Google trends versus traditional news and their effect on the European Stock Market

Abstract

This thesis investigates the causal relationship between the mentions of the word Grexit in both Traditional Media and the Google search queries and the closing prices of the General stock market indices of five Euro-periphery eurozone countries (namely the Greek, the Irish, the Italian, the Portuguese and the Spanish). We used daily data for the period spanning from February 2, 2012 till July 12, 2016 and also for the period from October 14, 2014 till July 12, 2016. We implemented the Orthodogonalized Impulse Response Functions and the Local Projection Impulse Response Functions introduced by Jordà (2005). Firstly, we examined the responses of each stock market to a shock in Traditional Media and in Google Trends in each period and both tests agree that the stock markets respond positively to a shock in the Google search queries and negatively to a shock in the mentions of the word Grexit in Traditional Media however most of these responses are not statistically significant. For the second phase, we introduced into the VAR all the stock market returns, our results remain the same with the only difference being that all stock markets respond positively and significantly to a Google Trends shock in the sub-sample period.