INTRODUCTION
In this article, we argue that these earlier models present a far too simple picture of the world. This defense would force an acquirer to win two consecutive corporate elections to gain control of the target.
TAKEOVER TECHNIQUES AND DEFENSES: THE BASICS
The poison pill stops a hostile bid because it threatens the acquirer with massive dilution of its stake in the target if it buys a controlling share of the target's stock without the support of the board of directors. Although the Delaware Business Combination Antitakeover Statute29 is not of the controlling shareholder type, it also contains provisions that may require a shareholder vote on whether the bidder's shares should have voting rights.
AN OVERVIEW OF EARLIER MODELS
Schwartz, supra note 3, at 10 (assuming as part of their argument that "target shareholders . . . make an informed choice in the proxy because they can assess the economic variables that affect the desirability of the underlying tender offer"). In summary, previous models make unrealistic assumptions about the homogeneity of shareholders, both in terms of the size of their holdings and their voting behavior.
OUR MODEL OF CORPORATE VOTING
Thus, the choice of the function r(1,.3,x) would indicate a shareholder whose vote will be highly correlated with the signal, but has an underlying bias in favor of management. Finally, we should briefly discuss how sensitive the model is to the choice of exogenous parameters.
APPLICATIONS OF THE M ODEL
Even if this value is reduced to 10 for both groups of investors, the calculated probabilities vary by 10 percent on average. If both values of k are reduced to 5, the resulting probabilities can vary by as much as 30 percent. Variations of any of these values by an amount of 0.2 result in a variation in the calculated probabilities of approximately 25 percent.
For technical reasons, we have ensured that the shareholders cannot be divided into two groups each holding exactly half of the shares. 46 With that exception, it follows from the discrete nature of the model that small changes in the distribution will not affect the calculated probabilities.
Ownership Structure Assumptions
In this scenario, management joins the company's founding family in opposing a hostile takeover.48 The bidder has built up a small stake in the company and could buy more shares without exceeding the poison pill threshold level of 15 percent . Institutional investors are believed to hold the overwhelming majority of the remaining shares, while individual shareholders (Public) own only 5 percent of the shares. In this scenario, the target management owns a very small percentage of the company's stock, but the company has an ESOP and/or has sold a block of stock to a friendly third party: a white squire.49 This may be the case with target companies with strong protection against takeovers, but low management share ownership.50 The dissident shareholder has purchased as many shares as possible given the 15 percent trigger of the target company's poison pill.
As in scenario A, institutional investors own the vast majority of the remaining shares, while individual shareholders own only 6 percent of the shares. For example, if management owned less than 1 percent of the target's stock, but placed another 14 percent with a white squire or ESOP, and the target company had a poison pill with a 15 percent trigger level, then this scenario would be closely match the situation. duplicated. Again, as with most large publicly traded companies, institutional investors own the majority of the target company's shares, along with individual investors.
We use the same distribution of institutions: Institution I (follows ISS's recommendations), Institution II (favors management in general but is influenced by ISS's recommendation) and Institution III (more opposed to management in general takeover situations, but still influenced by ISS recommendations). The remaining actors, Institution II and Institution III, will vote differently depending on the nature of the contest, although in each case their vote will be determined by the function r(k,t,x), as described earlier.
Types of Voting Contests
Routine Contests
The remaining shares are distributed in accordance with the previous distributions to the three groups of institutional investors and the individual investors. After describing the distribution of shares, we move on to describe how the different shareholders vote in different kinds of contests. We begin by noting that Management will always vote for Management with probability 1, Dissident will always vote for Management with probability 0, and Public will always vote for Management with probability 0.5.
Moreover, Institution I will vote according to the signal sent by the advisory agent ISS, that is, its vote will be drawn from the distribution f(x) if it receives a For signal and from g(x) if it receives a Against signal received. In other words, Institution I consists of those institutional investors who regularly follow the advice of their advisory agent and who automatically vote their shares in accordance with this signal. Previous research has shown that shareholder votes in these routine votes tend to strongly favor management's position, and that institutional investors will vote heavily against management only in unusual circumstances.54 ISS generally recommends in favor of stock option plans and on shareholder proposals regarding internal corporate governance issues.
In this situation, both Setting II and Setting III will vote according to the function r(20,.5,x). The choice of this function reflects a strong bias towards management (t=.5) and little sensitivity to the signal sent from the advising agent (k=20).
Proxy Contests for Corporate Control
Arbitrages are unlikely to get involved in this situation as there is no bidder making a premium offer to buy the target's shares. However, the presence of a distributed board will force dissidents to win two proxy contests to gain control of the company, as they can only elect a third of the board in a single election if there is a secret board. Previous evidence shows that shareholders are more likely to vote for dissidents than in routine contests, but less likely than in takeover contests involving a tender offer.56 To capture these preferences, we assume that Institution II votes according to the function r( 20,-. 1,x) and Institution III votes according to r(20, .2, x).
In the case of a two-part competition, such as the case where the board is distributed, the two groups will vote in the second competition according to the functions r(20,-.3,x) and r(20, O ,x). These choices reflect the intuition that shareholders will be more biased toward management in the second vote than in the initial vote. Arbitrageurs are not likely to actively engage in these battles as there is no pending tender offer at a premium over the market price.
Takeover Contests Using A Tender Offer
In this case, we will assume that institution II and institution III are most negative towards management and vote according to the function r(20, -.5, x) and r(20, -.2, x) respectively. If there is a second vote, as in the case of a staggered board, institution II will vote according to the same function, but institution III will vote according to the function r(20, -.5, x), reflecting a more negative evaluation of management. In this situation, institution III can be thought of as representing arbitrage, speculating on the probability that the offer will succeed.
Basic Voting Results
- R outine V ote
- Proxy Contest for Corporate Control
- Takeover Bid Using A Tender Offer
- Proxy Contests for Corporate Control
- Takeover Bids Using a Tender Offer
Regardless of the proxy signal, the dissidents win when they are allowed to collect 30 percent of the target company's stock. In this case, we have a bidder who has announced an offer for a target company and is forced to participate in a proxy contest due to the target group's poisoned pill. However, since the target company does not have a secret committee, we only need to consider the bidder's probability of succeeding in a particular election race.
Now a negative opinion from the proxy advisor leads to an almost certain victory for the bidder, except in the case of the founding family, where the odds are approximately equal. When will the independent proxy advisor make a recommendation to management if there is a more expensive offer for shareholders? Calculations not shown show that even in scenario A, if we give the bidder 15 percent of the target company's shares, the chance that management wins only decreases slightly.
ISS typically states its recommendations based on the transaction's value to the target company's shareholders and its value to diversified investors with holdings in both companies. Institutions are likely to make their own sales decisions internally without the use of representative counsel.
Summary of Basic Results
POLICY IMPLICATIONS
The important conclusion of the model is that the minority position can win elections if voters who hold that position are diligent enough, by which we mean that voters get a. There are two criticisms to be made of the conclusions drawn from this model by Gilson and Schwartz. Indeed, they may be able to increase the size of the private benefits they receive relative to those.
Gilson and Schwartz go on to show that the hypotheses of the Feddersen/Pesendorfer model do not hold. They also note that the Feddersen/Pesendorfer assumption that the distribution of shareholder utilities is common knowledge is false in the corporate context. To incorporate imperfect information among some shareholders about the magnitudes of the cash flows (Yi and YR), we assume that (YI,YR) follows a known probability distribution with mean (YI*, YR*) and support (0,Y) x.
A stable result means that the market makers will set a price in such a way that the outcome of the election results in the price they set. If R owns a large number of shares, then there is some way he can convey to the uninformed shareholders what YR is. which R holds. But in the course of the proxy contest, R can announce what price he is prepared to offer for the shares.
With an explicit YR signal, R can avoid the ambiguity inherent in market makers setting the stock price.
