In this paper we address the problem of forecasting non-gaussian portfolio returns over multiple time scales. We apply a relatively new technique for estimating portfolio returns by considering higher order mutual information. This technique is based on two methodologies: Independent Component Analysis and Gaussian mixtures. We apply this model to intraday data from the ASX. Our findings illustrate that this model is particularly useful in estimating portfolio returns over a short time scale when the distribution is highly non-Gaussian. © Springer-Verlag 2003.
CITATION STYLE
Lo, K., & Coggins, R. (2004). Intraday analysis of portfolios on the ASX using ICA. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2690, 934–938. https://doi.org/10.1007/978-3-540-45080-1_133
Mendeley helps you to discover research relevant for your work.