The Shareholder’s Dilemma

Theoretical and empirical research on shareholder voting has provided many exciting insights and guidance for debates on policy and regulation. The default assumption, though, is that shareholders have strong incentives to vote for alternatives that they think are best for the firm.   In our paper, available here, we show that, when shareholders have the opportunity to buy or sell shares after voting, there can be strong incentives to vote against the firm’s interest. The presence of this incentive confounds the potential for shareholder voting to aggregate information and serve as a check on management.

We develop a game-theoretic model of shareholder voting that captures the fact that shareholders are traders/investors and assumes that they ultimately seek to maximize the return on their investments. By embedding a model of shareholder voting in a model of market transactions, we are able to better understand the nature of voting and its relationship to the market. The presence of liquidity generates a shareholder dilemma: Voting for the choice that a shareholder thinks is optimal for the firm maximizes her return only if her vote is pivotal to the total vote’s outcome; in all other instances, it is better to protect her informational advantage over the market and vote against her information, moving the share price, if only a little, and then capitalizing on her informational advantage by buying or selling shares. The shareholder dilemma requires that the equilibrium informativeness of voting may be quite low. As the number of shareholders gets large, the votes become essentially uncorrelated with private information.  Moreover, in all equilibria, shareholders will have strong incentives to trade after they vote, and heterogeneity in which side of the market they are on will depend on their private information. The model does support very simple equilibria in which no information is contained in the votes and shareholders act as a rubber stamp on the questions that are brought to them.

A pervasive feature of corporate governance is that shareholders are provided with regular opportunities to shape the choices made by the firms that they invest in.  This tendency is not only a norm but a requirement codified  by regulations.  Moreover, if anything, the trend in policy-making is to expand the authority of shareholders to influence the firm in sometimes important decisions.  Reflexively, we might expect that (1) shareholders are motivated to see firm decisions congruent with maximizing share-value, (2) vote counts efficiently aggregate the information held by investors so as to maximize the chances that the profit-maximizing decisions are reached, and (3) there is a strong relationship between shareholder votes and market assessments of the firm. There is, however, an emerging consensus that the second and third points don’t pan out in the empirical record.

Although a budding theoretical literature addresses important questions about the effects of different institutional structures on the efficiency of voting in the shareholder context, and another set of papers is starting to frame alternatives to this informational theory in order to reconcile empirical patterns, there is, to our knowledge, an important conceptual gap.  The starting point for work that treats voting as informational is the assumption that shareholders seek to maximize the profitability of the firm.  If investors faced illiquidity, this assumption would be natural.  But, when the shareholder is free to change her status by buying or selling shares, it does not follow that maximizing the profitability of the firm coincides with maximizing the investors’ returns.

We present a theoretical account of shareholder voting as a means to potentially aggregate private information and find the presence of a sharp incentive problem.  The desirability of buying or selling shares depends on the price, and thus shareholders facing liquidity will care not just about influencing firm decisions (which affect the value of the shares they hold), they will care about influencing market prices (which affect the returns they can get from selling some of their shares or the cost of buying additional shares).  To flesh out the incentives faced by shareholder voters in the presence of liquidity, we develop a simple model of voting and trading.   We find that voting for the option that the shareholder believes to be optimal for the firm is only optimal for her if she turns out to be the pivotal voter (that is, the vote is nearly a tie).  In all other situations, her payoffs are maximized by casting a vote for the option she believes to be worse for the firm and capitalizing on the informational advantage she has over the market.  By voting against the policy that is ultimately chosen, she may lower the price, and by voting for the policy that is ultimately chosen, she may increase the price. She then trades at this distorted price.

Equilibrium then requires that voting must be sufficiently noisy relative to how strong shareholders’ private signals are so as to balance two factors: (i) how the market reacts to votes and (ii) the incentive to select the policy a shareholder thinks is best for the firm.  Thus, equilibrium will be consistent with the idea that there is not too much information contained in voting even if shareholders, in aggregate, possess a lot of relevant information.

Moreover, a central feature of the model is that shareholders will use their information, the policy chosen, and the market price adjustments based on vote counts to adjust their portfolio after voting.  Because votes are noisy signals of shareholders’ private information, shareholders have incentives to trade.  Shareholders that received private information supporting the policy chosen will believe the new market price is too low, and shareholders that receive private information contrary to the policy chosen will believe the market price is too high. This feature of the equilibrium offers a connection between informational models and recent empirical findings in Li, Maug and Schwartz-Ziv (2019). They interpret post vote heterogeneity in the trading behavior of shareholders as evidence against information aggregation models, and in support of opinion models. We find that this heterogeneity is a necessary feature of any equilibrium in an informational model where traders are allowed to trade after voting. The key here is that neither selling nor buying is in fact only optimal in close cases, and so unless there is complete convergence of prior beliefs a model that allows for post-vote trading will actually have post-vote trading in equilibrium.

Our work also provides some insights on the regulation of corporate voting. In 2003, the Securities and Exchange Commission required management investment companies to disclose proxy votes in the hope these funds, typically the largest voting bloc, would actively monitor management and improve corporate governance through voting. However, our work shows that it is not a foregone conclusion that mandatory voting will drive these large actors to cast ballots in ways that serve the remaining shareholders.  SEC rules can only push funds to vote, but they cannot control how they vote. Our model yields two types of equilibria.  In one, shareholder votes are very weakly correlated with the action the shareholders think is best and thus voting is a very week control on management. In the other, votes are consensual and unrelated to what the shareholders actually believe.   In both types of equilibria, voting is not capable of serving as a strong control on management.  Yet these patterns seem to pan out in the data.  Among 2025 companies in the Russell 3000 that had an annual meeting and say-on-pay vote in the calendar year 2012, 1,466 companies (72 percent) gained approval from over 90 percent of shareholders, and 381 companies (19 percent) gained approval from between 70 percent and 90 percent of shareholders, while only 53 companies (2.6 percent) failed to gain shareholder approval[1]. This snapshot suggests that we might be seeing voting that is not closely related to firm fundamentals.

Even though our findings are driven by the potential for shareholders to achieve informational rents from voting against the firm’s interest, in the equilibria in which voting is consensual and non-responsive, they won’t see these opportunities in equilibrium.  This is analogous to the logic behind no-arbitrage pricing.  We expect market prices to tend to be “correct” because, were there opportunities for arbitrage, opportunistic traders would take advantage of them. Our argument is based on the idea that, were voting informative, opportunistic shareholders would take advantage of the opportunities to capitalize on informational rents through post-vote trading.

A second area of research on shareholder voting that speaks to regulation pertains to opportunities for shareholders to gain additional votes beyond their number of shares.  Preventing this type of behavior does nothing to temper the incentive problem and predicted inefficiencies from our analysis.  The analysis in our paper does, however, provide some guidance for policymakers.  But rather than congruence of shares and votes the policy lever is liquidity, or at least a form of it.

The theoretical literature makes conflicting predictions about the impact of liquidity on corporate governance. Coffee (1991) and Bhide (1993) argue that stock market liquidity impairs governance. Maug (1998) argues that liquidity makes corporate governance more effective. Admati and Pfleiderer (2009), Edmans (2009), and Edmans and Manso (2011) show investors exiting by liquidating their shares, and this is a governance mechanism in itself.  The empirical literature around this question also yields mixed findings. Edmans, Fang, and Zur(2013) demonstrate that stock liquidity has a positive effect on blockholder governance. But Back, Li and Ljungqvist (2015) show that greater liquidity negatively affects governance on average. In this work, the focus is on the direct result of selling and not voting.  None of these papers consider the potential for incentives created by liquidity to distort voting and thus affect governance.  Our paper finds that the ability of shareholders to change their portfolios shortly after voting weakens their incentives to vote informatively and thus reduces the informativeness of corporate voting. In the equilibria where voting is consensual we see a severe breakdown of shareholder oversight.

The analysis, therefore provides guidance on the potential harm from a particular form of liquidity.  One way to make voting more effective may be to increase liquidation costs for all or just large shareholders in some period following important votes.

To conclude, shareholders do not have an absolute incentive to maximize the value of the firm that they have a stake in. Starting instead with the assumption that shareholders as investors seek to maximize their return, the possibility of trading introduces a potential wedge. In settings where shareholders can possess private information, incentives to use this information in their trading behavior and not reveal it when voting may lead to distortions in corporate policy-making. Equilibrium forces must balance out these incentives, and accounting for this provides an explanation for fairly uninformative voting even when shareholders have access to high-quality information. Moreover, because differences in private information may remain present at interim stages of the model, shareholders may be seen to behave heterogeneously in the market.  Accounting for this potential incentive problem may help guide future research as well as policymakers.


[1] The statistics are from Semler Brossy 2012 Say on Pay Results: Russell 3000 Shareholder Voting.


  1. Admati, Anat R., and Paul Pfleiderer. “The “Wall Street Walk” and shareholder activism: Exit as a form of voice.”The Review of Financial Studies 7 (2009): 2645-2685.
  2. Back, Kerry, Tao Li, and Alexander Ljungqvist. “Liquidity and governance.”ECGI-Finance Working Paper 388 (2015).
  3. Bhide, Amar. “The hidden costs of stock market liquidity.”Journal of financial economics 1 (1993): 31-51.
  4. Coffee, John C. “Liquidity versus control: The institutional investor as corporate monitor.”Columbia law review 6 (1991): 1277-1368.
  5. Edmans, Alex. “Blockholder trading, market efficiency, and managerial myopia.”The Journal of Finance 6 (2009): 2481-2513.
  6. Edmans, Alex, and Gustavo Manso. “Governance through trading and intervention: A theory of multiple blockholders.”The Review of Financial Studies 7 (2010): 2395-2428.
  7. Edmans, Alex, Vivian W. Fang, and Emanuel Zur. “The effect of liquidity on governance.”The Review of Financial Studies6 (2013): 1443-1482.
  8. Li, Sophia Zhengzi, Ernst G. Maug, and Miriam Schwartz-Ziv. “When Shareholders Disagree: Trading After Shareholder Meetings.”European Corporate Governance Institute (ECGI)-Finance Working Paper 594 (2019).
  9. Maug, Ernst. “Large shareholders as monitors: Is there a trade‐off between liquidity and control?.”The journal of finance1 (1998): 65-98.
  10. Semler Brossy, 2012 Say on Pay Results: Russell 3000 Shareholder Voting.

This post comes to us from Professor Adam Meirowitz at the University of Utah’s David Eccles School of Business and Shaoting Pi a PhD candidate at the school. It is based on their recent paper, “The Shareholder’s Dilemma: Voting and Trading,” available here.