Econofísica e o estudo da complexidade em mercados financeiros

• Ano de defesa: 2005
• Autor(a) principal: Tavares, Marcus Vinicius Moldes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Programa de Pós-graduação em Física; Física
• Palavras-chave em Português: Econofísica; Mercado financeiro; Algorítmo genético; Análise de Wavelets; CNPQ::CIENCIAS EXATAS E DA TERRA::FISICA
• Link de acesso: https://app.uff.br/riuff/handle/1/17281

In the last 10 years, a new area called Econophysics came along and since then it has questioning traditional models on Economy by confronting them to real financial time series. In this work, after a short review of this new field, we claim that these conflicts are not sustainable. Therefore, we proposed to study the behavior of markets should be done with tools that are frequently used on the studies of complex and/or chaotic systems. In the search of patterns on financial time series, we developed a tool based on a genetic algorithm (GA) for further analysis by other techniques, such wavelets, as we chose to use in this work. We have found that the most regular sections of the time series display at least the common features: persistence, large amplitudes and multifractality. These regions seems to appear between sharp changes on the time series. Finally, we proposed a microscopic model for the stock market. Using an agentbased simulation following the Bak-Sneppen dynamics, we have found the efficient market hypothesis does not hold, even for a closed just-one-active market. Our model show some features of real markets such intermittencies, fat tails and long range correlations on volatility series. According our simulations, the persistent behavior frequently found on price series is a consequence of agent reactions to the market price variations.