eThesis - School of Business electronic theses

This thesis explores the acquisition discounts of unlisted targets reported in US takeovers with a European high-tech focused dataset, and a specific view on the determinants of that discount. More specifically, I study the interrelatedness of patents, target shareholders’ demand for liquidity, and the information asymmetry as explanatory measures of the acquisition discount. To provide a more thorough view of the role of patents, liquidity, and information asymmetry in acquisitions, I also study the determinants of the target having patented its innovations prior to the acquisition announcement, and those of the acquirer abnormal announcement return. In the former, I proceed with a specific focus on dimensions of information asymmetry as reasons for a target having patents. In the latter, my focus is similar to the study of the acquisition discounts. On the one hand, my results should provide validation for those found in the US, and on the other, a more thorough understanding of the listing effect, and the role of patents, liquidity, and information asymmetry in acquisitions of unlisted high-tech targets.

I employ a unique combination of data related to acquisitions of European targets. First, I obtain deal data, and data related to M&A and IPO volumes, as well as valuation multiples related to the target from the Thomson-Reuters SDC Platinum database. Then, I merge that data with acquirer price and financials data from Datastream, as well as target patent data from the European Patent Office free databases. I end up with multiples data for 1538 deals, and acquirer return data for 637 deals. To facilitate valid hypothesis tests, the acquisition discount data are transformed from the left skewed economic discount to a normally distributed transformed discount using the square root transformation. Moreover, the MacKinnon-White (1985) covariance matrix is employed to account for heteroskedasticity in finite samples. In the announcement return model, I employ the wild bootstrap to account for heteroskedasticity and leptokurtosis in the dependent variable. Finally, while the selection model is logit, the model is resampled utilizing the jackknife procedure to increase the validity of the hypothesis tests and mitigate potential overrejection of the null hypotheses.

My results are consistent with the majority of my hypotheses. In fact, only the hypothesis that the mere existence of patents decrease acquisition discounts is not supported by the data. My findings are consistent with the notion that the acquisition discount is in part caused by decreased availability of liquidity to target shareholders. Moreover, my results support the notion that any means of mitigating information asymmetry is valuable to both the acquirer and the target. Finally, I find that acquirers are rational in trying to mitigate the information asymmetry in one dimension while it increases in another.