Search for content and authors
 

Analysis of times between events by methods of statistical physics

Mateusz Denys 1Tomasz Gubiec 1,2Ryszard Kutner 1Maciej Jagielski 1,2,3Eugene Stanley 2

1. Institute of Experimental Physics, Faculty of Physics, University of Warsaw (FUW), Pasteura 5, Warsaw 02-093, Poland
2. Center for Polymer Studies and Department of Physics, Boston University, 590 Commonwealth Ave, Boston, MA 02215, United States
3. Department of Management, Technology and Economics, ETH Zürich (ETHZ), Scheuchzerstrasse 7, Zürich CH-8092, Switzerland

Abstract

One of the most significant questions in the analysis of losses in financial market time series, closely related to an economic concept of value at risk (VaR), is description of times between subsequent losses of a particular magnitude (called interevent times). We provide such a description under two complementary approaches.

First, we present a model of superstatistics founded on the continuous-time random walk (CTRW) model and the extreme value theory (EVT). The model provides a closed analytic formula for the universal distribution of interevent times valid for excessive losses and profits (irrespective of the asset type or the time resolution of data) as well as for some geophysical data of earthquakes [1]. Our description is an alternative to the approach involving q-exponential functions [2].

Secondly, we propose an agent-based model of financial markets being a generalization of the Potts model from statistical mechanics [3]. A value of the spin variable in the model represents a short (–1), neutral (0), or long (+1) position taken by the investor on the market. The action, or decision, of the investor, i.e., buying or selling a stock is, in turn, defined as a change of the spin value (positive for buying and negative for selling). Thus, we identify the state of a spin with the actual market state of the investor (the position taken on the market), not the market action (buying or selling) as in the previous works. The model reproduces, inter alia, the empirical shapes of the autocorrelation function of both usual and absolute market returns, as well as the distribution of interevent times [4].

[1] M. Denys, T. Gubiec, R. Kutner, M. Jagielski, H.E. Stanley, Phys. Rev. E 94 (4), 042305 (2016).

[2] J. Ludescher, C. Tsallis, and A. Bunde, Europhys. Lett. 95 (6) 68002 (2011).

[3] M. Denys, T. Gubiec, R. Kutner, Acta Phys. Pol. A 123 (3) 513–517 (2013).

[4] M. Denys, T. Gubiec, R. Kutner, arXiv:1411.1689 (2014).

 

Legal notice
  • Legal notice:
 

Related papers

Presentation: Oral at Econophysics Colloquium 2017, Symposium A, by Mateusz Denys
See On-line Journal of Econophysics Colloquium 2017

Submitted: 2017-03-13 13:58
Revised:   2017-03-16 15:28