An Impact Analysis of the Positions of Brazilian Presidents Regarding the Brazilian Stock Market Using Machine Learning
Keywords:
Mercado Acionário Brasileiro, Discursos Presidenciais, Política e Finanças, Processamento de Linguagem Natural, Aprendizagem de Máquina, Análise de Sentimento, Large Language Models, Regressão Linear MúltiplaAbstract
This study investigated the correlation between the speeches of presidents
Dilma Rousseff, Michel Temer, and Jair Bolsonaro and the Brazilian stock
market, using Artificial Intelligence to classify the speeches by theme and sentiment.
The research concluded that while the speeches had a limited and specific
statistical correlation with asset prices, they showed a relevant correlation with
trading volume. This effect was particularly notable during the Bolsonaro administration,
where there is evidence that speeches with negative sentiment were
strongly correlated with an increase in market activity.
References
Ran Aroussi. yfinance, 2025. Versão 0.2.65. Disponível em: https://github.com/ ranaroussi/yfinance. Acesso em: 21 jul. 2025.
B3. Índice ibovespa - ibovespa - composição da carteira, 2025. Disponível em: https://www.b3.com.br/pt_br/ market-data-e-indices/indices/indices-amplos/
indice-ibovespa-ibovespa-composicao-da-carteira.htm. Acesso em: 21 jul. 2025.
S. Baek, S. K. Mohanty, and M. Glambosky. Covid-19 and stock market volatility: An industry level analysis. Finance Research Letters, 37:101748, nov 2020.
M. Baker and J. Wurgler. Investor sentiment in the stock market. Technical report, National Bureau of Economic Research, Cambridge, MA, jun 2007. Disponível em: https://doi.org/10.3386/w13189. Acesso em: 28 maio 2024.
Brasil. Biblioteca presidência da república, 2025. Disponível em: https://www. biblioteca.presidencia.gov.br/. Acesso em: 21 jul. 2025.
Google DeepMind. Gemma, 2025. Disponível em: https://deepmind.google/ models/gemma/?hl=pt-br. Acesso em: 21 jul. 2025.
R. S. Hudson and A. Gregoriou. Calculating and comparing security returns is harder than you think: A comparison between logarithmic and simple returns. International Review of Financial Analysis, 38:151–162, mar 2015.
Yuliia Kniazieva. Machine learning in the finance industry: is it worth taking the risk?, 2023. Disponível em: https://labelyourdata.com/articles/ machine-learning-in-finance. Acesso em: 30 maio 2024.
Wassima Lakhchini, Rachid Wahabi, and Mounime El Kabbouri. Artificial intelligence & machine learning in finance: a literature review. International Journal of Accounting, Finance, Auditing, Management and Economics, 2022.
A. Maligkris. Presidential candidates, political speeches and stock market re- turns, 2017. Disponível em: https://ssrn.com/abstract=2980475. DOI: 10.2139/ssrn.2980475.
Meta. Llama 3, 2025. Disponível em: https://www.llama.com/models/
llama-3/. Acesso em: 21 jul. 2025.
H. et al. Naveed. A comprehensive overview of large language models, jun 2025.
D. Shah, H. Isah, and F. Zulkernine. Stock market analysis: A review and taxonomy of prediction techniques. International Journal of Financial Studies, 7(2):26, maio 2019.
M. G. et al. Sousa. Bert for stock market sentiment analysis. In 2019 IEEE 31st Interna- tional Conference on Tools with Artificial Intelligence (ICTAI), nov 2019. Disponível em: https://doi.org/10.1109/ictai.2019.00231. Acesso em: 30 maio 2024.
Hariom Tatsat, Sahil Puri, and Brad Lookabaugh. Machine learning and data science blueprints for finance. O’Reilly Media, 2020.
A. Wagner, R. Zeckhauser, and A. Ziegler. Company stock reactions to the 2016 elec- tion shock: Trump, taxes and trade. Technical report, National Bureau of Economic Research, Cambridge, MA, fev 2017. Disponível em: https://doi.org/10. 3386/w23152. Acesso em: 28 jul. 2025.
Yahoo. Yahoo finanças, 2025. Disponível em: https://finance.yahoo.com/. Acesso em: 21 jul. 2025.
L. X. Yuan. distilbert base multilingual cased sentiments student · hugging face, 2025. Disponível em: https://huggingface.co/lxyuan/ distilbert-base-multilingual-cased-sentiments-student. Acesso em: 21 jul. 2025.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Pedro Santos Mendonça Neto, João Gabriel Gama Vila Nova
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Authors who publish in the Annals of Data Analysis retain
the copyright of their work and agree to license it under a Creative Commons Attribution-
NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license.
