The Decoupling of Voting and Economic Ownership (IVOR nr. 88) 2012/3.3.2.1:3.3.2.1 Sample Size Neglect

The Decoupling of Voting and Economic Ownership (IVOR nr. 88) 2012/3.3.2.1

3.3.2.1 Sample Size Neglect

Deze functie is alleen te gebruiken als je bent ingelogd.Deze functie is alleen te gebruiken als je bent ingelogd.Deze functie is alleen te gebruiken als je bent ingelogd.

According to a recent empirical study, the merger wave that swept the markets in the 1990s destroyed a staggering $216 billion in value for the shareholders of the acquiring firms 1 Who took the decision to make these acquisitions? Managers, to begin. Perhaps they suffered from the cognitive bias of overconfidence, which may have led them to overestimate the accuracy of their valuations and their ability to create value.2 But even if managers took the decision to make these acquisitions, shareholders will often have approved them.3 Perhaps in doing so, they suffered from cognitive biases as well. Overconfidence on the part of shareholders is unlikely to be the culprit, because overconfidence refers to confidence in one's own capabilities rather than the capabilities of someone else, such as management of the portfolio firm. But there is a cognitive bias that can lead people to put too much faith in someone else's capabilities: sample size neglect. Sample size neglect is caused by the so-called representativeness heuristic and generates the 'hot hand' phenomenon. This phenomenon can be witnessed when spons fans become convinced that a basketbal) player who has made three shots in a row is on a hot streak and will score again, even though there is no evidence of a hot hand in the data.4 Now substitute the scoring basketbal) player for a CEO who has made a number of successful prior acquisitions, and we can see how shareholders might be inclined to approve the next big deal presented to them even if those prior acquisitions are not fully representative of management's ability to make the proposed acquisition a success, for example because they were much smaller. The aforementioned study of large loss deals offers some evidence supporting this proposition: firms that made these deals, it turns out, were serial acquirers, and the acquisitions made in the two years prior to the large loss deal had created substantial shareholder value.5

Whether shareholders who approved the large loss deals were infiuenced by this positive track record is an empirical question, but it certainly seems plausible. One piece of anecdotal evidence stems from the $100 billion-plus acquisition of ABN Amro by a consortium of three European banks, billed as the largest banking deal ever. The acquisition took place just before the financial crisis erupted in the fall of 2007. For one consortium member in particular, Fortis, the acquisition entailed significant risks. Apart from the risk inherent to splitting a large financial institution such as ABN Amro in three parts and then to successfully integrate it, there was significant financing risk, which becomes clear from the fact that Fortis needed to raise about eighteen billion dollars in equity, an amount roughly corresponding to half of its own market capitalization. The risks were widely publicized at the time, and have arguably materialized as Fortis collapsed in 2008 and was nationalized to avoid a meltdown.6

And yet, the acquisition was approved by more than ninety percent of Fortis shareholders.7 Why? Possibly because they attached too much significance to management's prior successes, which were emphasized by management in its attempts to sell the deal." Indeed, one shareholder was quoted as saying that he voted in favor of the bid 8to support Fortis management" and that he had "rock solid" confidence in management's plans.9 More tellingly, many institutional investors were likely to have followed ISS 's recommendation to vote in favor of the acquisition, which was based in part on the fact that Fortis had a strong record of prior—but smaller—acquisitions.10

Deze functie is alleen te gebruiken als je bent ingelogd.