3.5.1 Empty Voting

In the previous section we assumed that votes are bought by agreeing with another shareholder that he or she will vote as instructed. But in modern financial markets, votes can effectively also be bought through a range of other techniques including borrowing shares (stock lending), reducing economic exposure through derivatives (hedging), and buying shares prior to the voting record date and selling them immediately afterwards (record date capture).1Vote buying has thus become relatively easy, and seems to be occurring more often.2 To assess the impact of the new vote buying (or empty voting as it is referred to) on voting efficiency, it is again useful to look at the stock market. There, we find a cousin to empty voting: short selling. Both are strategies that can be used to leverage superior information. Both are facilitated by derivatives and securities lending. And both are commonly denounced by policymakers even if they carry the potential to enhance efficiency.

The potential of short selling to enhance efficiency lies in the fact that it can help to quickly incorporate new information into share prices. Empirical evidence suggests that short selling actually fulfills this role.3 Policymakers have nevertheless long held a negative view of short selling.4 Accordingly, they have responded to the recent financial crisis by imposing hans on short selling. The hans were largely driven by the concern that traders might seek to make a quick profit by selling short and driving the share price down either by increasing supply (and creating a negative feedback loop) or, less subtly, by spreading false stortes. Theoretical as well as empirical studies of short-selling hans, however, suggest that such hans slow down price discovery.5 This suggests that we should be skeptical of limits on short selling.

By analogy, an analysis of empty voting should focus on its potential to enhance efficiency. Empty voting realizes this potential when it enables a shareholder with superior information to obtain greater voting power, thereby increasing the probability that a majority of the shares is voted in favor of the correct option. The potential of the new vote buying to enhance efficiency is thus similar to that of conventional vote buying. Scholars such as Susan Christoffersen, who examine stock lending activity recognize this and note that "[s]ince the dispersion of information can be a poor match to the dispersion of shareholdings, vote trading can improve the aggregation of this information."6 Consistent with this hypothesis, they document increased lending activity around voting record dates, as illustrated in Figure 3.

As with other arbitrage strategies, we can explore the limits of empty voting as an arbitrage strategy by identifying cost constraints and legal constraints. The costs will largely depend on the consideration that needs to be paid to the lender of the shares (in the case of stock lending), the counterparty to the derivative transaction (in the case of hedging), or, in the case of record date capture, the interest that needs to be paid when borrowing funds to temporarily purchase shares in the market. As a general matter, these costs do not seem prohibitive. Indeed, they may be lower than the cost of conventional vote buying.8 Moreover, finding a counterparty who is willing to lend money or shares or to take the long side in a derivative transaction will probably be easier than finding a shareholder willing to accept voting instructions against a payment.

Moving to legal constraints, we have already seen that courts are suspicious of conventional vote buying because it is susceptible of abuse. This suspicion is also warranted with respect to the new vote buying. In the extreme case where a shareholder uses derivatives to build a net short position, his interests clearly conflict with those of other shareholders, as he will prefer an outcome (share price decrease) that is the opposite from that preferred by other shareholders (share price increase).9 The conflicted shareholder will, to use our terminology, vote insincerely.

Interestingly, the Delaware Chancery Court recently addressed the question of whether the concept of vote buying, as developed by the courts, is broad enough to encompass the new vote buying. The answer is yes. When these techniques prove deleterious to stockholder voting, the court "can and should provide a remedy."10 In the case at hand, the court found that the voting buying was not a legal wrong, because the shareholder did not have any competing economic or personal interests that might have created an overall negative economic ownership.11 So again, it appears that a shareholder with superior information who engages in vote buying—this time, the new vote buying—to ensure that the correct option is chosen might survive judicial scrutiny, but still faces a litigation risk.

Contrary to the courts, policymakers have been slow to recognize that empty voting need not be deleterious per se, resulting in additional legal constraints. While ownership disclosure rules have not yet been updated across the board, they have been in a number of jurisdictions and will soon be in others.12 There are sound reasons to do so, both from a point of view of market efficiency and of corporate governance.13 Indeed, the benefits of increased transparency may well outweigh the costs deriving from imposing legal constraints on empty voting as an arbitrage mechanism.14 In respect of further-reaching measures aimed at curbing abuse, however, the cost-benefit analysis is ambiguous.

Henry Hu and Bernard Black propose several measures that would have the effect of restricting not only the possibility to engage in vote buying for abusive purposes, but also the possibility to use vote buying as an arbitrage strategy.15This suggests that even though they acknowledge that the net efficiency of the new vote buying is uncertain, they tend to focus on the risk that votes are bought for abusive purposes.16 However, it is far from clear from Hu and Black's overview of real world examples that in those cases empty voting led to an inefficient outcome, and indeed they make no such claim. So until the contrary is proven, we must at least be open to the possibility that the new vote buying is used more often for benign purposes than for abusive purposes.

At this point, it is useful to revisit the analogy with short selling. The findings of a recent study on short selling and the news suggest that traders who sell short and then spread false news play a significant role.17 Conceptually, this is equivalent to shareholders who engage in vote buying with the purpose of promoting a majority vote for the option that fails to maximize shareholder value. In both cases, the arbitrageur's behavior is not driven by the possession of superior information, and consequently his acts will not reduce information asymmetry. But the findings of the study also suggest the importance of traders who, by collecting and analyzing publicly available data, detect that an issuer 's share price exceeds its fundamental value, sell short, and then truthfully spread their conclusions.18 This is the functional equivalent of shareholders who engage in vote buying with the purpose of promoting a majority vote for the option that does maximize shareholder value. In both cases, the arbitrageur's behavior is driven by the possession of superior information, and his acts will reduce information asymmetry.

While research on short selling and the news is still in an early stage, the results so far offer no reason to assume that short sellers who spread false news are more prevalent than short sellers who spread true news.19 If there is no reason to assume that there is more abusive short selling than beneficial short selling, why should we assume that there is more abusive vote buying than beneficial vote buying? It's not obvious why we should, especially since the usual suspects who engage in short selling, hedge funds, are one and the same as those who are usually suspected of engaging in the new vote buying.20

The SEC has recently issued a draft release through which it seeks to obtain insight into how empty voting should be regulated.21 The preceding analysis suggests that, because empty voting can be used for both beneficial and abusive purposes, it generally makes sense to battle abusive empty voting through narrow ex post rules rather than through broad ex ante prohibitions of empty voting. This observation completes the analogy with short selling. Finance scholars are near unanimous in their disapproval of short selling hans, citing the unintended consequence of disabling the salutary effect of short sales increasing information efficiency.22 The new vote buying should also not be illegal per se. Rather, if it is established after the fact that a shareholder engaged in empty voting not to leverage superior information but to further his private interests by profiting from a majority decision that fails to maximize shareholder value, courts should intervene, just as regulators will intervene if it is established after the fact that a trader engaged in short selling not to leverage superior information but to further his private interests by profiting from the market's response to false news.23

To enable ex post scrutiny, transparency is key. To begin, a disclosure obligation discourages empty voting driven by insincere motives by increasing the risk of detection. Moreover, disclosure enables the market and the firm to detect actual abuse and commence litigation if need be.24 In securities markets, to enable detection of abusive short selling it suffices to require disclosure to the regulator only. This way, the profit potential from short selling is not unduly restricted and incentives to search for fundamental information are preserved. But notice that in principle, a shareholder who engages in empty voting with sincere purposes needn't be reluctant to disclose his increased voting power to the market. On the contrary, if such shareholder holds a significant economie stake, public disclosure sends a credible signal to other shareholders that the shareholder has superior information and thus offers a means to further leverage that information. In this sense, disclosure of empty voting positions may increase voting efficiency in the same way as disclosure of short selling may increase market efficiency, even if it induces a risk of herding behavior.25