The Effect of Minority Veto Rights on Controller Tunneling

Jesse Fried is the Dane Professor of Law at Harvard Law School; Ehud Kamar is Professor of Law at Tel Aviv University Buchmann Faculty of Law; and Yishay Yafeh is Professor at the Hebrew University of Jerusalem School of Business Administration. This post is based on their recent paper. Related research from the Program on Corporate Governance includes Executive Compensation in Controlled Companies by Kobi Kastiel (discussed on the Forum here) and Independent Directors and Controlling Shareholders by Lucian Bebchuk and Assaf Hamdani (discussed on the Forum here).

Most public firms around the world have a controlling shareholder (“controller”). In these firms, a key governance objective is to protect minority shareholders from controller tunneling.

Standard tools—independent director approval for related-party transactions and the duty of loyalty—are often insufficient. Independent directors typically serve at the pleasure of the controller, undermining their objectivity (Bebchuk and Hamdani, 2017). And hurdles to litigation and controller-friendly substantive law tend to erode the effectiveness of the duty of loyalty (Enriques et al., 2017).

A potentially more powerful tool is mandating minority approval for related-party transactions (Goshen 2003; Djankov et al. 2008). This approach, now favored by the OECD, has become the law in Israel, Canada, Australia, Hong Kong, India, Indonesia, Mexico, and Russia. Delaware follows a softer approach: while not requiring minority approval, it applies more deferential judicial review to a related-party transaction in which the controller voluntarily gives veto rights to the minority. However, there is scant empirical evidence on whether any form of minority approval works.

In a paper recently posted on SSRN, The Effect of Minority Veto Rights on Controller Tunneling, we exploit a 2011 Israeli reform to test the efficacy of minority veto rights. The reform gave minority shareholders veto rights over related-party transactions, including the pay of controllers and their relatives serving as officers (“controller executives”). In particular, a controller executive could not receive a pay package without approval by a majority of the minority votes cast in a shareholder meeting (“MoM approval”), and that approval was effective for only three years. Until 2011, a controller-executive pay package required the approval of only a third of the minority (“ToM approval”), and that approval held indefinitely. The reform did not alter the approval mechanisms for the pay of executives unrelated to controllers (“non-controller executives”).

We examine the reform’s effect on controller-executive pay by hand-collecting and analyzing data on thousands of executives, some related to controllers and others not, over an eight-year period around the reform.  We find that the reform is associated with an average decline of 13%–17% in controller-executive pay relative to the expected pay of the same executive based on firm size, profitability and other factors. We also find that this decline is due, at least in part, to a substantial increase in the frequency of pay cuts for controller executives.

We also examine the reform’s effect on the disappearance rate of controller executives from the list of the firm’s highest paid executives—due to such executives resigning or continuing to work for limited (or no) pay. We find that the relative likelihood of controller executives disappearing increased more than 50%, often in circumstances indicating that MoM approval would be unobtainable. Because we can measure only the pay of controller executives remaining on the firm’s list of highest-paid executives, our estimates of the magnitude of the reform’s effect on controller-executive pay are downward-biased.  Interestingly, non-controller executives replaced about 10% of these disappearing controller executives, suggesting that the reform affected not only pay levels but also the composition of certain management teams.

Our findings relate to several strands of the corporate governance literature. First, we shed light on the efficacy of minority veto rights. Most prior research concerns controllers who voluntarily grant the minority veto rights in Delaware freezeouts to limit judicial scrutiny of the deal (Subramanian, 2007; Restrepo, 2013; Restrepo and Subramanian, 2015). However, the timing of the deal and the decision to grant the minority veto rights are endogenous, and the grant of minority veto rights changes the legal treatment of the deal. In contrast, our study examines a largely exogenous event. Moreover, the availability of pay data for non-controller executives, combined with executive fixed effects, enables us to construct robust controls.

Second, we contribute to the study of pay tunneling. While pay may not be the most lucrative channel for tunneling, controllers and their relatives often work for the firm, and it is accepted that they be paid. Prior work has therefore sought to estimate pay tunneling by comparing the pay of controller executives to that of non-controller executives. This work has found pay tunneling in some jurisdictions, but not in others, and could not completely rule out the possibility that controller executives receive higher pay because they serve in higher positions.

Our setting enables us to measure pay tunneling directly by examining the effect on pay of a reform that introduced, in midstream, real bargaining over pay between the minority and the controller. Before the reform, there had been some constraint on pay, as controller executives needed ToM approval to obtain pay increases. However, the reform tightened the constraint by requiring all controller executives to obtain MoM approval every three years to continue receiving any pay. Our findings that the reform lowered relative pay levels and triggered pay cuts and turnover suggest that the minority perceived controller executives to be overpaid, and used its new veto rights to constrain their pay.

The complete paper is available here.

Both comments and trackbacks are currently closed.