Financial variables and real economic activity : empirical evidence from G7 and EU - 15 countries

Abstract

The purpose of this thesis is to examine the ability and the degree up to which a number of financial variables can predict the level of future economic activity. The variables that are going to be examined concerning their predictive power are stock market returns, their volatility, and term structure of interest rates as well as inflation. As measures of the real economic activity we are going to use the Real Gdp per capita and in some cases the percentage change in industrial production index. Our main target is to define the existence and then the direction of the dependence between finance and economics. In the first part of this thesis we will try to give a theoretical explanation concerning how financial variables can predict economic activity. At that point, there will be an extended presentation of relative to our subject researches (published in various journals) which will help us in our theoretical approach. The second part describes the data and the econometric methods. For the conduction of our research, we will use the Vector Autoregressive model and then we will move on to the variance decomposition and impulse response analysis in order to identify relationships between the variables under investigation. VAR model is used very often for the prediction of systems with interrelated time series and for the analysis of the shocks’ behaviour over the system of variables. This empirical research is going to be conducted for the EU-15 as well as the G-7 countries. The third part summarizes our main conclusions and provides the statistical tables of our econometric tests.