Analysis of economic growth fluctuations based on EEMD and causal decomposition

Xuegeng Mao*, Albert C. Yang, Chung Kang Peng, Pengjian Shang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

10 Scopus citations

Abstract

In this paper, we first apply ensemble Empirical Mode Decomposition (EEMD) to analyze the change rate of GDP time series for ten major countries. Each signal is decomposed into four modes with different scales and a residual trend. The variance contribution and averaged period cycle for each mode is then calculated. Results show that economic growth fluctuations for most of the countries mainly fluctuate at a short period between 3–5 years. Then the causal decomposition proposed by Yang et al. is utilized to detect the mutual causation among different countries. The novelty of this method is that the causal interaction is identified in instantaneous phase coherence at a specific time scale. The strengths and directions of causal relationship for each mode among countries differ from each other, which mirrors the fact that the world economic activities are fluctuated and changeable over varying periods. However, no visible causation for the fourth mode (long-term cycle) among countries is characterized, meaning that the development of economy over a long period for most countries depends on its own conditions.

Original languageEnglish
Article number124661
JournalPhysica A: Statistical Mechanics and its Applications
Volume553
DOIs
StatePublished - 1 Sep 2020

Keywords

  • Causal decomposition
  • Economic growth fluctuation
  • Empirical Mode Decomposition
  • GDP time series
  • Hilbert–Huang transform
  • Phase coherence

Fingerprint

Dive into the research topics of 'Analysis of economic growth fluctuations based on EEMD and causal decomposition'. Together they form a unique fingerprint.

Cite this