eThesis - School of Business electronic theses

OBJECTIVES OF THE STUDY:

In this thesis, I study investor attention around earnings announcements through novel proxies formed from companies' Google search volumes and Wikipedia page views. Firstly, I examine the validity of these data sets as investor attention proxies and how Google and Wikipedia activity change around earnings announcements. In addition, I study whether firm characteristics or investor distraction affect the level of attention investors pay to an announcement. Finally, I assess whether the level of investor attention before and on the announcement day affects earnings sensitivity (relation between earnings surprise and abnormal announcement return).

DATA AND METHODOLOGY:

I collect daily Google search volume and Wikipedia page view statistics from Internet databases for 577 European companies. Based on these data sets, I construct novel proxies for investor attention. I study the changes in these proxies around 8 702 earnings announcements between the years 2008 and 2012. I study all my hypotheses through both simple bivariate analysis and panel regression analysis where standard errors are calculated according to the recommendations of Petersen (2009).

FINDINS OF THE STUDY:

I find strong support for the validity of my investor attention proxies constructed from Google and Wikipedia data. Both of these proxies have significant positive correlation with traditional investor attention proxies. In addition, investor attention, measured with these proxies, starts to climb a couple of days before the earnings announcement, spikes up 20-30 % on the announcement day and stays at an elevated level for about a week.

I also find that the magnitude of the increase in investor attention around earnings announcements is affected by both firm characteristics and investor distraction. Big, highly-followed and low book-to-market companies are documented to have higher pre-announcement investor attention increases. Also, I find that both pre-announcement and announcement day investor attention is mitigated by other earnings announcements coinciding to the same day. Finally, I document a positive relation between announcement day investor attention and earnings sensitivity.