text modeling we incorporate in our framework the Bidirectional Encoder Representations from Transformers, BERT (Devlin et al., 2019). In an application to policy communication by the Brazilian Central Bank, our results reinforce the need for more than one sentiment dimension to comprehensively capture the relevant nuances of communication for monetary policy assessment. These results indicate that the added dimensions complement the usual hawk/dove gauge of policy sentiment, and can at times be more relevant in setting the overall tone of policy communication, eventually even preceding shifts in monetary policy stance.

COMMITTEE MEMBERS:
Externo à Instituição - Carlos Alexandre Piccioni - BACEN
Presidente - 1642911 - DANIEL OLIVEIRA CAJUEIRO
Externo à Instituição - JOSE ANGELO COSTA DO AMOR DIVINO - UCB
Interno - 1550794 - JOSE GUILHERME DE LARA RESENDE
Interno - 1229140 - ROBERTO DE GOES ELLERY JUNIOR