Takeover prediction: A study in predicting takeover targets in Sweden

Sammanfattning: Takeovers are in general associated with bid premiums, giving rise to substantial stock returns for shareholders. This thesis investigates whether it is possible to accurately predict takeovers of Swedish firms using public information, as this would provide an opportunity to exploit bid premiums. We first employ a logit model to identify takeover characteristics of Swedish listed firms between 2005-2014 and second use this estimated model to predict takeover targets in 2015. Even though few characteristics of Swedish targets are identified, we find conclusive evidence that targets are smaller in size and suggestive evidence that targets belong to industries with previous takeover activity. However, given these limited findings, the overall results of the predictive ability of the model therefore suggest that constructing a statistical model to correctly predict Swedish takeover targets is not possible using the methods currently employed in takeover prediction studies. The findings are in line with similar studies based on American and British settings.