Análise de desempenho de portfólios de investimentos, a partir de critérios ESG, no mercado acionário Brasileiro

• Ano de defesa: 2025
• Autor(a) principal: Garcia, Sávio Cardoso
• Outros Autores: https://orcid.org/0000-0002-1412-3074
• Orientador(a): Prado, José Willer do
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Lavras; Faculdade de Ciências Sociais Aplicadas (FCSA)
• Programa de Pós-Graduação: Programa de Pós-Graduação em Administração
• País: brasil
• Palavras-chave em Português: ESG; Precificação de ativos; Modelos ARCH; Mercado financeiro; Asset pricing; ARCH models; Financial market
• Área do conhecimento CNPq: Ciências Sociais e Aplicadas
• Link de acesso: https://repositorio.ufla.br/handle/1/60369

Various investment strategies in financial assets have been proposed over time. In this context, analysts, managers, and investors adopt different approaches in portfolio formation. On the other hand, investors are increasingly aware of environmental, social, and governance factors and, consequently, take such factors into account in allocation decisions. However, academic literature still lacks more studies on the effect of ESG criteria on investment portfolios, considering, primarily, an emerging country like Brazil. Therefore, the present research aims to investigate the performance of investment portfolios formed based on ESG criteria in the Brazilian stock market. To meet this objective, three articles will be developed: The first article conducted a thematic review on portfolio theory and asset pricing in the Web of Science and Scopus databases, using tools such as R Studio, Excel, and EndNote. The second article formed portfolios with ESG, E, S, and G criteria through Markowitz's (1952) mean-variance approach and the scores established by Thomson Reuters Refinitiv ESG. It investigated whether ESG criteria can be considered a risk factor and whether traditional asset pricing models can explain the returns of ESG portfolios. Lastly, the previously established filters will form ESG, E, S, and G factors, which will be incorporated into a pricing model. Finally, in the third article, the conditional volatility of ESG, E, S, and G portfolios, as well as sustainability and governance indices and the Ibovespa in the Brazilian stock market, was modeled using ARCH and GARCH model extensions (EGARCH, TGARCH, and APARCH). In the first article, bibliometric techniques found that seminal works such as Markowitz (1952) and Sharpe (1964) are still fundamental bases for other research. In the proposed theme, there is an increasing number of studies incorporating other risk factors, including ESG criteria in pricing models. In the second article, the portfolios had higher returns and lower drawdowns than the benchmark indices. The regression models revealed that ESG and Governance (GOV) factors have explanatory power over the returns of Brazilian stocks, suggesting that investors may be pricing sustainability and governance practices. Finally, in the last article, ESG portfolios showed higher volatility persistence compared to benchmark indices and leverage effects. Macroeconomic and political factors may be linked to the amplification of volatility in the analyzed time series. This research contributes to the academic literature on sustainable finance by supporting the relationship between ESG criteria investments, return, and risk in the Brazilian stock market, raising and exploring future research trends on the proposed theme (Article 1), evaluating the return and risk of these ESG portfolios (Article 2), and finally, conducting a more detailed risk analysis using conditional volatility models (Article 3).