O efeito do sentimento das notícias sobre o comportamento dos preços no mercado acionário brasileiro

• Autor(a) principal: Silva, Maria Daniella De Oliveira Pereira da
• Data de Publicação: 2018
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFPB
• Texto Completo: https://repositorio.ufpb.br/jspui/handle/123456789/12575
• topic: Notícias financeiras; Sentimento textual; Mercado acionário; Previsão do risco; Financial news; Textual sentiment; Stock market; Forecast risk; CNPQ::CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO::CIENCIAS CONTABEIS

This paper aimed to investigate the effect of the textual sentiment of the financial news on asset prices in the Brazilian Stock Exchange, seeking theoretical support in the agreement theorem, in the negativity bias and in the efficient market hypothesis. In order to analyze the effect of the tone of the news on asset prices. I verified the influence the textual sentiment of the news exerted on the IBOVESPA index return and the risk (volatility of the return); also I analyzed the influence of textual sentiment on the forecast market risk. This study used to daily data of the following index: IBOVESPA, DJIA, VIX and S&P500. As for the texts of financial columns from the newspaper Valor Econômico (45.304 articles). I analyzed the period of July 25, 2011 to June 30, 2017, corresponding to 1.470 daily observations. The choice of predictors of volatility was performed by l-penalized (LASSO) quantile regression method and the impact of the sentiment on the market was evaluated by quantile autoregression. As for the main results, this study shows that the sentiment of media pessimism predict reductions in the return of IBOVESPA of -28.72 basis points and optimism sentiment predict increase of 22 basis points on the market return when low performances market (low return). As for the volatility, I observed that media pessimism corroborates with the reduction of volatility. In the periods of high uncertainty, the sentiment of pessimism led to reductions in returns, in the circumstances of high or low market performance, predicting -76.21 basis points on the low performance (q.05), and reductions of -81.94 points (q.95) on the high performance. Regarding the forecast of volatility, the results that the textual sentiment is a predictor of volatility, which contributes to minimizes the forecast errors of volatility. Therefore, we can conclude that the information coming from newspapers in Brazil affect the perception of investors in moments of greater uncertainty in the market and in the economy. The work has contributed to the discussions about the role of the media in the stock market of emerging countries, raising evidence that textual sentiment of news can provide information for risk management.