A Multimarket Study of Efficient Market Hypothesis in Granger Causality Networks

Resumo

Granger causality analysis is a statistical tool used to identify predictive relationships between time series. This article applies this methodology to assets that make up indices of six countries—Germany, Turkey, Japan, India, Brazil, and the USA—with the aim of constructing a causality network for each of them that reveals the interdependencies between these assets. Using historical price data of the assets, the analysis aims to identify the directions and strengths of the predictive relationships of the method. By examining the global characteristics of the networks, this study seeks to advance mathematical knowledge about financial markets. Additionally, the analysis explores how the characteristics of the networks compare with the Hurst exponent of the respective assets and indices in order to identify relationships between their interpretations. The methodology applied in this study can also serve as a basis for future research on causality networks in different contexts and markets.