CMStatistics 2017: Start Registration
View Submission - CFE
A1700
Title: Predicting intraday returns based on overnight returns for the US stock market Authors:  Hao Li - University of Amsterdam (Netherlands) [presenting]
Cees Diks - University of Amsterdam (Netherlands)
Valentyn Panchenko - Univerisity of New South Wales (Australia)
Abstract: Research on high frequency data for the S\&P 500 Exchange-traded Fund (ETF) from 2003 to 2013, using Trade and Quote (TAQ) data, documents nonlinear dependence between overnight returns and subsequent intraday returns based on both parametric and nonparametric models. Specifically, the dependence between overnight returns and the first half-hour returns is significantly negative while the dependence between overnight returns and the last half-hour returns is significantly positive. However, although the returns in-between the first half-hour and the last half-hour do not linearly depend on overnight returns, using parametric as well as nonparametric methods we find them to depend on overnight returns nonlinearly.