Wednesday, September 22, 2004

Board Size debate continued: Faleye upholds tradition!

A more traditional view of Board size: Smaller is better

A great thing about academic conferences is that you can see multiple sides of most issues (group think is generally not a problem!) For instance at the upcoming FMA meetings Olubunmi Faleye will present a paper that represents the traditional view (at least since Yermack 1996) that smaller is better when it comes to board size. To quote myself in the August FinanceProfessor newsletter:

“Faleye reports that large boards of directors are less likely to replace existing CEOs and if the CEO replaced, less likely to find a successor from outside the firm. Moreover, when firms announce smaller boards, the firm's stock return is positive.” These findings lead Faleye to conclude "that a large size hinders the board's ability to perform its monitoring functions, and lends additional support to the current drive toward smaller boards."

That Faleye uses some cool techniques and has some important findings that help shed light on the board controversy is undebatable.

1. Rather than examining CEO turnover conditional on firm performance, the author looks at unconditional performance. Why? A good board would replace a CEO BEFORE performance suffered, not after. (The logic for this comes form Hermalin 2004.)

2. Larger boards are significantly less likely to replace managers! “The probability of CEO turnover during the period is significantly decreasing in boardsize: An additional director reduces the odds in favor of turnover by 13%!” (Emphasis is mine.) I would caution CEOs who would like to use this finding to entrench themselves that this is non linear or else putting 8 new board members would lead to life long entrenchment.

3. Replacement by smaller boards is tied to better stock performance: “announcement period abnormal return is significantly negatively related to board size, which implies that investors view CEO replacement decisions made by smaller boards more positively than those made by larger boards.” This could be because smaller boards are more apt to replace with an outsider.

So how does this paper fit with the papers by Coles, Daniel, and Naveen (CDN), or with Larcker, Tuna, and Richardson (LTR)? Interesting question!! They are not as different as they may first appear.

For starters that the findings are slightly different should not be surprising. The authors are looking at different samples and more importantly they are not looking at precisely the same thing (although close!). For example CDN use Q values as the dependent variable. LTR use “a wide set of dependent variables, e.g. abnormal accruals, excessive CEO compensation, debt ratings, analyst recommendations, Q, and over investment.” Faleye looks at the likelihood of CEO replacement and abnormal returns on the replacement.

That said, I think what I will take away from these papers is further evidence that smaller boards do appear to be better monitors. BUT (and this is big) better monitoring can come at a cost of expertise and advisement. This cost varies with firm specific factors.

What is still a bit troubling to me is that if the market knows this and incorporates it in the price ( for example LTR show that smaller boards are better where monitoring is more important), why is there still a difference in stock returns following the replacement? And it is more pronounced for smaller boards, the opposite of what would be consistent with the market anticipating the change.

Is this return somehow tied to firm specific factors (maybe industry specific) that is not being picked up by Faleye?

Obviously much left to be said on the topic. It is unfortunate that LTR and Fayele are not in the same session in New Orleans.

No comments: