Dresden 2006 – scientific programme

Parts | Days | Selection | Search | Downloads | Help

AKSOE: Physik sozio-ökonomischer Systeme

AKSOE 10: Poster Session (posters are expected to be displayed the full day 8:30-18:00)

AKSOE 10.7: Poster

Wednesday, March 29, 2006, 16:00–18:00, P2

Statistical Analysis of Stock-Market Fluctuations: A Single Parametric Formulation — •Pragya Shukla — Department of Physics, IIT Kharagpur-721302, West Bengal, India

An understanding of the stock-market dynamics requires a characterisation and quantification of the correlations in the prices of of various stocks at a series of time-steps. To obtain the mutual correlations between N such stocks, therefore, one needs to analyze the correlation matrix No-dqCNo-dq, with its elements as the correlation function between two stock-prices, represented by the time series. However the complexity of the stock market leads to complicated fluctuation of prices and an exact determination of the associated time-series is not always possible. The correlations can be obtained, therefore, only within a certain distribution, resulting in No-dqCNo-dq as a random matrix. The mutual dependence of various stocks and their effect on stock market fluctuations can thus be characterized by the nature of the associated correlation matrix, that is, by its eigenvalue and eigenvector statistics. Our analysis of the correlation matrices of various stock markets reveals the existence of a deep level of universality underlying their statistical behavior.

. ......