The Value-Decreasing Effect of Staggered Boards

Alma Cohen is a Professor of Empirical Practice at Harvard Law School and Charles Wang is an Assistant Professor of Business Administration at Harvard Business School. This post is a reply to a recent post by Yakov Amihud that is based on a paper on staggered boards with Stoyan Stoyanov, available here, which contests the conclusions of an article on the subject by Professors Cohen and Wang, available here. The post by Professors Cohen and Wang draws on their reply paper, available here.

In an article published in the Journal of Financial Economics in 2013, How Do Staggered Boards Affect Shareholder Value? Evidence from a Natural Experiment, we provided evidence that market participants perceive staggered boards to be, on average, value-decreasing. In an April 2016 paper, summarized in a recent post on the Forum (available here), Amihud and Stoyanov attempt to contest our findings. The Amihud-Stoyanov paper (hereinafter AS2016) puts forward several specifications that render our results not statistically significant (though the results largely retain their sign). In response to their work, we have carried out further empirical analysis and found that the non-significant “results” of AS2016 do not hold up when carefully examined. Indeed, the tests that we have conducted to address the issues raised by AS2016 provide a wide array of statistically significant results that are consistent with and reinforce the findings and conclusions of our 2013 JFE article.

In November 2015 Amihud and Stoyanov issued an earlier version of their paper (AS2015, available here) attempting to show that the significance of our results weakens when some observations are excluded. In December 2015 we issued a detailed response (Staggered Boards and Shareholder Value: A Reply to Amihud and Stoyanov, hereinafter the CW Reply), showing that the claims of AS2015 are unwarranted. In their current paper, AS2016, Amihud and Stoyanov drop some of their claims, but they retain other arguments that were already shown to be unwarranted in the CW Reply and add new claims.

Notably, with the exception of its replication of our 2013 specifications, none of the results presented in AS2016 are statistically significant, and the results based on our sample largely have a sign consistent with the conclusions of our 2013 JFE article. Thus, even assuming that the results reported by AS2016 hold up to scrutiny, the lack of statistical significance implies that they would equally be consistent with both the view that staggered boards are value-decreasing and the view that they are value-increasing.

In any case, this point is moot because the non-significant “results” of AS2016 do not hold up. As is true of any event study without a large number of observations, the statistical significance of our 2013 results could be sensitive to the removal of a small number of observations, especially if such removal is designed in a strategic way. However, in the case of our study, the results retain their statistical significance under a wide range of tests conducted to address the concerns raised by AS2016.

One line of argument pursued by both AS2015 and AS2016 is that the results of our 2013 JFE article are unduly driven by a few particular observations that have significant returns in a certain direction. However, as we already showed in the CW Reply, to address this concern without cherry-picking observations for exclusion, it is best to apply an unbiased and neutral method of excluding observations with unusually large returns from both sides of the distribution. Doing so, we obtain an array of results (tabulated in the CW Reply) that are statistically significant and consistent with the conclusions of our 2013 JFE article.

In a new line of argument not made in AS2015, AS2016 claims that our 2013 results become statistically insignificant when excluding some small companies and thus cannot inform the assessment of how staggered boards affect value in normal-sized firms. Our analysis shows that this claim too does not hold up when carefully examined. Indeed, we obtain an array of significant results for normal-sized firms, consistent with the results and conclusions of our 2013 JFE article. This full set of results will be included in a revision of the CW reply that we will issue.

Notably, AS2016 does not address, or even mention, empirical results presented in the CW Reply that contradict, or are directly responsive to, its claims. For example, AS2016 does not engage with or even mention the analysis reported in the CW Reply showing that AS2015’s non-significance results do not hold up, and the results retain their significance and consistency with the conclusions of our 2013 JFE article, when using two refined definitions of the set of treated firms are used.

Overall, AS2016 fails to cast doubt on the conclusions of our 2013 JFE article. Our re-examination of the data provides a wide array of statistically significant results that are consistent with and buttress the conclusions of our 2013 JFE article. A substantial part of these results is presented in the CW Reply available here, and the rest will be included in the revision of this reply that we will issue.

Both comments and trackbacks are currently closed.