Avaliação da eficácia de sinais da análise técnica no mercado de capitais brasileiro, no período de 2000 a 2010

• Ano de defesa: 2012
• Autor(a) principal: Petrokas, Leandro Augusto
• Orientador(a): Famá, Rubens
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica de São Paulo
• Programa de Pós-Graduação: Programa de Estudos Pós-Graduados em Administração
• Departamento: Faculdade de Economia, Administração, Contábeis e Atuariais
• País: BR
• Palavras-chave em Português: Análise técnica; Bolsa de valores; Eficiência de mercado
• Palavras-chave em Inglês: Technical analysis; Stock market; Market efficiency
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO
• Link de acesso: https://tede2.pucsp.br/handle/handle/1025

This study aims at assessing the technical analysis is capable of producing superior financial results to the model buy and hold, which recommends the purchase and sale of shares in the medium or long term, without the use of specific tools or criteria to guide such decision. The sample was composed of fourteen shares, in eight different sectors of the economy, the survey period was from 2000 to 2010 and the tests were conducted with five key indicators of technical analysis. It was decided to test whether the simple moving average of 233 periods could increase the profitability obtained by the signs. The impact of generating a buy signal in the returns of five actions, through the event study methodology and, finally, estimated the average duration of the operations performed by the signals studied. The results indicated that in a case 350, 61 only in the yield obtained with the signals of the technical analysis was superior to that obtained by the model buy and hold. By means of the chi-square, it was found that this frequency is not statistically equal to half the cases, therefore, conclude that technical analysis was not able to produce superior financial results to the buy and hold. The results showed that there was no significant improvement in profitability with the use of simple moving average of 233 days as a filter rule, therefore, was not statistically significant differences in mean returns obtained by each signal, with and without the filter. The event studies revealed that only one of the five events analyzed was a statistically significant and positive impact on stock returns. The analysis of the duration of the operations indicated that profitable operations have a longer duration when compared to non-profit for all signals except the IFR, when there was an opposite behavior of this pattern