Extracting the exponential behaviors in the market data
Physics and Society
2009-11-13 v1 Data Analysis, Statistics and Probability
Statistical Finance
Abstract
We introduce a mathematical criterion defining the bubbles or the crashes in financial market price fluctuations by considering exponential fitting of the given data. By applying this criterion we can automatically extract the periods in which bubbles and crashes are identified. From stock market data of so-called the Internet bubbles it is found that the characteristic length of bubble period is about 100 days.
Keywords
Cite
@article{arxiv.physics/0608008,
title = {Extracting the exponential behaviors in the market data},
author = {Kota Watanabe and Hideki Takayasu and Misako Takayasu},
journal= {arXiv preprint arXiv:physics/0608008},
year = {2009}
}
Comments
revtex4, 7 pages, 5 figures, proceedings of Apfa5 Conference