Non-Linear Noise Reduction and Detecting Chaos: Some Evidence from the S&P Composite Price Index

Donald George, Richard Harrison, W. Lu, L.T Oxley, Dahai Yu

Research output: Contribution to journalArticlepeer-review

Abstract

Academic and applied researchers in economics have, in the last 10 years, become increasingly interested in the topic of chaotic dynamics. In this paper we undertake non-linear dynamical analysis of one representative time series taken from financial markets, namely the Standard and Poor's (S&P) Composite Price Index. The data is based upon (adjusted) daily data from 1928 to 1987 comprising 16 127 observations. The results in the paper, based on the Grassberger–Procaccia (GP) correlation dimension measurement in conjunction with non-linear noise filtering and the surrogate technique, show strong evidence of chaos in one of these series, the S&P 500. The analysis shows that the accuracy of results improves with the increase in the number of recording points and the length of the time series, 5000 data points being sufficient to identify deterministic dynamics.
Original languageEnglish
JournalMathematics and Computers in Simulation
Issue number1
Publication statusPublished - Jun 1999

Keywords

  • time series analysis
  • stock exchange
  • Kolmogorov entropy
  • chaos theory

Fingerprint Dive into the research topics of 'Non-Linear Noise Reduction and Detecting Chaos: Some Evidence from the S&P Composite Price Index'. Together they form a unique fingerprint.

Cite this