Criação de um método para balanceamento de estratégias automatizadas intradiárias para transação de contratos de derivativos

Abstract

The stock market has experienced, in recent decades, a great architecture evolution due to the latest technological revolution, offering investment options, at a click, for individual investors or even large specialized companies. The current technological gain is the trading of assets through specialized algorithms. The objective of this study is demonstrating the method of analysis and definition of a portfolio of investor robots, analyzing, through backtests, indicators for defining the strategy that will be placed in a real account and the financial feasibility of a portfolio of day trade strategies. The research, also bibliographical, used published theoretical references, in addition to knowing and analyzing the cultural and scientific contributions of the past, existing on the subject. The study was based on a real case and dozens of automated strategies were created and the best thirty were selected by backtests that, in principle, were profitable during the two years tested. Those selected to participate in the study were operating simultaneously on the same asset, the BOVESPA mini-index, ticker WIN. Fifteen strategies were selected, which made up the so-called “A” portfolio. The other fifteen strategies made up the “B” portfolio. Both portfolios achieved high returns relative to any benchmark investment (with less risk involved). The experiment should be repeated for a longer time and with a larger number of strategies portfolio, preferably in other assets.