In the paper, The Disciplinary Effects of Proxy Contests, which was recently made publicly available on SSRN, I study the effect of potential proxy contests on corporate policies and performance. The agency problem created by separation of ownership and control in publicly traded corporations with dispersed ownership is at the heart of corporate governance literature, which focuses on mechanisms to discipline incumbents. These mechanisms range from informal “jawboning” to contests for corporate control, which are used to change management and to obtain control in publicly traded corporations. Such mechanisms play a disciplinary role if managers are more reluctant to take self-serving actions that increase the probability of shareholders’ intervention (e.g., Grossman and Hart, 1980).
In 1992, the regulatory burdens surrounding proxy fights were substantially liberalized thus reducing the cost of engaging a proxy contest. Specifically, the 1992 proxy reform reduced the costs of the proxy contest by relaxing constraints on communications among shareholders of public corporations (Bradley, Brav, Goldstein, and Jiang, 2010). As a result, the frequency of proxy contests increased significantly after 1992. The average number of proxy contests was 55 (80) per year during 1994-2008 (2006-2008) as compared to an average of 17 a year during 1979-1994. While shareholders more often rely on the proxy contest mechanism, evidence about the effectiveness of proxy contests is limited, since most of the existing literature uses pre-1992 proxy reform data and studies ex post effects only, ignoring the disciplinary effects.
Do proxy contests play a disciplinary role in the post-1992 period? Do shareholders benefit from proxy contests that materialize? These questions have important policy implications for debates on the optimal scope of shareholder rights in public companies (Bebchuk, 2007; Lipton and Savitt, 2007). Whereas a conclusion that the proxy contest mechanism is beneficial provides support for strengthening shareholder rights in proxy contests, the opposite conclusion provides support for constraining them (Bebchuk and Weisbach, 2010). The effort that the SEC, institutional investors, and business groups expended fighting for and against 2010 proxy access reform suggests that effectiveness of the proxy contest mechanism is of central importance.
This paper fills the gap in the literature and studies the effectiveness of the proxy contest mechanism using a manually collected data set of all proxy contests from 1994 to 2008. The paper studies not only companies that have experienced a proxy contest (ex post effects), but also companies that have never experienced a proxy contest (disciplinary effects). Studying both the ex post and the disciplinary effects is crucial (Easterbrook and Fischel, 1981). An analysis that is limited to the materialized events implicitly assumes that incumbents are passive and do not act until a contest materializes. If, however, the disciplinary effects are significant, expectations of potential events will affect corporate decisions.
I begin the analysis by studying the disciplinary effects of proxy contests. To address this important question, I examine whether companies change their financial policies in anticipation of the proxy contest. Specifically, I ask whether a public company changes corporate policies when the likelihood of a proxy contest increases. The evidence suggests that when the likelihood of a proxy contest increases, companies increase leverage, dividends, and CEO turnover. In addition, companies decrease investment in research and development, capital expenditures, and executive compensation. Following these changes, operating performance improves. Overall, the documented evidence suggests that the disciplinary effects of proxy contests mitigate agency costs. The strong disciplinary effects of proxy contests are consistent with both the relatively low frequency of materialized proxy contests (fewer companies end up being targeted because inefficiencies are resolved out-of-equilibrium) and the weak ex post effects on corporate policies (companies change corporate policies in anticipation of the proxy contests).
I study the disciplinary effects of proxy contests by estimating the effect of the likelihood of a proxy contest on corporate policies. This estimation procedure overcomes three issues. First, the likelihood of a proxy contest is a latent variable and therefore must be estimated. Second, the likelihood of a proxy contest is endogenously determined (i.e., it can be correlated with an unobserved component of corporate policies). Finally, the effect of the likelihood of a proxy contest cannot be estimated using the regular two-stage method that accounts for endogeneity because the likelihood of a proxy contest is a latent variable. The estimation procedure developed by Heckman (1978) and Amemiya (1978) addresses these issues. This procedure is applied as follows. First, I estimate a binary choice model (e.g., probit), where the dependent variable is a dummy variable that equals one when the company is targeted in the proxy contest. Next, using estimated coefficients, I construct a consistent estimator of the likelihood of a proxy contest for every public company. Finally, I assess the effect of the estimated likelihood of a proxy contest on the corporate policies. The estimation procedure requires an identification assumption. Specifically, the estimated likelihood of a proxy contest has to be constructed such that it includes at least one covariate that does not directly affect the corporate policies.
Theory suggests that liquid stock markets are generally beneficial for corporate governance. Kyle and Vila (1991), Bolton and von Thadden (1998), and Maug (1998) show that greater liquidity trading facilitates interventions by reducing the cost of acquiring shares in the open market. The general idea is that liquid stock markets make it easier for investors to accumulate large stakes without substantially affecting the stock price. Supporting evidence comes from Collin-Dufresne and Fos (2012), who use micro-level data on trades by activist shareholders and report that activist shareholders significantly benefit from stock liquidity when they purchase a stake in target company. Importantly, Collin-Dufresne and Fos (2012) show that the trading strategy of activist shareholders depends on stock liquidity: activist shareholders purchase shares of targeted companies more aggressively when stock liquidity is high. Consistent with the theoretical insight, empirically, stock liquidity is a significant positive determinant of the likelihood of a proxy contest. Therefore, I make the following identification assumption: stock liquidity satisfies the exclusion restriction and can serve as a source of exogenous variation in the likelihood of a proxy contest.
The final and important concern is that time-varying, unobservable firm-specific effects may confound the inference. For example, the results may be consistent with either a direct effect of liquidity on the outcome variables, as well as an omitted variable that affects stock liquidity and the outcome variables. In order to help rule out such stories, I perform a placebo test, which exploits a period of time when external intervention was almost impossible. The test exploits two changes in the legal environment. First, the cost of a hostile tender offer increases significantly after the second generation of state-level antitakeover laws in late 1980s (Bertrand and Mullainathan, 2003). Second, the 1992 proxy reform reduces the costs of communications among shareholders. As a result, the frequency of proxy contests increases significantly. These two changes suggest that the likelihood of a shareholder’s intervention is lower between late 1980s and 1992.
After I document the disciplinary effects of proxy contests, I study the ex post effects of a proxy contest. First, I study the ex post effects of proxy contests on corporate polices. The evidence suggests that it is hard to detect changes in corporate policy in the post-proxy contest years. However, companies do change corporate policies during the pre-proxy contest years: future targets increase leverage, spend less on R&D and decrease capital expenditures, increase dividend payouts, and decrease CEO compensation. Consistent with the disciplinary effects of proxy contests, future targets and companies that do not end up being targeted change corporate policies in the same direction. Second, I show that companies experience positive and significant stock returns when a proxy contest materializes, without reversals in the long run. Hence, shareholders of ex post targeted companies benefit from a proxy contest. Finally, I show that changes in corporate policies of both targeted and non-targeted companies are associated with stronger operating performance.
This paper contributes to corporate governance literature by showing that the proxy contest mechanism plays a strong disciplinary role. The evidence suggests that companies change corporate policies when the mere threat of a proxy contests increases. The rare proxy contests that actually occur are sufficient to create a threat, which provides companies with monitoring pressure. While it has been documented that hostile takeovers have played a significant disciplinary role, several changes in the regulatory environment and the structure of financial markets have decreased the disciplinary effects of hostile takeovers (Bertrand and Mullainathan, 2003). As this paper shows, consequent changes in the regulatory environment and the structure of financial markets led to a significant increase in the disciplinary effects of proxy contests. That is, there was a substitution between hostile takeovers and proxy contests.
The full paper is available for download here.