Ownership Structure, Voting, and Risk

In our paper Ownership Structure, Voting and Risk, forthcoming in the Review of Financial Studies, we investigate the interaction between the ownership structure of publicly traded firms and their risk profiles. In particular, we show how the potential for conflict of interest between shareholders on risk decisions may cause the emergence of activist mid-sized investors. In turn, ownership structure affects the risk decisions that firms make.

It is natural to believe that the choice of shares to hold in a company is a trade off between diversification and control: large size comes with control at the cost of diversification. Many firms, however, have mid-sized shareholders who are neither well diversified nor have control. For example, in the United States (where it is widely agreed that regulation helps dispersed ownership), 67% of public firms have more than one shareholder with a stake larger than 5%, while only 13% are widely held and 20% have only one blockholder (Dlugosz et al., 2006). In Europe (where concentrated ownership is the norm), in eight out of the nine largest stock markets of the European Union, the median size of the second largest voting block in large publicly listed companies exceeds five percent (data from the European Corporate Governance Network). Why do such mid-sized shareholders emerge?

Our paper offers a novel rationale for the existence of multiple mid-sized blockholders: they mitigate the conflicts of interest on risk between the largest blockholder and dispersed shareholders. To elaborate on this idea, suppose there is an initial owner of a firm who wants to issue shares to raise capital for an investment project. Suppose too that the initial owner cannot commit to taking a value-enhancing action in the future (e.g., monitoring the manager), unless he holds a sufficiently large stake. The owner and the other investors are risk averse. Outside investors demand shares conditional on the offer price set by the initial owner. Once the ownership structure is established, shareholders vote on the riskiness of the project that the firm subsequently undertakes. The investment technology is such that higher returns can only be achieved by increasing risk. Investors’ preferences over the available risk/return profiles depend endogenously on their chosen shareholding. The higher the ownership stake is, the lower the preferred risk exposure is, creating an endogenous conflict of interest between the initial owner, who must hold a large block, and the dispersed shareholders who have no such constraints.

We assume, realistically, that only a fraction of the small shareholders are active, in that they actually vote. In the United States, for example, voter turnout of shareholders is between 70% and 80%. The (partial) abstention from voting by small shareholders potentially drives a wedge between a shareholder’s cash flow rights and the power in decision-making. For example, if the initial owner is the only blockholder, he might have a minority of cash flow rights but a majority of votes cast. Dispersed shareholders thus might anticipate that the future investment decision will not be in their favor, which reduces their willingness to pay for the shares. Since the total fraction of shares they hold may be large, this reduction in share price hurts the initial owner.

The latter may then benefit from the emergence of another smaller blockholder, who would be pivotal and indirectly serve as a commitment device for the initial owner to shift the choice of risk toward an outcome that is more favorable to dispersed shareholders. Paradoxically, of course, as an additional shareholder acquires a larger stake, his preferences move closer to those of the initial owner, who holds a large stake. The additional blockholder thus allows the initial owner to sell shares at a higher price, without going as far as allowing small shareholders to implement their preferred choice of risk. Note, however, that because the benefit of choosing the project is outweighed by the cost of being poorly diversified, the mid-sized blockholder, in spite of being pivotal, benefits less from his shareholding than dispersed shareholders do. In this sense a large shareholder provides a public good to dispersed shareholders.

The model identifies a set of factors that affect ownership structure, and, through it, a firm’s risk profile. First, the degree of moral hazard linked to the value-enhancing action taken by the largest blockholder affects the ownership structure. Even though the value-enhancing action does not affect risk directly, it affects the size of the largest block and hence the firm’s risk preferences. If the moral-hazard problem is severe, the initial owner must retain a large stake, increasing the subsequent conflict of interest with dispersed shareholders. Conversely, when conflicts of interest are small (the moral-hazard problem is mild), a dispersed ownership structure may arise, in which the initial owner is the only blockholder, but he has a small stake and so will be outvoted by dispersed shareholders, who thus will be willing to pay a high price for their shares.

In particular, our theory predicts that for medium sizes of the largest blockholder, at least one other blockholder would emerge. Moreover, the size of the second-largest block increases in the size of the largest block (to counterbalance the voting power) and decreases in the number of blockholders. These predictions are confirmed empirically by Carlin and Mayer (2000).

Second, firms operating in very risky sectors can potentially suffer large conflicts of interest between the largest blockholder and small shareholders. Our model predicts multiple-blockholder ownership structures in such firms, whereas firms operating in sectors that are more mature should tend to have only one large shareholder with a fringe of small shareholders.

Third, regulation determining voting rules and possibilities of proxy voting affects the firm’s ownership structure. In the United States for example, broker non-votes and absenteeism are treated differently across states: In California, shares that are not voted are counted as voting with management, whereas they are counted as abstentions in Delaware (GMI Ratings for CalPERS 2013). Our model predicts that, all else equal, economic systems that thwart voting by minority shareholders should end up with firms that have concentrated ownership. This happens because a large initial owner can retain an outright majority of votes cast with a smaller fraction of shares owned, obviating the need to introduce a second blockholder. Conversely, ownership structures where effective control is maintained with dispersed shareholders are harder to sustain, as they require a larger fraction of total share ownership to be in the hands of small, dispersed shareholders. We are not aware of empirical work that directly tests for such a relationship but it seems worth exploring. There is, however, indirect evidence from the comparison of ownership structures in the United Kingdom and the United States. Becht et al. (2009) find higher risk taking and higher value creation in the U.K., and link it to U.K. pension fund activism. The U.K. legal system favors smaller but significant ownership stakes that can be thought of as mid-sized blocks.

Our model has implications for how decisions on risk are affected by ownership structure: in particular, decisions are effectively determined by the shareholder who is pivotal, and that may not be the largest shareholder but could be a mid-sized blockholder. In the model, multiple large shareholders are associated with higher risk. These predictions are consistent with the findings of several papers in the empirical literature, which show that firms with less-concentrated ownership invest in higher-risk projects, such as R&D and skill-intensive activities. Our model also predicts that multiple large shareholders choosing higher risk, should be positively related to higher firm value, whereas a single large shareholder is associated with lower firm value. Thus, mid-sized blockholders play a role in mitigating the conflicts of interests between the largest shareholder, who prefers to reduce risk at the expense of value and diversified shareholders who are value maximizing. Consistent with our theory, several papers find that firms with multiple blockholders have higher market value (e.g., Lehmann and Weigand, 2000; Volpin, 2002).

The full paper is available for download here.