eThesis - School of Business electronic theses

The purpose of this thesis is to study how mergers affect analysts' ability to predict earnings of the company by comparing pre-merger forecast errors to post-merger errors. In addition to examining the change in forecast accuracy after merger, I also study the permanence of this changed forecast accuracy as well as the factors affecting the change. I also aim to provide new information on how the information environment of the target company affects to forecast accuracy, since this is not studied in previous studies. Finally, I study the possible forecast bias after mergers. I aim to provide updated as well as new information about analyst forecast accuracy after mergers with significantly newer and wider data to help forecast users to assess the possible factors affecting forecast accuracy.

My sample consists of 810 mergers between listed US companies completed in the years 1991-2011. To test the change in the forecast accuracy I compare the forecast error in three post-merger years to forecast errors in the pre-merger years. In addition, I run a multiple regression analysis to find which factors affect the changes in the forecast accuracy.

I find that the accuracy of analyst earnings forecasts decreases significantly after mergers and the forecast accuracy prevails on the decreased level at least three years after the merger. As predicted the information complexity and uncertainty have a negative relation with forecast accuracy in post-merger years. My test shows that, in addition to variables studied in previous research, also target's information environment has a significant effect on analysts' forecast accuracy in the context of mergers. I find that the overprediction is significantly higher in post-merger years, however the explaining component is rather decreased precision than bias.