A dynamic study of the characteristics, predictive power and profitable opportunities underlying takeover prediction models : evidence from the UK and the USA

Abstract

Takeover prediction models can be defined as a cross-sectional regression using a firms corporate accounting information in order to measure its likelihood of becoming an acquisition target. Given the theoretical and empirical evidence that these models are unstable both over time and across economies, the research objectives of this thesis are to (i) determine persistent takeover characteristics over a ten-year time span, (ii) analyze the model's predictive accuracy and (iii) investigate whether long-term abnormal returns can be earned by investing in realistic portfolios based on the model's predictions and whether the model's profitability is correlated with its predictive ability.