Factor analysis of financial time series using EEMD-ICA based approach

• Main Authors: Lu Xian, Kaijian He, Chao Wang, Kin Keung Lai
• Format: Article
• Language: English
• Published: Elsevier 2020-01-01
• Series: Sustainable Futures
• Subjects: Ensemble empirical mode decomposition; Independent component analysis; Underlying factors; Financial time series
• Online Access: http://www.sciencedirect.com/science/article/pii/S2666188819300036

Analyses of financial time series and exploring its underlying characteristic factors are longstanding research problems. Ensemble empirical mode decomposition (EEMD) and independent component analysis (ICA) are two methods developed to deal with these problems in nonlinear and non-stationary time series. Recently, a new model integrating the two methods (called EEMD-ICA) has been proposed for single-channel signal processing. For better exploration of the underlying factors of single financial time series, this paper attempts to conduct the empirical analysis based on EEMD-ICA model for this task. In the proposed approach, the single financial time series is decomposed into several statistically independent components. The decomposed components reveal more information which include the supply and demand, cycle, economical development and other factors. We find the related economic variable for every decomposed component by analysis and comparison. Finally, the crude oil price is used as the typical financial time series for illustration and verification. The empirical results show that EEMD-ICA based analysis approach is a vital technique for exploring the underlying factors of single financial time series.