Abstract
In event studies the typical methodology utilized has been some form of residual analysis. Residual analysis depicts what happens, on average, to the sample firms and does not concern itself with each individual firm. We will use two kinds of models discussed in Section 4.3 which have been used before both in short-term and long-term event studies: the Mean Adjusted Return (MAR) model and the Industry Adjusted Market (IAM) model. Our approach is similar in spirit to the one used by Agrawal, Jaffe, and Mandelker (1992), although we have not adjusted for size. We realize that there might be a size effect due to the nature of the problem — targets are generally smaller than the acquirers — but there is no market capitalization data across event time readily available which is machine readable.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
See Franks, Harris, and Titman (1991), Agrawal, Jaffe, and Mandelker (1992), Barber and Lyon (1997), Kothari and Warner (1997), Rau and Vermaelen (1998), and Mitchell and Stafford (1998).
Brown and Warner (1980 and 1985) cite clustering of announcement dates as a potential problem with the MAR model. Examination of our sample shows that announcement dates are not clustered and assume that the individual announcements are independent of one another.
See Mitchell and Stafford (1998).
See Shanken (1985).
Langetieg (1978) used a two-factor model including industry and market factors.
In prior studies on longer term post-event abnormal returns performances the horizon length of examination has been up to 60 months (see, for example, Agrawal, Jaffe, and Mandelker (1992), Loderer and Martin (1992) and Loughran and Vijh (1997)). For a sample reference list on long-horizon post-event studies see Kothari and Warner (1997).
See, for example, Dodd (1980), Asquith (1983) or Copeland, Lemgruber, and Mayers (1987).
For example, Brown and Warner (1980) define the estimation period as t = 89 until 11, Brown and Warner (1995) use t = 239 until 6. Alexander, Benson, and Kampmeyer (1984) use as the estimation period t = 150 up to — 31. Dodd (1980) ends the estimation period at t = 41, whereas Dennis and McConnell (1986) end at t = — 16. Gerke, Garz, and Oerke (1995) define the estimation period as t = 200 up to — 41.
A year in Germany has approximately 250 trading days with a range of 246 to 253 days in our sample due to different regional holidays. Note that 250 trading days does not necessarily mean 250 returns or quotes due to missing values.
For a description of the indices and how they are calculated see Deutsche Börse AG (1995).
See Barnes (1984).
See Fama (1976).
Following Brown and Warner (1985).
See Brown and Warner (1980).
Technically, the performance could be equal to the benchmark, thus leading to a zero abnormal return. In that case, the investment, i.e. the acquisition transaction, has a NPV of zero. Managers who maximize the market value of their firm should undertake such an investment because it has a non-negative risk-adjusted net present value. Since an exact zero abnormal return is relatively unlikely, we have for all practical purposes a positive or negative abnormal return.
Thus, we do not treat the case where minority shareholders may “loose” to a certain degree because they do not get their full share of the achieved gains since some of them could possibly have been transferred to the majority shareholder. If positive abnormal returns were shown, we argue that the transaction was still a success from the point of view of the target shareholders.
Note that we mean here all target shareholders, thus both the majority and the minority ones.
Concerning synergistic effects in general, however, it is unlikely that all benefits go exclusively to the acquiring firm. Some synergies, especially those arising from cost reduction, should go or remain at the target company, at least partially.
This is applicable in general for German transactions and applies to 17 out of 34 companies in our data set. Nevertheless, if abnormal negative returns are found for the sub-sample with publicly traded majority shareholders, one could analyze their performance. Conclusions, however, would be subject to statistical sampling, relation target-buyer and most importantly, what other events did influence the majority shareholders’ share performance over the period in question.
For an exact definition of “industrial” refer to Chapter 1.
At least in the semi-strong form. See Fama (1976) for a discussion on efficient markets.
Rights and permissions
Copyright information
© 1999 Springer Fachmedien Wiesbaden
About this chapter
Cite this chapter
Scheller, K.R. (1999). Methodology and Hypothesis. In: Performance of Corporate Acquisitions over the Medium Term in Germany. Deutscher Universitätsverlag, Wiesbaden. https://doi.org/10.1007/978-3-663-08842-4_7
Download citation
DOI: https://doi.org/10.1007/978-3-663-08842-4_7
Publisher Name: Deutscher Universitätsverlag, Wiesbaden
Print ISBN: 978-3-8244-6921-5
Online ISBN: 978-3-663-08842-4
eBook Packages: Springer Book Archive