Impacto das negociações algorítmicas de alta frequência no mercado futuro de dólar

• Ano de defesa: 2014
• Autor(a) principal: Pereira, Meire Midori Hori
• Orientador(a): Pereira, Pedro L. Valls
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Algoritmos de negociação; Dados de alta frequência
• Link de acesso: http://hdl.handle.net/10438/11510

This work presents a study of the impact of algorithmic tradings in the process of price discovery in the foreing exchange market. It was used high-frequency data for the U.S. Dollar Futures Contract trade in the São Paulo Stock Exchange, from January 2013 to June 2013. In order to verify if algorithmic trading strategies are more dependent than those of non-algorithmic tradings, it was examined the frequency at which algorithmic traders negotiate with each other and compared it to a benchmark model that produces theorical probabilities for different types of traders. The results obtained for these minute by minute tradings present evidence that actions and strategies for algorithmic traders appear to be less diverse and more dependent than those held by non-algorithmic traders. And it was estimated a reduced form of a high-frequency vector autoregressive (VAR) to model the interaction between serial autocorrelations of returns and algorithmic trading. The estimates show that algorithmic trading activities cause an increase in the autocorrelations of returns, indicating they may contribute to a rise in the volatility.