Identification of clusters of companies in stock indices via Potts super-paramagnetic transitions

Abstract

The clustering of companies within a specific stock market index is studied\nby means of super-paramagnetic transitions of an appropriate q-state Potts\nmodel where the spins correspond to companies and the interactions are\nfunctions of the correlation coefficients determined from the time dependence\nof the companies' individual stock prices. The method is a generalization of\nthe clustering algorithm by Domany et. al. to the case of anti-ferromagnetic\ninteractions corresponding to anti-correlations. For the Dow Jones Industrial\nAverage where no anti-correlations were observed in the investigated time\nperiod, the previous results obtained by different tools were well reproduced.\nFor the Standard & Poor's 500, where anti-correlations occur, repulsion between\nstocks modify the cluster structure.\n