Trading por arbitragem estatística
Carregando...
Arquivos
Data
2016-08-12
Autores
Orientador(res)
Ruilova Terán, Juan Carlos
Métricas
Título da Revista
ISSN da Revista
Título de Volume
Resumo
This paper proposes a tool to detect statistical arbitrage opportunities in a particular pair of stocks in the Brazilian market. The technique is based on the construction of a synthetic asset that presents mean reversion process. The forecasts will be realized in the form of conditional probability density, which is based on econometric techniques such as autoregressive process (AR) and conditional variance of the residuals (GARCH). A trading system able to take advantage of mispricing observed by the synthetic asset dynamic is created. The approach will consider prices in one minute intervals and positions with limit orders and to market. Still be considered transaction costs and analysis of P&L for the cases addressed.