作者: D Grech , Z Mazur
DOI: 10.1016/J.PHYSA.2004.01.018
关键词:
摘要: Abstract We apply the Hurst exponent idea for investigation of DJIA index time-series data. The behavior local prior to drastic changes in financial series signal is analyzed. optimal length time-window over which this can be calculated order make some meaningful predictions discussed. Our prediction hypothesis verified with examples 1929 and 1987 crashes, as well more recent phenomena stock market from period 1995 2003. Some interesting agreements are found.