Optimal Corporate Governance in the Presence of an Activist Investor

In our paper, Optimal Corporate Governance in the Presence of an Activist Investor, forthcoming in the Review of Financial Studies, we provide a model of governance in which a board arbitrates between an activist investor and a manager facing reputational concerns. Shareholder activism to force policy changes at publicly-traded firms represents an increasingly important dimension of the market for corporate control. While activist investors represent a source of corporate governance that is external to a firm’s power structure, they differ dramatically from the corporate raiders that are the focus of earlier theories of external governance. In most cases, activist investors accumulate relatively small stakes and so cannot exert direct control. Rather, they must rely on persuasion and the firm’s internal governance mechanisms to implement changes. As Brav et al. (2008) show, activist hedge funds are often successful in influencing managers and boards, and their efforts have a substantial impact on firm value.

How does the presence of such an external governance force affect internal governance policy? To address this question, we analyze a model in which a board, recognizing that an activist shareholder may exert discipline on a manager, chooses an appropriate level of internal governance. To our knowledge, this is the first theoretical article to consider the interaction between an activist, the board, and a manager. We argue that the possibility of disputes between an activist and management creates a natural but novel role for the board of directors that has not been considered previously: It functions as an arbitrator between different stakeholders who wish to take the firm in different directions.

Disputes in our model arise because the manager faces reputational concerns that make him reluctant to reverse strategic decisions he has made in the past. As a result, he may ignore an activist’s efforts to implement strategic change at the firm, even if he believes such change would increase firm value. The first contribution of our article is to show that, while more stringent internal governance mitigates the effects of this agency conflict ex post, it can exacerbate the manager’s reputational concerns ex ante. As a result, it can worsen the very agency conflict that it is intended to solve. Our second contribution is to show that, unlike in standard governance models, internal governance and external governance provided by an activist are natural complements, and only become substitutes when external governance is relatively weak. We develop testable predictions from both theoretical results, and also propose an additional test based on the comparative statics of activist entry.

In our model, a firm chooses between two mutually-exclusive projects with uncertain payoffs. The manager obtains a noisy signal about the relative payoffs of the projects, the precision of which depends on whether the manager has high or low ability. Based on this signal, the manager embarks on one of the projects. Then, an outside investor decides whether to generate costly information about the projects. If acquired, the precision of the outsider’s signal lies between that of the high- and low-ability managers. The outsider is an activist in the sense that she only realizes a profit if she can induce the firm to implement value-creating change.

If the information of the outsider conflicts with that of the manager, the manager can choose to concede (switch to the other project) or fight (continue with his original project). The efficient outcome is to fight if his ability is high and concede if his ability is low. However, the manager cares both about firm value and his own short-run reputation, which depends on investors’ posterior beliefs over his type before the project pays off. Thus, the low-type manager has an incentive to mimic the high type by fighting.

The board plays two roles in our model. At the outset, it determines a level of ongoing screening over the manager. At a later stage, its screening technology yields information about the type of the manager. This is the only information that the board generates itself. If the activist pushes for change and the manager fights, the board has final authority over the choice of project; in particular, it either upholds or overrules the manager. The board’s authority results in a certification effect—a manager who fights and is upheld by the board is more likely to be a high type. The certification effect provides the insight behind our first contribution: Better governance sometimes increases the manager’s tendency to fight.

We find that, under some circumstances, more intensive governance by the board can worsen the same agency problem it is intended to correct. This differs from existing articles that identify negative consequences of stronger governance, which conclude that attempting to solve one agency conflict can exacerbate other frictions. For example, Burkart, Gromb, and Panunzi (1997) show that intervention by a board to limit a manager’s private benefits can cause the manager to under-invest in firm-specific human capital. Adams and Ferreira (2007) show that it can cause the manager to under-supply information to the board. In Almazan and Suarez (2003), underinvestment in human capital is caused by the board’s efforts to weed out low quality managers.

Our theory builds on previous work demonstrating that agents facing reputational concerns are reluctant to implement changes. Managers concerned about their reputations may act as if sunk costs matter (Kanodia, Bushman, and Dickhaut 1989) and are reluctant to deviate from their earlier investment levels when additional information becomes available (Prendergast and Stole 1996). Boot (1992) argues that a manager who has purchased assets will not divest often enough because of the reputational cost of reversing prior investment decisions. However, the threat of a hostile takeover can mitigate the agency conflict and induce divestitures.

Since both internal and external governance in our model discipline managers, one may expect them to be natural substitutes (e.g., Fama 1980; Fama and Jensen 1983; and Williamson 1983). However, the two are in fact natural complements—if the activist’s information improves, it is more valuable to the board to identify and overrule a low-type manager, so the board invests a greater amount in its screening technology. Indeed, they function as substitutes only when the board departs from the cash-flow maximizing level of governance: If the activist’s information is noisy, the board over-invests in internal governance to induce the activist to enter. As the precision of the activist’s signal improves, a smaller over-investment is needed, so the level of internal governance falls.

Our model generates several empirical predictions. First, consider an exogenous shock to the net cost of activism, either in absolute terms or relative to other forms of external governance. For example, state business combination laws in the 1980s decreased the cost of activism relative to hostile takeovers, and decimalization reduced the cost to the activist of acquiring an initial stake in a firm. We predict that an exogenous decrease (increase) in the cost of activism leads to increased (decreased) activism, increased (decreased) board vigilance, and more (fewer) disputes between managers and activists when managers have high reputational concerns or when the signals that outsiders can generate about optimal firm strategy are noisy.

Second, various policies on internal governance may be interpreted as requiring a minimal level of board vigilance. For example, the 2002 Sarbanes-OxleyAct increased director accountability, which induces the board to be more active. We predict that when managers have moderate reputational concerns, such an increase will make low-ability managers more intransigent and lead to more disputes between managers and activists. However, managers with high or low reputational concerns will be unaffected.

Third, our results on the relationship between internal and external governance imply that when managers have high reputational concerns, the board’s vigilance will decrease with the skill or industry experience of activists to generate information when this skill is low and increase with it when this skill is high. Testing this prediction may help resolve conflicting results in the literature on whether internal and external governance are substitutes or complements.

The full paper is available for download here.