Estratégias de negociação baseadas na existência de efeito de liderança e defasagem entre o índice bovespa e o índice bovespa futuro utilizando dados de alta frequência

• Autor(a) principal: Nelson Ferreira Fonseca
• Data de Publicação: 2010
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFMG
• Texto Completo: http://hdl.handle.net/1843/BUBD-9BFJK2
• topic: VECM; ARIMA; VAR; ARFIMA; Ibovespa futuro; Ibovespa; Mercado de capitais; Ações (Finanças); Investimentos; Bolsa de Valores de São Paulo

This paper aims at identifying trading strategies profitable based on the effects of lead and lag between the markets stock in cash, represented by the Bovespa index, and future, represented the Future Bovespa index in Brazil, using data from high frequency. To achieve this objective and based on historical data of the Bovespa index and Future of the Bovespa index, obtained from the CMA, we constructed four models econometricforecasting: ARIMA, ARFIMA, VAR and VECM. With models adjusted rates, the statistics for the evaluation of forecasting models. Then these were used with the intention to build money-making strategies negotiation. Trading strategies tested were: Liquid Trading Strategy - LTS, Buy and Hold Strategy - BHS and Filter Strategy - better than average return predicted - MFS. These strategies were tested for passive strategy, whichconsists of buying the index at the initial and subsequent sale in the final minute, and only two negotiations throughout the period. The period of analysis of this research lasted from October 3, 2006 to October 2, 2009 for the predicted models, consisting of 25,078 observations. The forecast period was 5 October 2009 to 16 October 2009, consisting of 297 observations. These observations, which comprise the index values every fifteen minutes. The results contradict the literature, showing that trading strategies were profitable in relation to a strategy passive when it considered the costs of trading. This work abnormal returns can be achieved with the use of negotiation strategies with the VAR model on the effects of lead and lag between the index Bovespa and future Bovespa index using high frequency data.