Time-dependent complexity measurement of causality in international equity markets: A spatial approach

Abstract

A nonlinear temporal complexity approach is proposed in order to properly model the evolution of randomness, self-similarity and information transmission for thirty-four international stock markets, grouped into four major geographical segments: America, Europe, Asia and Oceania. The causality between each type of time-dependent measures is investigated to assess the state system flows across all geographic segments. The empirical results show that self-similarity is vastly transmitted between financial markets. Moreover, significant emissions of entropy and self-similarity are found between America and Europe. Informational flows are observed only between Europe and Asia, and Europe and Oceania. Our findings may have important implications for portfolio management based on the spatial dimension of spillovers of stochasticity, self-similarity and system state informational content for world stock markets. These results would not have emerged by means of standard econometric approaches of causality investigation in financial returns.