Time varying beta and the Euro zone financial crisis: evidence from European Industrial and Financial sectors.

Abstract

In this study, we investigate the effects of Euro zone financial crisis at the industrial and the financial level. The empirical research was conducted using non-linear models from the GARCH family and OLS procedure in order to test the performance of the Capital Asset Pricing model and to estimate the time varying systematic risk. Weekly and monthly industrial sectors and financial sectors indices from seventeen European countries are utilized. The range of the data is from 2002 to 2012. We permit only one specification for the error distribution – Student’s distribution. The covariance is determined by using residuals of best-fit AR (p) and MA (q) models of individual and market portfolios. Overall, the mostly financial sectors present more volatility than the industrial sectors. Furthermore, it is observed that the OLS procedure overestimate betas, contrary to Maximum Likelihood. The mostly betas are defensive. Results indicate that the market risk increases sharply when the financial crisis hit each Euro zone country. The evidence from the rolling regression results generates severe doubts concerning the existence of structural breaks in betas coefficients. There is a relationship between the break in betas and the financial crisis since the most breaks in European betas had been occurred before the global financial crisis of 2007. Finally, it was provided that the break appear in financial sector at first and then in industrial sector. All these findings can pilot to significant inferences about the validity of the CAPM model.