Volume 95, Number 6, September 2011
|Number of page(s)||5|
|Section||Interdisciplinary Physics and Related Areas of Science and Technology|
|Published online||01 September 2011|
Universal behaviour of interoccurrence times between losses in financial markets: An analytical description
Institut für Theoretische Physik, Justus-Liebig-Universität Giessen - 35392 Giessen, Germany
2 Centro Brasileiro de Pesquisas Fisicas and National Institute of Science and Technology for Complex Systems 22290-180 Rio de Janeiro-RJ, Brazil
3 Santa Fe Institute - Santa Fe, NM 87501, USA
Accepted: 29 July 2011
We consider 16 representative financial records (stocks, indices, commodities, and exchange rates) and study the distribution PQ(r) of the interoccurrence times r between daily losses below negative thresholds −Q, for fixed mean interoccurrence time RQ. We find that in all cases, PQ(r) follows the form PQ(r)∝1/[(1+(q− 1)βr]1/(q−1), where β and q are universal constants that depend only on RQ, but not on a specific asset. While β depends only slightly on RQ, the q-value increases logarithmically with RQ, q=1+q0 ln(RQ/2), such that for RQ→2, PQ(r) approaches a simple exponential, PQ(r)≅2−r. The fact that PQ does not scale with RQ is due to the multifractality of the financial markets. The analytic form of PQ allows also to estimate both the risk function and the Value-at-Risk, and thus to improve the estimation of the financial risk.
PACS: 89.65.Gh – Economics; econophysics, financial markets, business and management / 05.40.-a – Fluctuation phenomena, random processes, noise, and Brownian motion / 89.75.Da – Systems obeying scaling laws
© EPLA, 2011
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.