Creative Commons

This work is done asthe final part of my studies at the International Hellenic University. It was written, among others, as an effort of developing an overall better understanding of the relationship between core macroeconomic variables and the stock market. The role of the financial sector within any economic environment and the basic variables of aggregate economic activity, together with their respective interrelations with various stock markets around the globe have been thoroughly studied in the empirical literature. However, as it will be seen extensively below, most of the traditional or even slightly more advanced econometric models that have been used, fail to consider the non-linear and asymmetric interdependence between macroeconomic and financial variables. Taking into account that the economic policy decision-making and the portfolio asset allocation are both regime-varying processes, we develop an analytical framework for dealing with this asymmetry by using the measures of transfer and partial transfer entropy with particular focus on their asymmetric versions. Starting with an unrestricted VECM that is accompanied by a standard Granger Causality analysis, we find significant evidence of lead-lag relationships. Furthermore, the model-free assumption of the proposed direct causality methods seems to offer additional value when testing for the true causal links between the variables on top of the non-linear and asymmetric nature of the macrofinance data.