Search for content and authors
 

Correlation matrix decomposition of intraday WIG20 fluctuations

Rafał Rak 1Stanisław Drożdż 1,2Jarosław Kwapień 2

1. University of Rzeszow, Institute of Physics, Rejtana 16, Rzeszów 35-310, Poland
2. Polish Academy of Sciences, Institute of Nuclear Physics (IFJ PAN), Radzikowskiego 152, Kraków 31-342, Poland

Abstract

Quantifying correlations among various financial assets is of great interest both for practical as well as for fundamental reasons. These relate for instance to the theory of risk management, and to the universal aspects of evolution of complex systems, respectively. The financial correlations can be classified as correlations in space (among companies comprised by a single stock market, among a group of subjects and even between different stock markets) and correlations in time, the later directly reflecting nature of the financial dynamics. Both types of correlations can be quantified by adopting some appropriately defined correlation matrices, such that the related procedure allows to identify the noise and collectivity components in the underlying dynamics. The random matrix theory (RMT) provides a useful null hypothesis in this respect. Noise is consistent with the RMT predictions. What however is even more interesting in this connection are deviations. These identify certain system-specific, non-random - thus perhaps deterministic - aspects of the dynamics. Due to strong nonstationarity of the parameters characterizing the financial world, the above interplay between collectivity and noise is also expected to vary in time and in space. In this contribution we concentrate on temporal aspects of the Warsaw Stock Market evolution as represented by the WIG20 index. This study is based on high frequency (1 min) WIG20 recordings over the time period between January 1999 and October 2005. One principal conclusion is that a bulk of this stock market dynamics is governed by the uncorrelated noise-like processes. The formalism of the correlation matrix allows however to filter out some small number of components of coherent short term repeatable structures in fluctuations that may generate some memory effects.

 

Legal notice
  • Legal notice:
 

Related papers

Presentation: Oral at 2 Ogólnopolskie Sympozjum "Fizyka w Ekonomii i Naukach Społecznych", Econophysics, by Rafał Rak
See On-line Journal of 2 Ogólnopolskie Sympozjum "Fizyka w Ekonomii i Naukach Społecznych"

Submitted: 2006-02-22 16:24
Revised:   2009-06-07 00:44