Testing Chaos and Fractal Properties in Economic Time Series

Maria I. Loffredo
Dipartimento di Matematica, Università di Siena, I-53100 Siena, Italy
e-mail : loffredo@unisi.it
 
Abstract
Search for empirical evidence of chaos and testing fractal and other statistical
properties in the framework of time series analysis are carried on as a
preparatory step in order to apply these concepts to data proper of Financial
Markets and deal with the puzzling failure of traditional economic theories.
Concepts like correlation dimension and Lyapunov exponents are discussed and
simple Mathematica programs are given for their evaluation. Before their
application to real economic data, a test on a well known nonlinear dynamical
system, through the correspondent reconstructed phase space and time series, is
carried out.