The ECGI blog is kindly supported by
Do Board Chairs Matter? Yes, but Not Always for the Reasons You Think
Traditional research on boards has focused on board independence, assuming that management may pursue its own interests rather than shareholder interests. Independent boards are then a way of mitigating these agency problems. Starting with the 1992 UK Cadbury Code, regulators have likewise advocated more independent boards of directors. Yet most studies do not find a robust and strong link between board independence and firm value or firm performance.
One reason may be endogeneity: board independence is unlikely to be chosen at random. A firm going through a bad patch may come under shareholder pressure to appoint more independent directors, raising the possibility of reverse causality.
Another possibility is that the “independent” directors may be independent in name only. Eighty-seven percent of the boards of Fortune 100 firms have been shown to have a majority of independent directors. When adjusting for social ties between the CEO and board, this percentage drops to 62%.
It could be that who matters most is not the entire board but the chair. Should the chair and CEO positions then be combined or separated? Theory is split on whether combining the two roles, so-called CEO duality, is good or bad. On the one hand, agency theory predicts that the costs from CEO duality are greater than its benefits, leading to CEO entrenchment and weak accountability. On the other hand, resource dependence and stewardship theories argue that CEO duality has net benefits. Firms choose CEO duality – or “unified leadership” – if they need to make quick decisions, such as during crises. Practice also differs widely: while 41% of S&P 500 firms have CEO duality, only five FTSE 150 firms have an executive chair. Similar to board independence, there is little evidence that CEO duality is correlated with higher firm value or performance.
A study I conducted with Peter Limbach and Meik Scholz investigates the reasons why S&P 500 firms combine or separate the two roles. Our analysis of first-time disclosures of these reasons shows that firms with CEO duality provide lengthier and more defensive justifications for their leadership structure. Stock prices did not react to the disclosures made by firms with a separate CEO and chair. While overall the market reacted negatively to the disclosures by firms with CEO duality, the market reaction was positive for small, hi-tech, and R&D intensive firms citing “unified leadership”. This suggests that investors consider that CEO duality is optimal in some contexts.
In ongoing work with Steffen Burkert, Aino Tenhiälä, and Jukka Partanen, we argue that board chairs occupy a central position in corporate governance, as they are the channel through which information flows from the executives to the board. This information is crucial for the board to fulfil its roles. We study the sudden hospitalisations of chairs, which reduce their ability to perform their role. As these hospitalisations are sudden, this gets us around the endogeneity issue mentioned earlier.
We expect the effect on performance to be negative and to be even more negative if the chair and CEO are more like each other. We expect greater similarity in age, education and gender to foster trust between the chair and CEO, leading to mutual understanding and better coordination.
We find that chair hospitalisations have a negative impact on next year’s return on assets, and this negative effect is stronger when the chair and CEO are more similar. This implies that a greater relational fit has a downside, as it makes firms more vulnerable when the chair-CEO relationship is disrupted. The bigger the drop when the chair is absent, the more the relationship was worth.
Still, relational closeness may also create opportunities for abuse. Chairs’ privileged information can create value, but it could also be used for opportunistic insider trading.
Chairs do trade frequently, albeit not as frequently as CEOs. The market reaction to their trades is relatively high, and the value of their trades is also relatively high. In ongoing work with Luc Renneboog and Yang Zhao, we find that directors who sit on more than one board and who are subject to a trading ban in one of their firms are more likely to trade in their other firms. We find a market reaction in the firm that is subject to the trading ban, and it moves with the market reaction in the traded firm. This suggests that information travels across boards during trading bans. Although these trades are not illegal, they are ethically grey.
While traditional research advocates an arm’s length relationship between the CEO and the board or board chair, more recent research suggests that a greater relational fit between the CEO and chair might create value – at least, in certain contexts. However, a greater fit may also create privileged access to information, which creates scope for opportunistic insider trading.
What we understand far less well is when relational closeness creates value and when it becomes a governance risk. Ways forward include measuring relationships more directly – e.g., how chairs and CEOs communicate – rather than inferring them from board structure, and studying settings where the letter of the rules is followed but not their spirit. This leads to two open questions. How can boards become more resilient to the loss of a key relationship? How can we prevent chairs from creating the agency problems they are meant to mitigate?
Marc Goergen is a Professor of Finance at IE Business School in Madrid and an ECGI Research Member.
This blog post is based on a keynote presentation at the 7th Annual RCF-ECGI Corporate Finance and Governance Conference, hosted by Universidad CEU San Pablo in partnership with ECGI, and supported by the British Academy of Management’s Corporate Governance Special Interest Group and the Institute for Humane Studies. Visit the event page to explore more conference-related blogs.
The ECGI does not, consistent with its constitutional purpose, have a view or opinion. If you wish to respond to this article, you can submit a blog article or 'letter to the editor' by clicking here.