6533b823fe1ef96bd127f7b5
RESEARCH PRODUCT
Identification of clusters of companies in stock indices via Potts super-paramagnetic transitions
L. KullmannJános KertészRosario N. Mantegnasubject
Statistics and ProbabilityStatistical Mechanics (cond-mat.stat-mech)SpinsFOS: Physical sciencesCondensed Matter PhysicsStock market indexParamagnetismCluster (physics)Statistical physicsCluster analysisStock (geology)Condensed Matter - Statistical MechanicsPotts modelMathematicsdescription
The clustering of companies within a specific stock market index is studied by means of super-paramagnetic transitions of an appropriate q-state Potts model where the spins correspond to companies and the interactions are functions of the correlation coefficients determined from the time dependence of the companies' individual stock prices. The method is a generalization of the clustering algorithm by Domany et. al. to the case of anti-ferromagnetic interactions corresponding to anti-correlations. For the Dow Jones Industrial Average where no anti-correlations were observed in the investigated time period, the previous results obtained by different tools were well reproduced. For the Standard & Poor's 500, where anti-correlations occur, repulsion between stocks modify the cluster structure.
| year | journal | country | edition | language |
|---|---|---|---|---|
| 2000-02-16 |