Does the Economic Policy Uncertainty affect the Belgian stock market?

Details

Supervisors

D'Hondt, Catherine

Faculty

Louvain School of Management

Degree label

Master [120] en sciences de gestion, à finalité spécialisée

Abstract

This master’s thesis aims to study the impact of economic policy uncertainty, through the 3 versions of the EPU index for Belgium, namely the national, the French-speaking and the Dutch-speaking ones, on the Belgium stock market. In addition to the study of the predictive power of EPU indices on Belgium’s national stock index (i.e., the BEL 20), we also studied whether this effect is size-dependent and sector dependent, as well as whether the predictive power is different during crisis periods. For all our analyses we worked with a lag of one month, using the excess stock returns of the month t with the EPU data of the month t-1. Studying the predictive power of EPU indices on the BEL 20 between January 2000 and September 2022 we found a statistically significant impact. But contrary to our first hypothesis it was a positive impact, meaning that when the EPU level rose in month t-1, the excess stock returns in month t tended to rise. Concerning the size effect, we studied the predictive power of EPU indices on the BEL 20 and BEL Small excess returns between December 2002 and September 2022. Again, we found a positive impact on both stock indices, but contrary to our second hypothesis, the impact on large caps (i.e., the BEL 20) seems to be more important than on small caps (i.e., the BEL Small), as the beta coefficients linked to the BEL 20 are greater and more statistically significant than those linked to the BEL Small. For our third analysis, we regressed the excess returns of the Belgian sector indices on the EPU indices, between December 2008 and September 2022, to study whether the predictive power is sector dependent or not. We found that indeed, as we assumed in our third hypothesis, it is well sector dependent. Again, almost all the beta coefficients are positive, and the strongest impact is on the consumer discretionary and the industrials sectors. For our last analysis we wanted to study the predictive power of EPU indices during crises. To do so, we used two samples, one for the Subprime crisis between September 2008 and August 2010 and the second for the Covid crisis between March 2020 and August 2021. For our GFC sample we found that all the beta coefficients were positive but only one was statistically significant (between the BEL 20 and the French-speaking EPU). And for our Covid sample we found some positive and some negative links, but none were statistically significant. Moreover, in all our analyses we found no major and constantly better predictive power between the different versions of the EPU index for Belgium.