Frozen Charters - Harvard Law School

Frozen Charters - Harvard Law School

FROZEN CHARTERS Scott Hirst* Abstract The corporate charters of a substantial number of U.S. public companies are frozen. These corporations are unab...

440KB Sizes 0 Downloads 5 Views

Recommend Documents

HARVARD LAW SCHOOL
1945. Robert H. Young. 1948. George W. McManus, Jr.**. 1951. Martin D. Cohn. William Cousins, Jr. George Vetter. 1952. J

property - Harvard Law School
century. Later on in the century the Statute Quia Emptores, 18 Edw. l, c. l (1290), is generally regarded as having rein

Executive Pensions - Harvard Law School - Harvard University
Apr 14, 2006 - Lucian A. Bebchuk and Robert J. Jackson, Jr., "Executive Pensions," 30 ... Lucian A. Bebchuk and Jesse M.

610 PDF - Harvard Law School - Harvard University
The circus mesmerized, recalled the Ringling brothers. .... 12 See David Lewis Hammarstrom, Big Top Boss: John Ringling

593: PDF - Harvard Law School - Harvard University
May 5, 2008 - nounced a plan to diversify the Trust's investment portfolio by selling the. Trust's controlling interest

asymmetric information - Harvard Law School - Harvard University
Rand journal of Economics. Voi, 15, No. 3, Autumn 1984. Litigation and settlement under imperfect information. Lucian Ar

588: PDF - Harvard Law School
three distinct domains: compliance costs, stock price reactions, and firms' decisions ... small companies have historica

The Judiciary - Harvard Law School
Honorable Vernon S. Broderick. 1988 Honorable Julia ... Honorable Henry H. Kennedy, Jr. 1973 Honorable ... Honorable Hen

Session Times - Harvard Law School Bicentennial - Harvard University
Session​ ​One​ ​-​ ​9:00am. A​ ​Conversation​ ​with​ ​Bob​ ​Zoellick. Hosts:​ ​​Philip

Remainders and Reversions - Harvard Law School
It could be, for example, a remainder in fee simple absolute, fee tail, fee simple .... of vested remainders following a

FROZEN CHARTERS Scott Hirst*

Abstract The corporate charters of a substantial number of U.S. public companies are frozen. These corporations are unable to amend their charters, even when amendments are supported by the boards and the overwhelming majority of shareholders of the corporations. This is the consequence of a recent New York Stock Exchange (NYSE) ‘Information Memorandum 12-4’ (IM 12-4), which restricted broker voting on charter and bylaw amendments. Implemented with the ostensible goal of improving corporate governance and protecting investors, IM 12-4 has in fact had the opposite effect. This paper is the first to describe frozen charters, and to examine the significant unintended consequences of IM 12-4. The empirical evidence I present shows that eliminating uninstructed broker voting has resulted in distortions of charter amendment votes, causing a substantial number of charter amendments to have failed since 2012 despite receiving over 90% shareholder support. IM 12-4 has increased the likelihood of failure for charter amendments in a statistically and economically significant manner. I also provide empirical evidence of two related types of ‘distorted fail’ outcomes resulting from IM 12-4. However, this analysis significantly underestimates the true effect of IM 12-4. I estimate that there are hundreds more U.S. public corporations that have been rendered unable to amend part of their charter as a result of IM 12-4. I compare these shareholder-harming consequences to the intended effect of IM 12-4, preventing ‘distorted pass’ outcomes. I show that there have been no charter amendments, and only one other management proposal, where IM 12-4 has prevented distortion. From the investor perspective implied by IM 12-4, the harm from preventing value-enhancing amendments outweighs the benefits in preventing ‘distorted passes.’ I put forward a number of potential solutions to address the frozen charters created by IM 12-4, while maintaining its limited benefits. Proportional voting by brokers would eliminate both kinds of distortion, as would allowing broker voting only on those corporate governance changes where frozen charters are most likely, and where ‘distorted pass’ outcomes are not likely. The NYSE should reverse IM 12-4, and implement one of these solutions to undo the investor-harming effects of IM 12-4. Keywords: Broker voting, shareholder voting, charter amendment, bylaw amendment JEL Classification: D22, G30, G32, G34, G38, K22.

*

Associate Director of the Program on Institutional Investors, Lecturer on Law and Terence M. Considine Fellow in Corporate Governance, Harvard Law School. For helpful suggestions and discussions, I am grateful to Lucian Bebchuk, Robert Clark, John Coates, Louis Kaplow, Kobi Kastiel, Reinier Kraakman, Yaron Nili, Steven Shavell, Nitzan Shilon, Holger Spamann, Guhan Subramanian, and Andrew Tuch, as well as participants in the Harvard Law School Law & Economics Workshop. All errors are my own. I would like to thank the John M. Olin Center for Law, Economics and Business at Harvard Law School and the Harvard Law School Program on Corporate Governance for their financial support.

FROZEN CHARTERS

Table of Contents I. Introduction ................................................................................................................ 4 II. Charter Amendments and Broker Voting ................................................................... 7 A. Amending Corporate Charters ............................................................................... 7 B. Broker Voting...................................................................................................... 10 B. NYSE Rules on Discretionary Broker Voting....................................................... 13 1. Rule 452........................................................................................................... 13 2. 2003 and 2010 Amendments to Rule 452.......................................................... 14 3. Information Memorandum 12-4........................................................................ 16 III. The Consequences of IM 12-4 .................................................................................17 A. Broker Voting Distortions, ‘Distorted Passes’ and ‘Distorted Fails’ ..................... 17 B. ‘Distorted Fails’ Resulting from IM 12-4 ............................................................. 18 1. Frozen Charters ................................................................................................ 18 2. Bylaws Not Amendable by Shareholders .......................................................... 21 3. Insider Vetoes .................................................................................................. 22 C. The Reduction in ‘Distorted Pass’ Resulting from IM 12-4 .................................. 24 IV. Estimating the Companies Affected by IM 12-4 ......................................................27 A. Companies with Frozen Charters from IM 12-4 ................................................... 27 B. Companies Where Shareholders Cannot Amend Bylaws from IM 12-4 ................ 28 C. Companies with Insider Vetoes from IM 12-4 ...................................................... 29 D. Reductions in Potential ‘Distorted Pass’ Results from IM 12-4 ............................ 30 V. Evaluating IM 12-4 ..................................................................................................32 A. Evaluating IM 12-4 from an Investor Perspective ................................................ 32 B. Evaluating IM 12-4 from a Firm Value Perspective.............................................. 34 C. Evaluating IM 12-4 From a Contractarian Perspective ......................................... 34 D. Evaluating IM 12-4 from Directors’ and Managers’ Perspectives ......................... 35 E. Evaluating IM 12-4 from a Procedural Perspective ............................................... 37 VI. Fixing IM 12-4........................................................................................................38 A. Reversing IM 12-4............................................................................................... 39 B. Reducing Undirected Broker Votes ...................................................................... 40 C. Proportional Voting ............................................................................................. 41 D. Permitting Broker Voting on Certain Charter Amendments.................................. 44 D. The Procedure of Reform .................................................................................... 46 VII. Conclusion ............................................................................................................47

2

FROZEN CHARTERS

Appendices Appendix A: Certain ‘Distorted Fail’ Outcomes, 2012-2013 .........................................49 Appendix B: IM 12-4 and the Likelihood of Charter Amendment Failure ......................52 Appendix C: Management Proposals Receiving <50% Shareholder Support 2012-2013 53 Appendix D: Charter Amendments and Director Election Turnout ................................54 Appendix E: Estimates of Companies with Potential ‘Distorted Fail’ and ‘Distorted Pass’ Results ..........................................................................................................................55

3

FROZEN CHARTERS

I. Introduction In 2011, a client represented by the Shareholder Rights Project (with which I am affiliated)1 signed an agreement with an S&P 500 company for the company to amend its charter, the central document establishing the internal rules of the corporation. The board of directors and management supported and recommended the change. At the required shareholder vote on the amendment, more than 99% of the votes cast were in favor of the amendment. But the amendment failed. Directors, managers and shareholders supported a change in the company’s charter, but were unable to change it. The company’s charter is frozen. This result is the consequence of a recent New York Stock Exchange (NYSE) rule change relating to broker voting, ‘Information Memorandum 12-4’ (IM 12-4). Although IM 12-4 was intended to protect investors and improve corporate governance and accountability, it has had the opposite effect. This is the first paper to identify the problem of frozen charters, and the other significant unintended consequences of IM 12-4. Because about 85% of investors hold shares through brokers, and because many of those investors do not instruct brokers how to vote their shares at annual meetings, uninstructed broker votes represent a substantial proportion of the outstanding shares of many corporations (an average of 10% in my sample). In order to protect shareholders from the potentially distortive effects of voting by brokers (who do not have an economic interest in the corporation), the NYSE has progressively limited the instances in which brokers may vote shares. In 2012, in IM 12-4, the NYSE restricted brokers from voting on charter and bylaw amendments. However, although the intention of IM 12-4 was to protect investors and enfranchise shareholders, I show that it has had the opposite effect. Although its intention was to remove the possibility of distortion, by preventing broker voting, it has created a different kind of distortion – resulting in proposals that have several different kinds of ‘distorted fails.’ A certain percentage of uninstructed shares would support a proposal. Preventing brokers from voting uninstructed shares means that none of those shares will be voted in favor of the proposal. Where the proposal would have passed had the uninstructed shares that supported the proposal voted, the outcome will be a ‘distorted fail.’ The main type of ‘distorted fail’ results from IM 12-4 are ‘frozen charters.’ Despite strong support from shareholders and directors, there are a number of corporations which, because of high supermajority requirements to amend parts of their charters, are unable to reach these thresholds without the broker votes that IM 12-4 has prohibited. I gather empirical data on the number of failed charter amendments since IM 12-4 was implemented, which shows that, in the two years since IM 12-4 took effect,

1

For further information on the work of the Shareholder Rights Project, see Lucian Bebchuk, Scott Hirst and June Rhee, “Towards the Declassification of S&P 500 Boards”, 3 Harv. Bus. L. Rev. 157 (2013).

4

FROZEN CHARTERS 35 of the 42 companies where charter amendments failed despite receiving overwhelming shareholder support would have had their amendments passed had IM 12-4 not been implemented. My analysis also shows that IM 12-4 has significantly increased the proportion of charter amendments that fail despite receiving overwhelming shareholder support, and that IM 12-4 has increased the likelihood of failure for charter amendments in a statistically and economically significant manner. I also provide empirical evidence of two other related kinds of ‘distorted fail’ results from effects of IM 12-4. Similar to frozen charters, the elimination of discretionary broker voting has prevented shareholders of a number of corporations from amending the bylaws of those corporations. And in a number of other companies, the prohibition on broker voting has left insiders with a de facto veto right over certain charter amendments. I compare the unintended, shareholder-harming ‘distorted fail’ results from IM 12-4 with its intended consequence, of preventing ‘distorted passes’ resulting from broker voting. Broker votes will result in a ‘distorted pass’ when a majority of the shareholders of the corporation prefer that a proposal fail, but broker votes in favor of the proposal cause it to pass. However, because most charter amendments are strongly supported by shareholders, and because brokers would follow management recommendations to support these charter amendments, in the overwhelming majority of cases there would be no distortive effects of broker voting on charter amendments. The number of cases where there may be ‘distorted pass’ outcomes is very small: there have been no charter amendments since IM 12-4 came into effect that would have had ‘distorted pass’ outcomes had broker votes been permitted, and only one other management proposal (0.2% of all management proposals during that period). However, the foregoing analysis significantly underestimates the true effect of IM 12-4, because only a small number of charter amendments go to a vote each year. I develop a model to estimate the number of companies that are affected by the consequences of IM 12-4, using corporations’ voting requirements and turnout in director elections. Based on the most conservative assumptions, I estimate that about 12% of U.S. corporations have been rendered unable to amend part of their charter as a result of IM 12-4. I further estimate that about 9% of corporations have an insider veto as a result of IM 12-4, and that about 11% of corporations now have board-only bylaw amendments as a result of IM 12-4. I also estimate the number of companies where eliminating broker voting has prevented ‘distorted pass’ results, and find that, based on current patterns of support for corporate governance proposals, IM 12-4 would prevent ‘distorted pass’ results at only 0.1% of companies. I use these empirical results to evaluate IM 12-4 against its own policy goals of protecting investors, enfranchising shareholders and improving corporate governance and accountability. The substantial negative effects of IM 12-4 in creating ‘distorted fail’ results — preventing charter and bylaw amendments desired by shareholders and giving insiders a veto over charter amendments — significantly outweigh its very limited benefits in preventing ‘distorted pass’ results. Considered from a firm value perspective, IM 12-4 prevents value-enhancing changes to the corporation. And from the contractarian

5

FROZEN CHARTERS framework of corporate law scholarship, IM 12-4 is problematic in that it changes an implicit term on which the amendment provisions contained in corporate charters rely, thereby changing the terms of the corporate contract without the agreement of the parties. I also examine IM 12-4 from the perspective of directors and managers, and explain why certain directors and managers may actually prefer that particular charter amendments be stymied by IM 12-4. Finally, I examine the manner in which IM 12-4 was implemented, and show that, compared to the standard process by which Rule 452 has been amended in the past, the IM 12-4 process prevented public comment and any consideration of the effects of IM 12-4 on investors, which may otherwise have made clear the negative consequences of the rule. I consider a number of possible alternatives to address the ‘distorted fail’ results created by IM 12-4. Although the current pattern of support for charter amendments and other corporate governance proposals means there are very few potential ‘distorted pass’ results, this is not necessarily the case, so I consider the effects of the solutions I consider in both reducing ‘distorted fail’ results and avoiding ‘distorted pass’ results. The simplest solution, reversing IM 12-4, would eliminate the ‘distorted fail’ results of frozen charters, failed bylaw votes and insider vetoes created by IM 12-4. However, it would reinstate the possibility of small number of ‘distorted pass’ results. Another possibility is to reduce the number of undirected broker votes, which would reduce both ‘distorted pass’ and ‘distorted fail’ results. However this may be costly, and is unlikely to provide a complete solution to the problem. A more promising alternative, proportional voting, whereby brokers voted uninstructed shares in the same proportion as other shareholders at the meeting, or other shareholder clients of the broker, has the possibility of eliminating both kinds of distortion. Finally, broker voting rules could be amended to allow broker voting on certain types of corporate governance proposals where ‘distorted fail’ outcomes would otherwise be likely, but where ‘distorted pass’ outcomes are not, such as amendments to remove supermajority provisions, amendments that shareholders generally support, or amendments for which a supermajority is required. These solutions allow the possibility of reducing the instances of ‘distorted fail’ results, while also minimizing the possibility of ‘distorted pass’ results. I explain that in order to avoid the procedural shortcomings involved in implementing IM 12-4, any reform should take place through the NYSE rulemaking process, with notice and public comment, and a Securities and Exchange Commission (SEC) determination that the new rule has the effect of protecting investors. In the interim, the NYSE (or if the NYSE is unwilling, the SEC) should reverse IM 12-4 to prevent the harm it is currently causing to investors. My analysis is structured as follows. Part II explains the charter amendment process, broker voting and the broker voting rules, including the recent changes that set out the policy on which IM 12-4 is based, and IM 12-4 itself. Part III examines the kinds of distortion created by broker voting, ‘distorted pass’ outcomes and ‘distorted fail’ outcomes, and uses these to demonstrate the consequences of IM 12-4 on charter and bylaw amendments in the period since it was implemented. Part IV expands this analysis by considering those companies that are affected by IM 12-4, but may not yet have brought charter or bylaw amendments. Part V evaluates IM 12-4 from its own perspective

6

FROZEN CHARTERS of investor protection, as well as several other perspectives. Part VI considers potential solutions for addressing the negative consequences of IM 12-4, and Part VII concludes.

II. Charter Amendments and Broker Voting This paper examines the phenomenon of frozen charters, and links it to the effect of recent changes to broker voting rules. In this Part II, I first set out the background of shareholder voting to approve charter amendments. I then focus on part of the shareholder vote – voting of uninstructed shares by brokers. I examine the rules that govern broker voting, most notably NYSE Rule 452, and then move to consider recent changes to the broker voting rules, including Information Memorandum 12-4.

A. Amending Corporate Charters The general procedure for amending the charter of a corporation is for the board of directors of the corporation to approve an amendment to the charter, and for directors to put forward a proposal that the amendment be approved by a vote of shareholders.2 The proposal is then voted on at a meeting of shareholders.3 If the amendment is approved, it is formalized by a filing with the secretary of state of the corporation’s state of incorporation. The vote required to approve an amendment varies from state to state, and from company to company. Delaware, California and New York, as well as a number of other states, require the affirmative vote of a majority of the outstanding stock entitled to be voted on the amendment.4 A number of other states require the affirmative vote of two-

2

3

4

e.g. 8 Del. C. § 242(b)(1) (“[The] board of directors shall adopt a resolution setting forth the amendment proposed, declaring its advisability, and either calling a special meeting … or directing that the amendment proposed be considered at the next annual meeting of the stockholders. … At the meeting a vote of the stockholders entitled to vote thereon shall be taken for and against the proposed amendment.”) Generally the meeting at which this occurs is an annual meeting, though it can also be voted on at a special meeting of shareholders. Some states may allow shareholder approval by written consent in lieu of a meeting – see, e.g., 8 Del. C. § 228. e.g. 8 Del. C. § 242(b)(1) (“If a majority of the outstanding stock entitled to vote thereon, and a majority of the outstanding stock of each class entitled to vote thereon as a class has been voted in favor of the amendment…”). The other states in this group are Alaska (Alaska Stat. § 10.06.504(a)), California (Cal Corp Code § 902), Hawaii (HRS § 414-283), Kansas (K.S.A. § 17-6602), Maine (13-C M.R.S. § 1003), Michigan (MCLS § 450.1611), Missouri (§ 351.090 R.S.Mo.), Nevada (Nev. Rev. Stat. Ann. § 78.390), New Jersey (N.J. Stat. § 14A:9-2), New Mexico (N.M. Stat. Ann. § 53-13-2), New York (NY CLS Bus Corp § 803), Oklahoma (18 Okl. St. § 1077), Pennsylvania (15 Pa.C.S. § 1914), Rhode Island (R.I. Gen. Laws § 7-1.2-903) and Washington (Rev. Code Wash. (ARCW) § 23B.10.030).

7

FROZEN CHARTERS thirds of the outstanding shares. 5 A third group of states that follow the provisions of the Model Business Corporation Act (MBCA) require a majority of the votes cast at a meeting at which a quorum of the majority of the outstanding shares are present.6 All states permit a corporation to require a greater vote requirement by including a specific provision in their charter,7 and a significant number of corporations take advantage of this provision by requiring votes of, for example, 66%, 75% or 80% of outstanding shares in order to approve an amendment to the charter as a whole, or to certain parts of the charter.8 Although it is possible for a supermajority provision to require that any part of the charter cannot be amended without meeting the supermajority vote requirement, the provisions more frequently apply to particular parts of the charter, especially anti-takeover provisions. The distribution of voting requirements for the sample of companies I use in this paper – companies incorporated in the United States that are part of the Russell 3000 index9 – is set out in Table 1 below.

5

6

7 8

9

Illinois (805 ILCS 5/10.20), Louisiana (La. R.S. 12:31), Maryland (Md. Corps. & Assocs. Code Ann. § 2-604), Massachusetts (ALM GL ch. 156D, § 10.03), Ohio (ORC Ann. 1701.71), South Carolina (S.C. Code Ann. § 33-10-103), Texas (Tex. Business Organizations Code § 21.364) and Virginia (Va. Code Ann. § 13.1-707). Alabama (Code of Ala. § 10A-2-10.03), Arizona (A.R.S. § 10-1003), Arkansas (A.C.A. § 4-271003), Colorado (C.R.S. 7-110-103), Connecticut (Conn. Gen. Stat. § 33-797), District of Columbia (D.C. Code § 29-308.03), Florida (Fla. Stat. § 607.1003), Georgia (O.C.G.A. § 14-21003), Idaho (Idaho Code § 30-1-1003), Indiana (IN Code § 23-1-38-3), Iowa (Iowa Code § 490.1003), Kentucky (KRS § 271B.10-030), Minnesota (Minn. Stat. § 302A.135; ), Mississippi (Miss. Code Ann. § 79-4-10.03), Montana (Mont. Code Anno., § 35-1-227), Nebraska (R.R.S. Neb. § 21-20,118), New Hampshire (RSA 293-A:10.03), North Carolina (N.C. Gen. Stat. § 5510-03), North Dakota (N.D. Cent. Code, § 10-19.1-19; N.D. Cent. Code, § 10-19.1-74), Oregon (ORS § 60.437), South Dakota (S.D. Codified Laws § 47-1A-1003), Tennessee (Tenn. Code Ann. § 48-20-103), Utah (Utah Code Ann. § 16-10a-1003), Vermont (11A V.S.A. § 10.03), West Virginia (W. Va. Code § 31D-10-1003), Wisconsin (Wis. Stat. § 180.1003) and Wyoming (Wyo. Stat. § 17-16-1003). e.g., e.g. 8 Del. C. § 102(b)(4). See, e.g., Restated Certificate of Incorporation of Chesapeake Energy Corporation, filed as an exhibit to Quarterly Statement on Form 10-Q, filed August 10, 2001, (“Notwithstanding anything contained in this Certificate of Incorporation to the contrary, the affirmative vote of the holders of at least sixty-six and two-thirds percent (66 2/3%) of the issued and outstanding stock having voting power, voting together as a single class, shall be required to amend, repeal or adopt any provision inconsistent with Articles V, VI, VII, VIII and this Article IX of this Certificate of Incorporation.") For an explanation of why I focus on the Russell 3000, see the discussion at Note 45, infra.

8

FROZEN CHARTERS Table 1: Voting Requirements for U.S.-incorporated Russell 3000 Corporations Vote Requirement

Companies

%

122

5.0%

885

36.4%

50% - 59%

1

0.0%

60% - 69%

790

32.5%

70% - 79%

226

9.3%

80% - 89%

394

16.2%

90% - 100%

15

0.6%

2,433

100.0%

Majority of Votes Cast 50%

Shares Outstanding

Total

As Table 1 shows, 26.1% of the companies in my sample require approval of 70% or more of the outstanding shares of the corporation to amend part of their charters. It’s important to note that most of the supermajority requirements discussed above, and the default requirements of many states (including Delaware, New York and California), are a percentage of shares outstanding, rather of the votes cast on the proposal. Of course, if all outstanding shares were voted, there would be no difference between the two figures. However, the reality is that the proportion of outstanding shares that are voted by shareholders at annual meetings – which I refer to as ‘shareholder turnout’10 – is less than 100%, since a substantial percentage of shareholders do not vote. The likelihood that a shareholder will vote varies with the level of institutional ownership of the corporation. Of shares held by institutions, an average of 90% were voted at annual meetings in 2013, whereas only 30% of shares owned by individual or ‘retail’ investors were voted.11 The distribution of shareholder turnout (excluding uninstructed broker votes) for director elections in 2012 for the sample that I consider in Part IV is set out in Figure 1, below.

10

I exclude broker votes from ‘shareholder turnout’, so shareholder turnout is the number of shares that are voted for, against or abstain on a proposal (but not uninstructed broker shares), divided by the number of outstanding shares of the corporation. 11 See Broadridge & PwC ProxyPulse, ‘2013 Proxy Season Recap’, Third Edition (2013). Note that the data published by Broadridge is based on shares held in street name.

9

FROZEN CHARTERS Figure 1: Voter Turnout in Director Elections for Russell 3000 Companies, 2012

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

As Figure 2 makes clear, although turnout is generally high, there are a substantial number of corporations that have turnout below some of the significant supermajority thresholds mentioned in Table 1 – 45% had turnout below 80% of shares outstanding.12 As a result, the treatment of uninstructed broker votes can be central to whether or not charter amendments pass.

B. Broker Voting The right to vote shares of the corporation belongs to the registered owner of the shares, who is listed in the share register of the corporation. While it is possible for an investor to be the ‘record holder’ or registered owner of the shares (known as owning through ‘direct registration’), many investors hold shares through securities intermediaries,13 such as a broker, bank or custodian (known as holding in ‘street name’), in which case the intermediary is the registered owner of the shares, and the investor’s interest as the beneficial owner is recorded in the books of the intermediary.14 Investors

12

As explained further in Part IV, in order to make a conservative estimate of frozen charters, my data uses the maximum turnout for a company between 2010 and 2012, so the distribution in Figure 2 is likely to overestimate turnout, or underestimate the proportion of companies that have turnout below the high supermajority thresholds. 13 ‘Securities intermediary’ is defined in Act Rule 17Ad-20 of the General Rules and Regulations (henceforth, the “Exchange Act Rules”) promulgated pursuant to the Securities and Exchange of 1934 (henceforth, the “Exchange Act”)) to include a clearing agency registered under Section 17A of the Exchange Act or a person, including a bank, broker or dealer, that maintains securities accounts for others. 14 For simplicity, I will generally refer to intermediaries as ‘brokers,’ though in actual fact particular investors may have intermediaries that are brokers, banks or custodians, or some

10

FROZEN CHARTERS generally buy shares through brokers, since brokers have access to stock exchange trading platforms and can place buy orders on behalf of the investor. Title to shares is transferred from the seller’s broker to the buyer’s broker.15 It is possible for investors to request that brokers transfer the registration to the investor. However, both investors and brokers disfavor this approach, as it increases the difficulty, cost and time required to transfer the shares, and brokers often charge the investor an additional fee to transfer the registration. As a result, brokers and other intermediaries are the registered owners of most shares owned by investors – approximately 85% of public company shares.16 Shareholder voting takes place by proxy. The corporation distributes proxy materials and proxy forms to shareholders.17 Shareholders complete the proxy forms, authorizing the directors of the corporation to vote on their behalf, in the manner that the shareholder specifies.18 Where the shareholder holds their shares through direct registration, the shareholder receives a proxy form directly from the corporation, and completes the proxy form, and submits it to the corporation. However, where the shareholder holds their shares in ‘street name’ the broker is the registered owner of the shares, and rules applicable to the broker govern the voting of the shares. Broker voting is governed by the exchanges of which the broker is a member, especially the New York Stock Exchange Rules. The NYSE Rules apply to all brokers that are members of the NYSE – including with respect to the brokers’ voting of shares of companies that are listed on other exchanges.19 Because almost all U.S. brokers are members of the NYSE, the NYSE Rules therefore govern essentially all broker voting of companies listed on U.S. exchanges.

combination thereof. The Exchange Act and NYSE Rules use the term ‘broker-dealer’. A ‘broker’ is defined as a person who undertakes transactions in securities for the account of others (see Exchange Act Section 3(a)(4)), whereas a ‘dealer’ is one who buys and sells securities for their own account (see Exchange Act Section 3(a)(5)). 15 The shares are often held by another intermediary, a ‘custodian.’ In the case of large retail brokers, custodian entities are often under common ownership with the broker entity. 16 See, e.g., Securities and Exchange Commission, Roundtable on Proxy Voting Mechanics (“Approximately 85% of exchange-traded securities are held by securities intermediaries”), May 23, 2007 (available at http://www.sec.gov/spotlight/proxyprocess/proxyvotingbrief.htm). 17 It is also possible that the corporation may give notice of where to access proxy materials online, in lieu of sending the materials themselves. 18 For many corporations, shareholder may also give their proxy electronically or telephonically as well as in physical form. 19 See NYSE Rule 452 “Giving Proxies by Member Organization” (henceforth, “Rule 452”); see also NYSE MKT Company Guide Section 723 “Giving Proxies by Member Organization”, relating to members of NYSE MKT exchange (formerly, the American Stock Exchange), which is substantively identical to Rule 452. FINRA Rule 2251 “Forwarding of Proxy and Other Issuer-Related Materials,” which applies to brokers that are members of the NASDAQ exchange, does not provide substantive guidance regarding whether a member may vote proxies it does not beneficially own, but provides that “a member may give a proxy to vote any stock pursuant to the rules of any national securities exchange of which it is a member provided that the records of the member clearly indicate the procedure it is following.”

11

FROZEN CHARTERS Where shares are held in street name, because the broker is the registered owner of the shares, the broker receives the proxy materials sent by the company.20 NYSE Rule 451 requires the broker to send the proxy materials to the beneficial owner or their investment adviser. The broker may send a blank, signed proxy form to the beneficial owner, in which case the broker has no further role in the voting of the proxy; if the beneficial owner wishes to vote, they complete the proxy card and submit it to the company. Alternatively, the broker must request that the beneficial owner provide the broker with instructions on how to vote the proxy (together with the deadline for doing so).21 If the broker receives instructions from the beneficial owner by the tenth day before the company’s meeting, NYSE Rule 452 requires the broker to vote the shares as directed by the beneficial owner. The question of broker discretionary voting arises in the situation where the broker has not received voting instructions by the deadline. Broker discretionary voting is important because many shareholders – especially retail shareholders – do not instruct brokers how to vote their shares. As a result, uninstructed broker votes represent a substantial number of shares at many companies. For the sample of companies that I consider in Part IV, the average level of uninstructed broker votes was 10.4% of outstanding shares. The full distribution of broker votes for director elections in 2012 for the companies in the sample I describe in Part IV is set out in Figure 2, below.22

20

These include the company’s annual report and proxy statement, which includes disclosure required by the Exchange Act about the matters to be voted on, and the proxy card, to be sent in to the company to give a proxy to vote the shares. 21 NYSE Rule 451 “Transmission of Proxy Material”. If the proxy soliciting material is transmitted to the beneficial owner or its adviser twenty-five days or more before the meeting, the statement must state that the proxy may be given fifteen days before the meeting at the discretion of the owner of record of the stock. 22 The sample is 2,136 companies disclosing broker non-votes for annual elections between 2010 and 2013.

12

FROZEN CHARTERS Figure 2: Uninstructed Broker Votes in Director Elections of U.S. Russell 3000 Companies, 2012

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

B. NYSE Rules on Discretionary Broker Voting 1. Rule 452 Discretionary broker voting is governed by NYSE Rule 452. The rule provides that a broker that has sent proxy materials to the beneficial owner of shares and hasn’t received instructions on voting the shares by the deadline: may give or authorize the giving of a proxy to [vote] such stock, provided the person in the member organization giving or authorizing the giving of the proxy has no knowledge of any contest as to the action to be taken at the meeting and provided such action is adequately disclosed to stockholders and does not include authorization for a merger, consolidation or any other matter which may affect substantially the rights or privileges of such stock.23 The supplementary materials to Rule 452 give further guidance on as to when brokers are not permitted to vote without instructions from the beneficial owner, and list 21 matters on which brokers are not permitted to vote discretionarily. These include contested proposals, shareholder proposals opposed by management, proposals relating to

23

NYSE Rule 452; see also NYSE Rule 452 Supplementary Material .10 “When member organization may vote without customer instructions”, substantially restating the provisions of Rule 452.

13

FROZEN CHARTERS mergers, proposals involving appraisal rights, preemptive rights, or voting provisions, proposals relating to executive compensation, and proposals for the election of directors.24 Rule 452 reflects a concern about distortion by brokers, based on widely held understanding that brokers vote overwhelmingly in the manner recommended by directors.25 The rule limits broker voting in situations where there is most concern about potential distortion – where there is likely to be divergence between what managers recommend (and what brokers are likely to support) and what shareholders might prefer – such as mergers, or contested elections. 2. 2003 and 2010 Amendments to Rule 452 Since the precursor to Rule 452 was adopted in 1937,26 there have been several changes to the matters on which brokers are permitted to vote discretionarily. The rationale for the most recent of these changes, in 2003 and 2010, help explain the rationale for the changes made in IM 12-4. In 2003, following the implementation of the requirement for shareholders to approve equity compensation plans, Rule 452 was amended to prevent broker discretionary voting on equity compensation plans.27 In approving the rule, the SEC considered numerous comments on the proposed rule, from corporations, investors and others,28 and determined that the importance of ensuring that votes on executive compensation matters reflect the views of beneficial shareholders outweighed potential difficulties or costs in obtaining necessary approvals, and that therefore the amendments served to protect shareholders and were in the public interest.29 Following the implementation of the 2003 changes, the SEC created a ‘Proxy Working Group’30 to review the rules regulating the proxy voting process, and in particular, Rule 452.31 The Proxy Working Group engaged in a comprehensive

24

NYSE Rule 452 Supplementary Material .11 “When member organization may not vote without customer instructions.” 25 See the Report and Recommendations of the Proxy Working Group to the New York Stock Exchange (henceforth, “Proxy Working Group Report”), June 5, 2006, at 14. See also Jennifer Bethel & Stuart Gillan, The Impact of the Institutional and Regulatory Environment on Shareholder Voting, 31 Fin. Mgmt. 29 (2002). 26 See Proxy Working Group Report, supra note 25, at 10. 27 See SEC Release No. 34-48108, “Self-Regulatory Organizations; New York Stock Exchange, Inc. and National Association of Securities Dealers, Inc.; Order Approving NYSE and Nasdaq Proposed Rule Changes and Nasdaq Amendment No. 1 and Notice of Filing and Order Granting Accelerated Approval to NYSE Amendments No. 1 and 2 and Nasdaq Amendments No. 2 and 3 Thereto Relating to Equity Compensation Plans”, June 30, 2003 (68 Fed. Reg. 39,995, July 3, 2003). 28 See Footnote 5 to SEC Release No. 34-48108, supra note 27, at 39,995. 29 See SEC Release No. 34-48108, supra note 27, at 40,008-9. 30 The Proxy Working Group included representatives of a number of corporations, institutional investors and attorneys, and their sessions were also attended by representatives of the SEC, the NYSE and the National Association of Securities Dealers, the predecessor of the Financial Industry Regulatory Authority (FINRA). 31 See Proxy Working Group Report, supra note 25, at 1.

14

FROZEN CHARTERS consideration of Rule 452, and possible ways to improve the rule. The Proxy Working Group recognized the potentially distorting effect of broker votes: [T]he problem with broker voting is that it allows someone (i.e. the broker) who does not have an economic interest in the corporation the opportunity to vote on the corporation’s business. A second problem is that historically brokers have generally cast uninstructed shares overwhelmingly in support of the board’s recommendations, which provides a significant advantage to the incumbent board in director elections and other matters.32 The Proxy Working Group weighed this against the cost of disallowing discretionary broker voting, in particular, the increased difficulty in obtaining necessary quorums.33 The Proxy Working Group concluded that, in the interests of “better corporate governance and transparency of the election process,” broker voting on uncontested elections should be eliminated.34 When the Dodd Frank Wall Street Reform and Consumer Protection Act was enacted, it included a provision amending the Securities Exchange Act of 1934 to require that exchanges prevent their members from making uninstructed votes on elections of directors or executive compensation matters (or any other significant matter as determined by the SEC).35 The SEC subsequently approved the NYSE’s proposed amendments, concluding that the rules would “better enfranchise shareholders, and thereby enhance corporate governance and accountability” and should “protect investors and the public interest.”36 As a result, since the rules came into effect on January 1, 2010, brokers have no longer been able to vote uninstructed shares on director elections. In September 2010, the SEC subsequently approved proposed NYSE rules extending the restrictions on voting on executive compensation matters mandated by the Dodd-Frank Act, concluding that “the proposal will further investor protection and the public interest” and “should enhance corporate governance and accountability to shareholders.”37 For these and other votes on which brokers do not vote uninstructed

32

See Proxy Working Group Report, supra note 25, at 14. See Proxy Working Group Report, supra note 25, at 15. 34 See Proxy Working Group Report, supra note 25, at 15. 35 Dodd Frank Wall Street Reform and Consumer Protection Act, Section 957, amending Section 6(b)(10) of the Securities Exchange Act of 1934 (15 U.S. 78F). 36 SEC Release No. 34-60215, “Self-Regulatory Organizations; New York Stock Exchange LLC; Order Approving Proposed Rule Change, as modified by Amendment No. 4, to Amend NYSE Rule 452 and Corresponding Listed Company Manual Section 402.08 to Eliminate Broker Discretionary Voting for the Election of Directors, Except for Companies Registered under the Investment Company Act of 1940, and to Codify Two Previously Published Interpretations that Do Not Permit Broker Discretionary Voting for Material Amendments to Investment Advisory Contracts with an Investment Company”, July 1, 2009 (henceforth, the “2009 SEC Release”) (74 Fed. Reg. 33,293, Jul. 10, 2009), 12. 37 See SEC Release 34-62874, “Self-Regulatory Organizations; New York Stock Exchange LLC; Notice of Filing and Order Granting Accelerated Approval of a Proposed Rule Change to 33

15

FROZEN CHARTERS shares, the number of uninstructed shares not voting are disclosed in the company’s election results as ‘broker non-votes.’ 3. Information Memorandum 12-4 This brings us to the change in the NYSE rules which is the subject of this paper. On January 25, 2012, the NYSE announced that it would no longer treat corporate governance matters as ‘routine’ for the purposes of Rule 452. The announcement was made by Information Memorandum 12-4 (henceforth, “IM 12-4”).38 IM 12-4 stated that “[m]ore recently, the approach to broker voting of uninstructed shares has narrowed through changes in Exchange rules as well as through legislative action”, and noted the restrictions on broker voting for director elections and executive compensation. IM 12-4 then continued: In light of these and other recent congressional and public policy trends disfavoring broker voting of uninstructed shares, the Exchange has determined that it will no longer continue its previous approach under Rule 452 of allowing member organizations to vote on such proposals without specific client instructions. Accordingly, proposals that the Exchange previously ruled as “Broker May Vote” including, for example, proposals to de-stagger the board of directors, majority voting in the election of directors, eliminating supermajority voting requirements, providing for the use of consents, providing rights to call a special meeting, and certain types of antitakeover provision overrides, that are included on proxy statements going forward will be treated as “Broker May Not Vote” matters.39 As a result, since January 25, 2012,40 brokers have no longer been able to vote uninstructed shares on charter amendments, as well as bylaw amendments and other corporate governance proposals.41

Amend NYSE Rule 452 and Listed Company Manual Section 402.08 to Eliminate Broker Discretionary Voting on Executive Compensation Matters”, September 9, 2010 (henceforth, the “2010 SEC Release”) (75 Fed. Reg. 56,152, September 15). 38 NYSE Information Memorandum 12-4, “Application of Rule 452 to Certain Types of Corporate Governance Proxy Proposals” (January 25, 2012). 39 NYSE Information Memorandum 12-4, supra note 38. 40 IM 12-4 does not indicate a phase-in period or start date. It does indicate that it applies to proposals “that are included on proxy statements going forward” A number of companies that had filed proxy statements prior to January 25, 2012 for meeting that took place after that date that included charter amendments did not separate broker votes. The last of these meetings, that of Qualcomm Incorporated, took place on March 6, 2012. 41 Although all of the examples listed in IM 12-4 are amendments that reduce takeover defenses, from subsequent amendment proposals where broker votes have been excluded, it appears to have been interpreted to also apply to other kinds of corporate governance amendments, such as ones that put in place takeover defenses.

16

FROZEN CHARTERS

III. The Consequences of IM 12-4 In this part, I consider the consequences of IM 12-4. I start by explaining how broker voting can distort the outcome of shareholder votes in two ways – ‘distorted pass’ results where broker voting is permitted, and ‘distorted fail’ results where broker voting is not permitted. I first consider the ‘distorted fail’ consequences of IM 12-4. The main ‘distorted fail’ consequence of IM 12-4 is frozen charters – because of IM 12-4, many companies are unable to amend their charters. There are two related consequences – bylaws not amendable by shareholders, and vetoes by insiders. All of these consequences reduce shareholder welfare, the opposite effect from that which IM 12-4 intended. I then consider the positive effects of IM 12-4 – preventing ‘distorted pass’ results – and show why IM 12-4 has had a limited effect on increasing shareholder welfare.42

A. Broker Voting Distortions, ‘Distorted Passes’ and ‘Distorted Fails’ Broker voting causes distortions because the proportion of shares voted by brokers is likely to be different from the proportion that would be voted if the beneficial owners voted themselves. If shareholders of all shares held by brokers gave voting instructions, there would be no distortion. Some proportion of the shareholders would vote in favor of a proposal. However, if shareholders do not give voting instructions, there will be some divergence between the actual vote and how those shareholders would have voted themselves. To illustrate, let us assume that the proportion of shareholders in favor of a proposal is 60%, and that holders of 15% of the outstanding shares of the corporation do not given instructions to their broker. Consider first the situation where brokers are permitted to vote uninstructed shares. I will assume (consistent with the evidence and widely held views) that broker vote the uninstructed shares overwhelmingly as directors recommend. 15% of the shares of the company will therefore be voted by the brokers in favor of the proposal. However, had the shareholders themselves voted, only 60% of that 15%, or 9% of the outstanding shares, would have voted in favor. As a result, broker voting has positively distorted the total vote by 6%. Consider now the situation where broker voting is not permitted. Now none of the uninstructed shares are voted in favor of the proposal. As a result, prohibiting broker voting has negatively distorted the total vote by 9%. However, more important than the vote tally is the outcome of the vote, which is binary – if the proposal receives more votes than the vote requirement, the proposal will

42

The issues I am discussing in this Part are separate from those discussed in the existing literature on the problem of distortions and the prevention of efficient changes that come from manager control of charter amendments. See, e.g., Lucian Bebchuk, The Case for Increasing Shareholder Power, 118 Harv. L. Rev. 833-917 (2005); Lucian Bebchuk, Limiting Contractual Freedom in Corporate Law: the Desirable Constraints on Charter Amendments 102 Harv. L. Rev. 1820-1860 (1989).

17

FROZEN CHARTERS pass, otherwise it will fail. This binary outcome can be distorted in two different ways, analogous to the analytical concepts of a ‘false positive’ and a ‘false negative.’ Either a proposal could pass even though the undistorted proportion of votes would cause it to fail; or a proposal could fail even though the undistorted proportion of votes would cause it to pass. Because of the assumption that brokers vote overwhelmingly as directors recommend, broker voting can only be distortive in the direction of director recommendations – which are almost exclusively in favor of charter and bylaw amendments.43 As a result, permitting broker voting can cause a ‘distorted pass’ in circumstances where shareholders would prefer that the proposal fails, but broker votes increase the vote beyond the vote required and cause it to pass. In contrast, if shareholders preferred the proposal to pass and broker votes increased the vote, the proposal would still pass and there would be no distortion in the outcome. To illustrate, assume that a proposal is supported by 45% of shareholders, and requires 50% to pass. If the outcome is not distorted, the proposal would fail. However, if shareholders of 15% of outstanding shares do not instruct their broker how to vote, and the brokers vote 100% of the shares in favor of the proposal, then the percentage of shares in favor will be 85% x 45% + 15%, or 53.25%, and the result will be a ‘distorted pass.’ The converse is true where broker voting is not permitted. If more than the required proportion of shareholders prefer that a vote passes, but the prohibition on brokers voting uninstructed shares means that less than the required proportion is actually voted in favor, then the result will be a ‘distorted fail.’ To illustrate, assume that a proposal is supported by 90% of shareholders, and requires 80% approval to pass. If the outcome is not distorted, the proposal would pass. However, if shareholders of 15% of the shares do not instruct their broker how to vote, then the percentage of shares in favor will be 0.85% x 90%, or 76.5%, and the result will be a ‘distorted fail.’ Following from this analysis, the effect of IM 12-4, in disallowing broker votes, has been to eliminate the possibility of ‘distorted passes,’ but to allow the possibility of ‘distorted fails.’ Sections B and C consider these two kinds of consequences.

B. ‘Distorted Fails’ Resulting from IM 12-4 1. Frozen Charters The main type of ‘distorted fail’ outcome from IM 12-4 is frozen charters. Broker votes represent, on average, 10.4% of the outstanding shares of corporations in my sample. For many corporations, particularly those with high supermajority requirements for certain charter amendments, turnout at annual meetings is such that the corporations are unable to reach those requirements without broker votes. Therefore, as a result of

43

There is a very small number of cases where directors have given no recommendation, or recommended against, corporate governance proposals they have approved. See the discussion in Part V.D and Table 11 for further details.

18

FROZEN CHARTERS IM 12-4, those corporations are no longer able to amend certain parts of their charters. Their charters are frozen. Although the great majority of charter amendments that go to a vote pass, a significant number of companies have failed in their attempts to amend their charters since IM 12-4 came into effect. Of the 433 proposals to amend corporate charters put forward by corporations44 in 2012 and 2013, 83 (19.1%) have failed. Of those, 54 (12.5% of the total) have failed despite receiving greater than 90% support.45 The charters of these companies are frozen – despite the support of directors and an overwhelming majority of shareholders, the corporations are unable to amend these parts of their charters. These amendments generally related to the removal of takeover protections, such as declassification of the board or reductions in supermajority requirements to amend the company’s charter or bylaws or take other corporate action.46 This can also be expressed algebraically. Throughout this paper, I will use α to represent a proportion of outstanding shares, with a subscript to represent the numerator – so αfor for the proportion of outstanding shares in favor, αag for the proportion of outstanding shares against, αbv for the proportion of outstanding shares that are uninstructed broker votes (or non-votes), and αreq for the proportion of outstanding shares required for a vote to be approved. I will use β to represent a proportion of votes cast (i.e., votes cast for, against and abstained, but not including broker votes), for instance, βfor for the proportion of votes cast in favor. Where necessary, I will use θ for the raw number of votes – for instance, θfor for the number of raw votes cast in favor. Using these symbols, we can say that a company has a frozen charter if the percentage of outstanding shares voted in favor (αfor) for the company is less than the percentage of outstanding shares required (αreq) for the company, and the percentage of votes cast in favor of the proposal (βfor) is greater than 90%, i.e.: βfor > 0.9 and αfor < αreq

44

This excludes shareholder proposals, as they cannot, by themselves, effect amendments to charters, and since the overwhelming majority of them are precatory, generally do not, by themselves, effect changes to bylaws or corporate governance policies. 45 Data is drawn from the FactSet TrueCourse, Inc. SharkRepellent.net Proxy Database. The FactSet SharkRepellent database contains data for approximately 4,000 U.S. public companies, including those in the Russell 3000 indices. The Russell 3000 index covers the largest 3,000 U.S. companies, which according to Russell Investments, which publishes the index, comprises 98% of the investable U.S. equity market; as of May 31, 2012, the market capitalization of companies in the Russell 3000 index ranged from $540 billion to $101 million (see http://www.russell.com/indexes/data/fact_sheets/us/russell_3000_index.asp.). Although there are approximately 700 additional companies covered by SharkRepellent, because the database was originally set up to track poison pills, many of the companies includes that are outside the Russell 3000 index are included because they have poison pills. To avoid biasing my sample, I limit my consideration here to companies in the Russell 3000 at the time of the meeting where the proposal was considered. Since charter amendment rules are governed by place of incorporation, I exclude non-U.S. companies from my sample. 46 Many of the other proposals that failed (i.e., those that received less than 90% support) were attempts to add takeover protection of the kind that are generally disfavored by shareholders.

19

FROZEN CHARTERS To what extent are these frozen charters due to IM 12-4? Consistent with the views of the NYSE Proxy Working Group and the evidence of Bethel and Gillan that brokers overwhelmingly follow management recommendations, let us assume that 100% of broker votes (αbv) would vote in accordance with management recommendations if they are permitted to vote. Using this assumption, we can describe the counterfactual situation where IM 12-4 had not come into effect. In that case, the percentage of outstanding shares in favor for each company would be the percentage actually voted in favor (αfor) plus the number of broker votes (αbv). A frozen charter will be the result of IM 12-4 where: βfor > 0.9, αfor < αreq and αfor + αbv > αreq

Take, for instance, the example of the management proposal of Akamai Technologies, Inc. in 2013. The proposal received support (αfor) of 73.2% of the shares outstanding. However, the voting requirement for the proposal to pass (αreq) was 75%. Uninstructed broker votes (αbv) represented 10.9% of shares outstanding. Had those shares been voted in favor of the proposal by brokers, the total in favor (αfor+ αbv) would have been 84.1%, and the proposal would have passed. Panels 1 and 2 of Appendix A lists the companies where charter amendments have failed since IM 12-4 in 2012 and 2013 (respectively), and show that of the charter amendments voted on since IM 12-4, 35 of the 42 companies47 (83%) where proposals failed despite receiving at least 90% support would have had their proposals pass if IM 12-4 had not applied. To put this into historical proportion, I calculate the proportion of management proposals that failed despite receiving greater than 90% support of votes cast before IM 12-4 came into effect and after it came into effect, which is shown in Figure 3, below.

47

One of the 43 companies listed in Panels 1 and 2 of Appendix A, Chesapeake Energy Corporation, had proposals fail despite receiving more than 90% support in both 2012 and 2013.

20

FROZEN CHARTERS Figure 3: Percentage of Companies Bringing Proposals Where Proposals Failed Despite Receiving >90% Shareholder Support 16% No broker votes

14% 12% 10% 8% 6% 4%

With broker votes

Average 2005 - 2011

2% 0% 2005

2006

2007

2008

2009

2010

2011

2012

2013

The average percentage of companies with proposals receiving 90% support that had proposals fail from 2005 to 2011 was 3.0%. Had broker votes been permitted in 2012 and 2013, I estimate that 2.1% and 1.1% of companies with proposals receiving 90% support would have failed, respectively. However, without broker votes, the actual percentages were 10.8% and 14.9%, respectively. Another way of examining the effect of IM 12-4 on frozen charters is to consider whether the likelihood of failure for proposals increased following the implementation of the rule. To consider this, I construct a regression model for the relationship between whether the vote passes or fails, as the dependent variable, and whether the vote took place in 2012 or 2013 (a binary variable taking the value 1 for 2012 or 2013 and 0 for previous years). The model is described in Appendix B. I find that whether or not the vote took place after IM 12-4 is strongly significant (at the 1% level) in determining the probability of the vote failing. I estimate that the odds of a proposal failing increase 8.2% if the proposal took place after IM 12-4. Together with the results above, these results show that IM 12-4 had a significant, and economically meaningful, impact on the likelihood of amendment proposals failing, and the number of amendment proposals that have failed. 2. Bylaws Not Amendable by Shareholders So far I have focused on the effect of IM 12-4 on charter amendments. The rule change also applies to bylaw amendments, which differ from charter amendments in several important respects.

21

FROZEN CHARTERS While board approval is necessary to amend a charter,48 state law allows the bylaws of a corporation to be amended by action of shareholders, without the approval of the board.49 In a significant number of corporations, bylaws provide that they can be also amended by the board of directors without a vote of shareholders.50 Because bylaws can be amended by shareholders without the assent of the board, they provide a degree of protection for shareholders in the event of a disagreement between the shareholders and directors of a corporation regarding whether a certain bylaw amendment may be in the best interests of shareholders. In such a situation, shareholders could amend the bylaws of their own volition, without the assent of the board.51 However, in the same way that IM 12-4 has resulted in frozen charters, it has also resulted in shareholders of many corporations being unable to amend the bylaws of the corporation because of the restriction on broker voting. If the voting requirement for shareholders to amend part of a corporation’s bylaws (αreq) is greater than the shareholder turnout (αto)52, then it will be impossible for shareholders to amend the bylaws of the corporation. This state is the result of IM 12-4 where turnout plus broker votes (αbv) is also greater than the votes required – i.e.: αto < αreq and αto + αbv > αreq Panel 3 of Appendix A lists the companies where bylaw amendments have failed since IM 12-4 came into effect that received greater than 90% support of votes cast, and shows that 5 of the 6 companies would have had their bylaw amendment pass if IM 12-4 had not applied.53 This is not as significant an effect as the frozen charters described above, since the number of bylaw amendments put forward for shareholder approval is lower, and since in some of those cases, boards could act unilaterally to amend the bylaws. However, as above, the effect has been to reduce the number of corporate governance changes preferred by shareholders. 3. Insider Vetoes In the same way as IM 12-4 has had the effect of making it impossible to amend parts of the charters of many corporations, it has also created new vetoes by insiders – i.e., situations in which all of the shareholders could formerly have amended the charter if

48

e.g., 8 Del. C. § 241. e.g., 8 Del. C. § 109(a). 50 A notable exception in Delaware is for bylaw provisions classifying the board of the corporation, which can only be added upon a vote of stockholders of the corporation – see – 8 Del. C. § 141(b). 51 In practice, instances of shareholders attempting to unilaterally amend the bylaws of a corporation are rare. Only 7 shareholder proposals submitted in 2012 (2.6% of all shareholder proposals submitted that year) were binding bylaw proposals, and no proposals submitted in 2013 were binding. For a discussion of the reasons why this is the case, see Lucian Bebchuk and Scott Hirst, Private Ordering and the Proxy Access Debate, 65 Bus. Lawyer 329 (2010). 52 Note that shareholder turnout (αto) excludes broker votes. 53 Of the four other companies, one ModusLink Global Solutions, Inc. failed to disclose the number of broker votes, and another, L-3 Communications Holdings, Inc. claimed that a unanimous vote was required to approve the proposal. 49

22

FROZEN CHARTERS they so desired, but whereby an insider block can now prevent the company from amending its charter. Take, as an example, the case of Cerner Corporation. At its 2013 annual meeting of shareholders, the company put forward a management proposal to amend its charter to declassify its board of directors.54 The company has in place a supermajority requirement requiring a vote of 80% of outstanding shares to amend the relevant provision of its charter. The proposal received support of 73.2% of shares outstanding (representing 86% of the votes cast), and failed. 6.9% of shares outstanding were held by brokers, who did not vote. With broker votes not voted, there were only 93.1% of the shares outstanding available to vote. According to its 2013 proxy statement, the officers and directors of Cerner held 13.7% of the outstanding shares of the corporation. If all of the directors and officers opposed an amendment proposal, even if all other shareholders voted in favor of the proposal, it could receive at most 79.4% of the vote, insufficient to amend the relevant provisions of the charter. As a result of Rule IM 12-4, directors and officers had a veto over amendments to those provisions of the charter. Expressed algebraically, Rule IM 12-4 will result in an insider veto where the shareholder turnout (αto) less the proportion of shares held by insiders (αins) is less than the vote requirement (αreq), but the shareholder turnout less the proportion of shares held by insiders plus uninstructed broker votes (αbv) would be greater than the vote requirement: αto - αins < αreq and αto - αins + αbv > αreq

The possibility of insiders vetoing charter amendments is especially problematic because in a set of potential charter amendments, managements’ preferences regarding amendments may diverge acutely from the preferences of the substantial majority of shareholders. These cases are where the company’s charter contains anti-takeover measures, such as a classified board. These represent a very large proportion of the charter amendment votes that fail despite receiving strong shareholder support.55 Where a company has an anti-takeover provision in its charter, insiders may have different interests regarding amendment of that provision from other shareholders. It is possible that other shareholders may prefer to amend the charter to remove such anti-takeover defenses, as the threat of a takeover may encourage management to perform more effectively, and shareholders may benefit from the potential premium paid in the case of a takeover. Consequently, shareholder proposals put forward in 2013 requesting that companies remove classified boards, a key takeover defense, received average support of

54

For full disclosure, this proposal resulted from engagement by a client represented by the Shareholder Rights Project, with which I am affiliated. For further information, including a full list of corporations bringing charter amendments as a result of engagement by clients represented by the Shareholder Rights Project, see Lucian Bebchuk, Scott Hirst and June Rhee, “Towards the Declassification of S&P 500 Boards”, supra note 1. 56 See Georgeson, 2013 Annual Corporate Governance Review, available at http://www.georgeson.com/us/resource/Pages/acgr.aspx, at 22.

23

FROZEN CHARTERS 80% of votes cast.56 However, managers are likely to prefer not to amend such charter provisions, for corresponding reasons – the threat of a takeover that might lead to their replacement will reduce their job security, and put more pressure on them to perform than they may otherwise prefer. In these cases, a partially-frozen charter allows management to veto a charter amendment that would otherwise be value-enhancing for the substantial majority of shareholders of the company. So far I have been describing the set of negative consequences of IM 12-4’s elimination of broker voting, the ‘distorted fail’ results. I now turn to consider the positive effects IM 12-4 has had in eliminating ‘distorted pass’ results.

C. The Reduction in ‘Distorted Pass’ Resulting from IM 12-4 As I described in Part II, IM 12-4 restricted broker voting on corporate governance matters such as charter amendments based on a concern that – because brokers generally follow management recommendations – broker voting could distort the outcomes of shareholder votes where shareholders disagree with management recommendations. In the terminology put forward in Section A, these are ‘distorted pass’ results. In this section I consider the positive consequences of IM 12-4 in eliminating such ‘distorted pass’ proposals. In order to evaluate this scenario, it is necessary to consider the preferences of the shareholders as a whole, including those who hold their own shares and do not vote, and those that hold their shares through a broker and do not instruct their broker how to vote. Because the non-voting shareholders do not vote, their preferences are necessarily unknowable. It is also difficult to gather data on whether the characteristics of non-voting shareholders differ from other shareholders, and therefore whether their preferences may differ from other shareholders. In the absence of any basis on which to believe otherwise, I will assume that the preferences of shareholders that do not vote (and do not instruct their brokers to vote) are the same as the preferences of the shareholders that do vote. As a result, I take the proportion of votes cast that are in favor of a proposal (βfor) as an indicator of the preferences of all shareholders with respect to the proposal. The method for determining whether a vote would have been positively distorted by broker votes will vary between companies that have ‘shares outstanding’ and ‘votes cast’ requirements. For a company with a vote requirement that is a percentage of shares outstanding, a vote would be distorted by broker votes if less than 50% of votes cast are in favor of the proposal, but the votes cast in favor of the proposal (αfor) plus broker votes (αbv) would be greater than the vote requirement (αreq), i.e.,: βfor < 0.5 and αfor + αbv > αreq For those companies that have a vote requirement that is a percentage of votes cast, a vote would be distorted by broker votes if less than 50% of votes cast are in favor 56

See Georgeson, 2013 Annual Corporate Governance Review, available at http://www.georgeson.com/us/resource/Pages/acgr.aspx, at 22.

24

FROZEN CHARTERS of the proposal, and the number of votes cast in favor and broker votes (as a proportion of all votes cast and broker votes - βfor+bv)57 is greater than the vote required (βreq), i.e.: βfor < 0.5 and βfor+bv > βreq Appendix C shows the seven proposals put forward in 2012 and 2013 that received support of less than 50% of votes cast. Of those proposals, six required a majority of votes cast for approval,58 and one required a majority of outstanding shares. As Appendix C shows, allowing broker voting would have distorted the outcome of only one of these proposals, that of Pacific Sunwear of California, Inc. The proposal required a majority of votes cast. 48% of votes cast were in favor of the proposal, which failed. Had the 12% of outstanding shares held by brokers been voted in favor of the proposal, 56% of votes cast would have been in favor, and the proposal would have passed. Note that the percentage of votes cast was already very close to 50%. In the other six instances, the vote was not close enough to 50% for broker votes to have distorted the outcome. As a result, it is difficult to conclude that the changes to broker voting in IM 12-4 have had a significant effect in preventing distortion of shareholder voting on management proposals. The very small number of management proposals where a majority of votes cast were against the proposal is important. A management proposal requires the approval of directors. Bringing a management proposal is costly – directors must spend time considering the proposal, may get legal advice regarding the proposal, and must oversee drafting of disclosure regarding the proposal for the proxy statement. Having a management proposal fail may also have negative reputational costs for directors and managers. As a result, directors are unlikely to put forward a management proposal that they think is unlikely to succeed. Shareholders generally follow directors’ recommendations, unless they have reason to believe that directors interests differ from their own (such as with takeover defenses, where directors and managers may have selfinterested reasons for their recommendation). Consistent with this theory, all seven failing proposals that received less than 50% support from shareholders were proposals to authorize takeover defenses.59 As a result of these factors, there are very few management

57

Since adding broker votes changes the denominator as well as the numerator of the percentage of votes cast, βfor+bv is calculated as the it total number of votes cast in favor (θfor) plus the total number of broker votes (θbv), divided by the total number of votes cast for (θfor), against (θag), abstained (θab) and by brokers (θbv), i.e.: θ +θ β = θ + θ + θ + θ

58

Five proposals were to approve the use of a shareholder rights plan (or poison pill); because shareholder rights plans are not contained in the charter, they only require the approval of a majority of votes cast. Green Plains Renewable Energy, Inc. is an Iowa company, and Iowa follows the MBCA, the vote requirement to amend the company’s charter is a majority of votes cast. 59 Well-advised corporations are likely to understand that, in the absence of a large blockholder who is in favor of the proposal, the likelihood of such proposals passing may be low, and are less likely to bring a proposal. Six of the seven companies – all except Cameron International

25

FROZEN CHARTERS proposals that fail because of low levels of shareholder support, and so the likely magnitude of the risk of distortion by the inclusion of broker votes is extremely limited. Figure 4 below shows the distribution of shareholder support for the 569 management proposals that have been voted on at U.S.-incorporated Russell 3000 companies since the implementation of IM 12-4. Figure 4: Distribution of Shareholder Support for Management Proposals 2012-2013

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

As Figure 4 shows, most management proposals receive very high levels of support - the median level of support was 98.6%. Only the seven proposals listed in Appendix C, 1.2% of the total, received support of less than 50% of votes cast. Since proposals can only be distorted if they receive less than 50% support, and the likelihood of receiving less than 50% support is extremely low, the chances of distortion occurring as a result of broker voting are also extremely low: the single management proposal since IM 12-4 came into effect that could have resulted in distortion, that of Pacific Sunwear of California, Inc., represents 0.18% of the management proposals brought during this period. This is significantly lower than the chances of a company having its charter frozen by an order of magnitude.

Corporation – are small capitalization companies (outside the Russell 1000), and may have not have had access to high quality advice regarding the likelihood of success of the proposals.

26

FROZEN CHARTERS

IV. Estimating the Companies Affected by IM 12-4 In Part III I confined my analysis to amendments that have been brought to a vote since IM 12-4 was implemented. However, since only a small proportion of companies have brought such proposals to a vote (15.7% of the sample I consider in this part), the results presented in Part III underestimate the true effects of IM 12-4. In this part, I consider the effects of IM 12-4 on those companies that have not brought such a proposal to a vote. I consider the consequences explained in Part III – frozen charters, insider vetoes, board-only bylaw amendments, and the potential reduction in distortion from broker votes – and estimate the companies that are affected by each consequence. This allows a full consideration of the negative and positive consequences of IM 12-4, which I consider in Part V, below.

A. Companies with Frozen Charters from IM 12-4 In circumstances where there has not yet been a charter amendment proposed, we cannot be certain how an amendment would fare, or the effect of IM 12-4 on the likelihood of such an amendment. However, we can be certain that a potential amendment would fail if all of the shareholders that would vote (the shareholder turnout, or αto) voted in favor of the amendment, and such shareholders would still be insufficient to meet the voting requirement necessary to amend the charter (αreq).60 Expressed algebraically, we can be certain that a company61 has a frozen charter if: αto < αreq62 I gather the voting requirements for each U.S. company in the Russell 3000 index, including whether the company has a supermajority provision to amend its charter or bylaws, and if not, the default charter amendment requirement in its state of incorporation. In order to estimate the effects of broker voting on potential charter amendments, it is also necessary to estimate shareholder turnout and broker votes for a potential charter amendment. Since Rule 452 was amended to prevent broker voting on director elections in 2010, corporations have been required to disclose broker vote non-vote figures for director elections.

60

This assumption means the analysis below will necessarily underestimate the number of frozen charters, since it will exclude proposals that would pass if they received 100% support, but would fail if they received between 90% and 100% support. 61 This assumes that the company has a ‘shares outstanding’ standard. Though since almost all ‘votes cast’ standard require a majority of votes cast, if more than 90% of votes cast are in favor of a proposal at such a company, the proposal will pass. 62 This is consistent with the analysis above, where I used αfor < αreq and βfor > 0.9. αfor can be expressed as αto. βfor, so the first condition is αto. βfor < αreq. Since we are assuming βfor = 1, the first condition simplifies to αto < αreq. Similarly, assuming that βfor = 1, the second condition, βfor > 0.9, is always satisfied.

27

FROZEN CHARTERS However, using these figures as estimators of turnout for a potential charter amendment proposal presents two potential problems. Director election turnout (and charter amendment turnout) may vary from year to year. To be conservative, I therefore use as an estimator the year with the largest turnout for each company over the period 2010 to 2013. A second problem may occur if there is systematically higher turnout for meetings with votes involving charter amendments than for meetings without charter amendment votes. Intuitively, it makes sense that if a company has an amendment proposal on the ballot and is concerned that the amendment may fail, management of the company may employ a proxy solicitor to try and get a greater number of shareholders to fail. To determine whether it is necessary to adjust for this effect, I construct a regression model of whether the presence of a corporate governance proposal affects shareholder turnout. The analysis is set out in Appendix D. The results show that the presence of a corporate governance proposal on the ballot at an annual meeting does not have a significant effect on shareholder turnout at the annual meeting. I limit my sample to companies that are currently in the Russell 3000 index. I exclude companies with missing turnout data. I also exclude meetings with contested elections, where shareholder votes may be split with competing candidates. I also eliminate companies where cumulative voting or multiple classes of shares with different voting rights makes it difficult to estimate likely shareholder amendment turnout.63 This leaves a sample of 2,433 companies. For each of these companies, I determine if shareholder turnout is less than the vote required to amend the charter, and if so, whether shareholder turnout plus broker votes would be greater than the vote required to amend the charter – i.e.: αto < αreq and αto + αbv > αreq I find that, of the 2,433 companies, IM 12-4 has caused 283 companies, or 11.6% of the sample, to be unable to amend part of their charters. These results are set out in Panel 1 of Appendix E shows the distribution of these companies by their market capitalization and voting requirements of the companies.

B. Companies Where Shareholders Cannot Amend Bylaws from IM 12-4 I use the same methodology as in Part A to determine the number of corporations where IM 12-4 has made it impossible for shareholders to amend the bylaws of the corporation. As before, I evaluate the number of companies where: αto < αreq and αto + αbv > αreq64

63

Although these companies are not identified as such in the SharkRepellent database, I identify them as those where with significant variation in turnout among directors – which I evaluate as the standard deviation of director turnout greater than 10% of the mean director turnout. 64 In this instance I use αreq to represent the vote required for the bylaw amendment, rather than the charter amendment.

28

FROZEN CHARTERS The default requirement in most states for a bylaw amendment is only a majority of votes cast,65 however a majority of companies provide for supermajority requirements for shareholder votes to amend certain bylaw provisions.66 The FactSet SharkRepellent database provides data regarding supermajority requirements for bylaw amendment. However, the database does not differentiate between corporations that require a majority of outstanding shares to amend certain bylaw provisions, and those that require a majority of votes cast. As a result, to be conservative, I assume that all companies without a supermajority require only a majority of votes cast. As a result, my analysis is likely to significantly underestimate the number of companies with board-only bylaw amendments.67 Of the 2,433 companies in my sample, my analysis shows that in 209 (8.6% of my sample), IM 12-4 has made it impossible for shareholders to amend the bylaws of the corporation. Panel 2 of Appendix E shows the distribution of these companies by market capitalization and bylaw amendment requirement.

C. Companies with Insider Vetoes from IM 12-4 As with frozen charters, it is possible to estimate the number of companies where IM 12-4 has given insiders a veto over certain charter amendments. Based on the same simplifying assumption that all shareholder support a particular resolution (βfor = 1), insiders will have a veto as a result of IM 12-4 where turnout less the insider block (αins) is less than the vote required, but turnout plus broker votes less the insider vote would be greater than the vote required – i.e.: αto - αins < αreq and αto - αins + αbv > αreq Using the same methodology and data as above (including insider holdings taken from the FactSet SharkRepellent database), I estimate that IM 12-4 has resulted in 227 companies, or 9.3% of my sample, where insiders now have a veto over certain charter amendments. Panel 3 of Appendix E shows the distribution of these companies by market capitalization and voting requirements. Of course, since shareholder support for a particular resolution will be less than 100%, (βfor < 1), there will be a greater set of companies where IM 12-4 has given insiders a de facto veto over certain amendments.

65

e.g., 8 Del. C. § 109. This assumes that the likelihood of that level of support is independent of the ‘distortability’ of the company submitting the proposal. In reality, companies that are distortable may be slightly more likely to put forward lower support proposals, and therefore the estimates in Table 13 may underestimate the level of expected distortion, though likely by less than an order of magnitude. 68 This assumes that the likelihood of that level of support is independent of the ‘distortability’ of the company submitting the proposal. In reality, companies that are distortable may be slightly more likely to put forward lower support proposals, and therefore the estimates in Table 13 may underestimate the level of expected distortion, though likely by less than an order of magnitude. 68

29

FROZEN CHARTERS

D. Reductions in Potential ‘Distorted Pass’ Results from IM 12-4 In the same way that I extended my consideration of frozen charters to the large majority of companies that have not had charter amendments go to a vote, it is possible to consider potential ‘distorted pass’ distortions that have not yet had proposals voted on. As discussed in Part III.D above, broker votes would positively distort the outcome of a proposal if less than a majority of voting shareholders vote for the proposal (βfor < 0.5) yet the proposal would pass if broker votes were included. As in Part III.D above, it is necessary to consider companies with ‘shares outstanding’ and ‘votes cast’ requirements separately. For companies with outstanding share requirements, the proposal will be distorted if: βfor < 0.5, αfor < αreq and αfor + αbv > αreq Since the percentage of outstanding for is the same as the turnout multiplied by the percentage of turnout for (αfor = βfor.αto), and since βfor cannot be greater than 0.5, we in the most generous case the above conditions will be true if: βfor < 0.5, 0.5.αto < αreq and 0.5.αto + αbv > αreq For companies with votes cast requirements, broker votes could conceivably cause a proposal to result in a ‘distorted pass’ if the absolute number of shares in favor (θfor), plus the number of broker votes, (θbv) as a percentage of the number of votes cast (θto) plus the number of broker votes, is greater than the vote required, βreq: βfor < 0.5 and





Since the number of shares cast in favor cannot be more than half the number of shares cast (θfor < 0.5. θto), in the most generous case this simplifies to: βfor < 0.5 and

. .



I find that 463 companies (19.0%) could potentially have a ‘distorted pass’ result. Panel 4 of Appendix E shows the distribution of these companies, by type of vote requirement and supermajority requirement. However, these are only potentially distortable companies. To estimate the likely number of ‘distorted pass’ distortions, it is necessary to consider the level of support that would be necessary for the companies above to have votes distorted, and the likelihood of those levels of support occurring. The level of support necessary for the vote to be distorted will be a function of level of broker votes and the level of turnout for the vote. The lower the broker votes as a proportion of the turnout, the closer the level of support must be to 0.5 for the outcome to be distorted. For companies with outstanding shares requirements, rearranging the formula above, we can see that: β

>

α

30

−α α

FROZEN CHARTERS For companies with votes cast requirements: β

>

β

θ +θ θ

− θ

Table 2 below shows the distortable companies by the range of support at which they could be distorted. Table 2: Expected Number of Distorted Companies Minimum Support for Distortion

Companies

Range of Support for Distortion

Likelihood of Support in Range

Expected Number of ‘distorted pass’ Companies

45%-50%

289

45%-50%

0.5%

1.6

40%-45%

83

40%-50%

0.9%

0.8

35%-40%

47

35%-50%

1.1%

0.5

30%-35%

15

30%-50%

1.1%

0.2

20%-30%

15

20%-50%

1.1%

0.2

10%-20%

10

10%-50%

1.3%

0.1

0%-10%

4

0%-50%

1.3%

0.1

Total

463

3.3

As well as the breakdown of distortable companies by the minimum level of support necessary for distortion, Table 2 also shows the likelihood of proposal support being between that level and 0.5, based on the proportion of all management proposals that received support in that range in 2012 or 2013.68 By multiplying this likelihood by the number of companies requiring that level of support, we can estimate the number of companies that are likely to have a ‘distorted pass’ outcome as a result of broker votes. As Table 2 shows, only 3.3 companies, or 0.1% of the companies in my sample, can be expected to have a ‘distorted pass’ outcomes as a result of broker votes. This is an artifact of the low number of management proposals that are likely to receive less than 50% of shareholder support that are brought by directors. However, if this were to change, the number of potential ‘distorted pass’ proposals would also change. As a result, since the possibility can’t be ruled out, any solution to the problem of distortion should minimize the risk of ‘distorted pass’ distortions as well as reducing the incidence of ‘distorted fail’ distortions.

68

This assumes that the likelihood of that level of support is independent of the ‘distortability’ of the company submitting the proposal. In reality, companies that are distortable may be slightly more likely to put forward lower support proposals, and therefore the estimates in Table 13 may underestimate the level of expected distortion, though likely by less than an order of magnitude.

31

FROZEN CHARTERS

V. Evaluating IM 12-4 I turn now to consider IM 12-4 from a normative perspective. I evaluate IM 12-4 from several frames of references. First, consistent with its own goals, I consider IM from an investor perspective. Second, I consider IM 12-4 from a firm value perspective. Third, I consider IM 12-4 from the perspective of the contractarian theory that underlies corporate law. Fourth, I consider IM 12-4 from the perspective of directors and managers. I conclude with some words about the procedure by which IM 12-4 was put in place.

A. Evaluating IM 12-4 from an Investor Perspective Although it doesn’t state it succinctly, IM 12-4 makes clear that its underlying goal is investor protection. Because IM 12-4 is so short – about half a page of text – it doesn’t set out the reasons underlying its prohibition on uninstructed broker voting for certain corporate governance proposals. Instead, it references “recent changes in Exchange rules as well as through legislative action,” gives the example of the 2010 prohibition on broker voting of uninstructed shares in the election of directors and executive compensation matters, and indicates that the changes are being made “in light of these and other recent congressional and public policy trends disfavoring broker voting of uninstructed shares.”69 To understand the goals of IM 12-4, it is therefore necessary to refer to the reasons underlying the 2010 changes prohibiting uninstructed broker voting on director elections and executive compensation. The underlying rationales cited in the SEC orders approving the 2010 prohibitions on uninstructed broker voting on director elections and broker voting on executive compensation were to “better enfranchise shareholders”70 and thereby “further investor protection and the public interest” 71 and “enhance corporate governance and accountability to shareholders.” 72 I therefore consider first how IM 12-4 fares against its own implicit goal of investor protection. Taking into account the consequences of IM 12-4 outlined in Parts III and IV, how does IM 12-4 fare when evaluated against these aims? Here it is necessary to weigh the benefits of IM 12-4 for investors in eliminating ‘distorted pass’ results, with its costs to investors in causing ‘distorted fail’ outcomes for proposals that investors consider to be value enhancing. As Part III shows, the number of ‘distorted fail’ outcomes resulting from IM 12-4, particularly frozen charters, clearly outweighs the small number of ‘distorted pass’ results that IM 12-4 has prevented. And on a prospective basis, as described in Part IV, the number of companies that are likely to have frozen charters without broker voting, and other ‘false fail’ outcomes likely to result from the elimination of broker voting, clearly outweighs the small number of companies where broker voting

69

See Information Memorandum 12-4, supra note 38, at 1. See the 2009 SEC Release, supra note 36, at 12. 71 See the 2010 SEC Release, supra note 37, at 9. 72 See the 2010 SEC Release, supra note 37, at 9; see also the 2009 SEC Release, supra note 36, at 12 (“enhance corporate governance and accountability”). 70

32

FROZEN CHARTERS could result in ‘distorted pass’ outcomes (assuming the distribution of support for corporate governance proposals remains constant). By freezing charters, IM 12-4 has had the effect of disenfranchising these shareholders. To the extent that these shareholders believe that these amendments would enhance corporate governance and accountability to shareholders, IM 12-4 has prevented such enhancement. As Sections B and C of Part III illustrated, there are other ways in which, rather than protecting shareholders from distortion in favor of insiders, IM 12-4 has actually harmed investors – by giving certain insiders veto power over certain amendments that shareholders may believe to be in their interest, and by taking away from shareholders the ability to amend certain bylaws. How has IM 12-4 so clearly failed against the policy goals it references? Charter amendment votes differ in two important respects from director elections and executive compensation votes IM 12-4 refers to, neither of which are acknowledged or considered in IM 12-4. First, as noted in Section III.D above, in most cases there is general alignment between director recommendations and shareholder preferences for charter amendments, because directors are unlikely to bring charter amendments if they believe they will not receive shareholder support. In contrast, with broker voting in director elections and on compensation matters, there is the possibility of withhold campaigns against directors, or moves to reject compensation plans put forward by directors – in which case there would be an obvious conflict between the recommendations of directors and the preferences of objecting shareholders. Second, there is also a significant difference among the requirement for votes to pass. For uncontested director elections, the default rule in almost all states73 is that a director will be elected either if there is a plurality.74 Even for those corporations that have adopted majority voting standards, a director will be elected, if the director receives a majority of the votes cast.75 Similarly, an executive compensation proposal will pass if it receives a majority of the ‘for’ and ‘against’ votes cast. However, as discussed in Section III.A, the default rule for most charter amendments is to require a majority of shares outstanding.76 As was explained in Part II above, many corporations have supermajority requirements requiring a higher proportion of votes outstanding. With a ‘votes cast’ standard, preventing broker voting will have a limited effect on the chance of the proposal passing, as broker votes will be eliminated from both the numerator and denominator of the proportion of votes cast. Indeed, Akyol, Raff and Verwijmeren found that the 2010 amendments to Rule 452 eliminating uninstructed broker voting for director elections did not decrease approval

74

That is, the director receives more votes than any other contestant. In an uncontested election with a plurality rule, a director will therefore be elected if they receive any votes at all. 74 That is, the director receives more votes than any other contestant. In an uncontested election with a plurality rule, a director will therefore be elected if they receive any votes at all. 75 A variation on this rule is a ‘resignation policy,’ where a director would be required to submit a resignation if they did not receive a majority of the votes cast. 76 This excludes those companies incorporated in states governed by MBCA-based statutes that have not overridden the default amendment rule in those states, though as discussed in Section III.A, these represent only 5% of the sample I consider.

33

FROZEN CHARTERS rates for directors.77 However, as demonstrated by the number of frozen charters observed in Part III above, with a ‘shares outstanding’ standard, preventing broker voting will have a much greater effect on the chance of the proposal passing.

B. Evaluating IM 12-4 from a Firm Value Perspective The consequences of IM 12-4 have been to prevent changes in the corporation’s governance. Given the support of managers, directors and shareholders, we can presume that these changes are value-enhancing for the corporation and its shareholders. However, in a number of these corporations, frozen charters have prevented the corporation from making such amendments. To the extent this is the result of IM 12-4, the rule change has prevented the maximization of firm value.

C. Evaluating IM 12-4 From a Contractarian Perspective Another way to evaluate IM 12-4 is from the perspective of the contractarian view of corporate law. Economists and corporate law scholars have long understood the corporation as a ‘nexus of contracts’ among different parties in the corporation.78 To the extent this analogy holds, the corporate charter is the central part of that contract. The charter defines the key terms of the contract among the corporation and its shareholders (which I will refer to as the ‘corporate contract’) – either by incorporating those terms explicitly, or by remaining silent and therefore adopting the default terms set out in state law. One of the central terms in the contract is the process for its amendment. The corporate law of most states provides certain mandatory requirements for charter amendments such as the requirement for director and shareholder approval of charter amendments, although all states allow these to be modified in certain ways in the charter, such as increasing the voting requirement for shareholder approval. Charter terms do not explicitly deal with broker voting. However, as the discussion in Part II indicates, the treatment of broker voting is central to the results of shareholder approval votes, because of its influence on the level of turnout, and because brokers overwhelmingly vote in favor of management proposals. As a result, the term of the corporate contract that deals with the requirements for shareholder approval of amendments is predicated on certain expectations about the treatment of broker voting.

77

See Ali Akyol, Konrad Raff and Patrick Verwijmeren, The Elimination of Broker Voting in Director Elections, Working Paper, available at http://ssrn.com/abstract=1973558, at 16 (“[W]e observe no decrease in approval rates after the change in Rule 452. In fact, the increase in approval rates in annual meetings after 2009 is statistically significant at the 1% level.”) 78 See, for instance, Alchian & Demsetz, Production, Information Costs, and Economic Organization, 62 Am. Econ. Rev. 777 (1972); Jensen & Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. Fin. Econ. 305 (1976); Frank H. Easterbrook & Daniel R. Fischel, The Corporate Contract, 89 Colum. L. Rev. 1416 (1989).

34

FROZEN CHARTERS The amendment terms of the charters in almost all of the companies in my sample were adopted prior to January 2012.79 These amendment terms were predicated on broker discretionary voting being permitted on charter amendments. As discussed above, it was also generally understood that brokers voted overwhelmingly in favor of management,80 and that broker votes represented, on average, 10.4% of the shares outstanding.81 The changes wrought by IM 12-4 in prohibiting broker voting on charter amendments implicitly modify the amendment requirements of corporate charters from how they were understood by directors and shareholders. To the extent this resulted in frozen charters, the expectation of directors and shareholders that the charter could be amended in a certain way has been thwarted.82 Thwarting the intention of the corporate contract is value-reducing because it results in a move away from contractual terms agreed to by all of the parties to the contractual nexus of the corporation. As a result, it is likely that changing the terms of the contract from those that were understood by the parties results in a less efficient corporate contract.83 Even if the corporate contract is not completely efficient, taking away the ability to amend the contract would make the contract more efficient only in a very narrow and unlikely set of circumstances.84

D. Evaluating IM 12-4 from Directors’ and Managers’ Perspectives Throughout this paper I have assumed that, because charter amendments require the approval of directors, directors (and managers) support such charter amendments. However, there may be reasons to believe that directors (and managers) notwithstanding

79

Ten companies in my sample successfully amended their charters to reduce or eliminate supermajority requirements for charter amendments at their 2012 annual meetings. 80 See the Proxy Working Group Report, supra note 25, at 14. 81 Jennifer Bethel & Stuart Gillan, The Impact of the Institutional and Regulatory Environment on Shareholder Voting, supra note 25, at 42. 82 The significance of this point is diminished by many other changes to the factors that influence the difficulty of amending corporate charters since those charters were entered into – for instance, the rise of institutional investors. However many such changes are endogenous to the shareholders or the corporation, whereas IM 12-4 is exogenous. 83 A long-standing debate in corporate law concerns whether the initial corporate contract is efficient. Corporate contracts are understood to be efficient because the parties designing the corporate contract cannot benefit from introducing inefficient terms, because the other parties are informed of their value, and will price the inefficient term accordingly. Therefore the parties to the initial charter will draft value-maximizing terms.83 If the corporate contract is efficient, then any unintended change to the corporate contract – such as a change in the effect of the amendment term – will make the contract less efficient. 84 Having the option to amend the contract is likely to be efficient, as discussed in Section B below. Evidence for the efficiency of amendment can be found in the fact that public company charters could effectively prevent amendments by requiring 100% unanimity, however only one company in my sample contains this requirement.84 In order for a charter preventing amendment to be efficient, shareholders would have to believe that there is some benefit to preventing themselves from amending the charter, and that that benefit outweighs the cost of preventing current or future amendments to the charter. Of course, despite these beliefs, shareholders could not have already taken the efficient action to prevent future amendments. In any other set of circumstances, changing the amendment term to prevent amendment will reduce the value of the company.

35

FROZEN CHARTERS approving the amendment, directors prefer that the amendments fail. There are a small number of management proposals where directors have not given a recommendation either for or against the proposal, or have recommended against the proposal. Table 3, below, sets out the number of such recommendations for charter and bylaw amendments from 2005 to 2013. Table 3: Number of ‘No Recommendation’ and ‘Against’ Recommendations by Type of Management Proposal, 2005-2013 No Recommendation

‘Against’ Recommendation

All proposals

Charter Amendments85

7

2

1,826

Bylaw Amendments

2

4

359

Other

8

1

260

Total

17

7

2,445

Although there are very few instances where directors do not recommend in favor of a management proposal,86 there are a large number of situations where management proposals approved by directors were preceded at the previous annual meeting by shareholder proposals that directors recommended against, and put forward numerous arguments against, casting doubt that the directors’ approval and recommendation of a management proposal signal their true preference. Why might directors approve amendments that they personally disfavor? This may be the case if a precatory shareholder proposal requesting the amendment has previously been approved by the shareholders of the company. The voting guidelines of many institutional investors indicate that, if that is the case, they will withhold votes from directors that fail to implement the request contained in the shareholder proposal.87 Similarly, the policies of the major proxy advisory firms are to recommend withhold votes against directors that fail to implement the request contained in a successful

85

These include amendments of both charter and bylaws. One reason there may be so few instances where directors make no recommendation, or recommend against a proposal, is because failing to recommend the proposal may cast doubt on whether the directors believe the proposal is in the best interests of the company, and therefore whether their decision to approve the proposal was consistent with their fiduciary duties. In other cases, directors may recommend in favor, but signal their own preferences against the proposal. For instance, in 2013 Costco Wholesale Corporation brought a proposal to declassify the board of the corporation, wherein the directors recommended in favor of the proposal but explained that, in their capacity as shareholders of the corporation, they intended to vote against the proposal. 89 See Georgeson, 2013 Annual Corporate Governance Review, supra note 56, at 22. 86

36

FROZEN CHARTERS shareholder proposal.88 Directors wishing to avoid having a significant proportion of votes withheld in their own elections may therefore comply with shareholder wishes that an amendment be put forward, even though their personal preference is against the amendment. This is most likely to be the case for those potential charter amendments, where directors’ (and managers’) preferences regarding amendments diverge from the preferences of the substantial majority of shareholders, such as charter amendments to remove anti-takeover measures, like classified boards. In that case, it is possible that shareholders may prefer to amend the charter to remove such anti-takeover defenses, based on the belief that the threat of a takeover may encourage management to perform more effectively, and shareholders may benefit from the potential premium paid in the case of a takeover. Consequently, shareholder proposals put forward in 2013 requesting that companies remove classified boards, a key takeover defense, received average support of 80% of votes cast.89 However, managers may prefer not to amend the charter to remove anti-takeover provisions, for corresponding reasons – the threat of a takeover that might lead to their replacement will reduce their job security, and put more pressure on them to perform than they may otherwise prefer.

E. Evaluating IM 12-4 from a Procedural Perspective Although IM 12-4 had significant consequences, its form, and the procedure by which it was put in place, differed substantially from the procedure by which other amendments to Rule 452 had previously been made. Amendments to NYSE rules (and those of other exchanges) are governed by Section 19(b) of the Exchange Act and Rule 19b-4 of the General Rules and Regulations promulgated thereunder. These set out comprehensive requirements for proposed amendments of exchange rules, including: 1. The proposed rule must be filed on a specified form, with certain required exhibits.90 2. The information provided by the exchange must enable the public to provide meaningful comment on the proposal and for the SEC to determine whether the proposed rule is consistent with the Exchange Act.91 3. The SEC is required to publish proposed rule changes and requests public comment during a comment period.92 4. The SEC must consider the proposed rule, as well as public comments on the rule, and determine whether the proposed rule satisfies the requirements of

89

See Georgeson, 2013 Annual Corporate Governance Review, supra note 56, at 22. See Georgeson, 2013 Annual Corporate Governance Review, supra note 56, at 22. 90 See Form 19b-4 to the Exchange Act Rules. 91 See Form 19b-4 to the Exchange Act Rules, supra note 90, at 2. 92 See Section 19(b)(1) of the Exchange Act. 89

37

FROZEN CHARTERS the Exchange Act and rules, including being designed “in general, to protect investors and the public interest.”93 All of the other recent restrictions on broker voting in Rule 452 – the 2003 restriction relating to executive compensation matters, and the 2010 restrictions relating to director elections and executive compensation – went through the rule amendment process, and therefore had detailed explanatory materials, public comment,94 and lengthy consideration of the rules and the comments by the SEC, on the basis of which the SEC determined the rules had the effect of protecting investors and the public interest In contrast, IM 12-4 did not follow these rules.95 IM 12-4 was a two-page memorandum, containing approximately half a page of text. IM 12-4 gave very little explanation of the reasoning behind the changes it implemented, or consideration of their potential effects. There was no advanced notice of the changes in IM 12-4, nor any ability to publicly comment. There was no consideration of the rule change by the SEC or other regulatory body, and no conclusion drawn whether the rule change would protect investors and the public interest. These procedural shortcomings may explain why the consequences described in Part III above, and the broader effects described in Part IV above, were unforeseen. Had the changes made in IM 12-4 been made through a formal amendment to Rule 452, the approval process described above may have led to these consequences being identified and avoided.

VI. Fixing IM 12-4 In this section I consider how the problems describe above may be mitigated. The aim of this section is to set out a number of potential solutions to the problem. I consider four kinds of solutions. Most obviously, the changes in IM 12-4 could be reversed. Alternatively, steps could be taken to reduce the level of uninstructed broker votes. If uninstructed broker votes cannot be eliminated, a proportional system of voting could be implemented, or a system could be devised to permit uninstructed broker voting in particular circumstances. I conclude with some comments on the procedure by which any reform should be undertaken. I evaluate each of the solutions presented below against the framework developed in Part III.A above: to reduce both ‘distorted fail’ proposals and ‘distorted

93

See Section 6(b)(5) of the Exchange Act. With respect to the 2009 SEC Release, 153 comment letters were received and reviewed. See the 2009 SEC Release, supra note 36, at 33,293. 95 Section 19 of the Exchange Act permits a proposed rule change to avoid the normal process if designed by the exchange as “constituting a stated policy, practice, or interpretation with respect to the meaning, administration, or enforcement of an existing rule of the self-regulatory organization”. 94

38

FROZEN CHARTERS pass’ proposals. As discussed in Part III.C above, there were only a small number of management proposals that received low levels of shareholder support in 2012 and 2013, which limited the positive impact of eliminating ‘distorted pass’ proposals during that period. However, it’s not possible to conclude that this distribution of shareholder support is the natural state of the world. Indeed, it’s possible to conceive of circumstances where a much larger number of proposals might receive low shareholder support.96 For the reasons set out in Part III.C, I don’t believe it’s likely that directors of a significant number of companies would bring proposals that had a significant chance of failing. However, since the possibility can’t be ruled out, any solution to the problem of distortion should minimize the risk of ‘distorted pass’ distortions as well as reducing the incidence of ‘distorted fail’ distortions. The ideal solution would therefore undo the effects of IM 12-4 in freezing corporate charters97 (as well as other ‘false fails’ for bylaw amendments, and insider vetoes), while maintaining the potential benefits of IM 12-4 in reducing ‘distorted pass’ distortions. I also comment on the potential cost and workability of the solutions.

A. Reversing IM 12-4 The most obvious solution to the problems outlined in Parts III and IV is to reverse IM 12-4, and return to the situation before IM 12-4 was implemented. This would have the opposite consequences to those described in Parts III and IV above. The ‘distorted pass’ results caused by IM 12-4 would disappear: the frozen charters caused by IM 12-4 would be defrosted, along with failed bylaw amendments, and the number of insider vetoes would be reduced. However, the possibility of ‘distorted fail’ proposals would be resurrected. Proposals that may be opposed by a majority of shareholders could conceivably pass with the support of broker votes. As discussed above, given the current state of the world where very few corporations bring management proposals that receive less than a majority of votes case, the overall effect would be beneficial – the benefit from eliminating those frozen charters and other ‘distorted fail’ proposals caused by not counting broker votes would outweigh the few instances of ‘distorted pass’ proposals that might occur. However, there remains the possibility that the number of ‘distorted pass’ proposals could increase, especially given that the reinstated potential for distortion from broker votes could increase the likelihood of such proposals passing. One solution to this could be to continue to require the disclosure of the number of uninstructed broker votes being voted, so that their distortive effect could be observed, and appropriate steps taken if ‘distorted pass’ results became a significant problem.

96

For example, during the 1980s and 1990s, a significant number of companies amended their charters to put in place takeover defenses, such as staggered boards. These types of amendments may have received significantly lower levels of shareholders support. 97 There are a number of companies that have frozen charters that are not the result of IM 12-4. Because of high supermajority requirements and/or low shareholder turnout, these companies would have frozen charters even if broker votes were permitted. Remedying these would require some other kind of intervention – for instance, court intervention to invalidate the supermajority requirement of the charter., though consideration of these is outside the scope of this paper.

39

FROZEN CHARTERS Reversing IM 12-4 could be implemented inexpensively, by the NYSE issuing further guidance reinstating charter amendments to the status of “routine items,” thereby allowing brokers discretion to vote without authority from beneficial owners. Given the simplicity of this solution and the ease by which it could be implemented, IM 12-4 could be reversed as an interim measure to reduce harm to shareholders while more comprehensive reforms are debated.

B. Reducing Undirected Broker Votes An alternative solution to the problem of distortive broker votes would be to reduce or eliminate uninstructed broker votes. If uninstructed broker votes could be eliminated, this would obviate the need to choose between ‘distorted fail’ distortions and ‘distorted pass’ distortions; both would be eliminated. Even if uninstructed broker votes could not be eliminated and could only be reduced, this would still reduce the intensity of both kinds of distortions, and therefore the likelihood that either would affect voting outcomes. The number of uninstructed broker votes could be reduced in a number of ways, some of which could be implemented by corporations, others of which could be implemented by brokers. As discussed in Part II, most uninstructed shares held through brokers are beneficially owned by retail investors. Corporations can already take steps to increase the response level from retail shareholders. Response rates among retail investors are higher when corporations mail proxy materials in full paper format, rather than electronic notification, or the ‘notice and access’ methods permitted by the SEC.98 In addition, corporations can hire proxy solicitors to telephone individual retail investors to encourage them to vote. However, neither of these techniques are a permanent solution – rather, they would need to be implemented for every annual meeting of the corporation, and would increase the cost associated with each such meetings. A more permanent solution is to reduce the number of retail investors. Many corporations have buyback programs targeted at small lots of shares. However, this requires retail investors to choose to tender into the buyback, which is unlikely to be universal. Other transactions could compulsorily acquire small shareholdings. In a reverse stock split, a corporation reduces the number of its outstanding shares by combining shares in a particular ratio. If a corporation undertook a reverse stock split with a high ratio, for example, requiring that 100 old shares be exchanged for 1 new share (a 1-for-100 reverse split), those shareholders with lots smaller than the ratio would receive cash for their shares. However, this is likely to be an expensive undertaking for a corporation, and since it will affect the number of shares outstanding, may have undesired effects on the liquidity of the company. Its costs are therefore likely to outweigh the benefits in reducing the number of small shareholders.

98

See Broadridge & PwC ProxyPulse, ‘How well do you know your shareholders’, First Edition (2013) (noting that only about 17% of retail shares receiving a notice were voted from 2007 to 2012, compared to 36% of shares receiving a full paper package).

40

FROZEN CHARTERS An alternative solution to those implemented by corporations with respect to their shareholders is for brokers to implement a solution with respect to their clients. The most promising solution is client-directed voting.99 Brokers could require their clients to direct how their shares should be voted if they fail to give instructions for a particular meeting. For example, clients could instruct that their shares be voted as management recommends, or against management’s recommendation. Alternatively, clients could choose that their shares be voted proportionally (as further discussed below). To the extent this could be implemented when clients join a broker it could significantly reduce the number of uninstructed broker shares. However, there would inevitably be some brokers that did not choose to participate. In addition, it may be difficult for brokers to require existing clients to make an election, or to force a default election upon existing clients. It’s also likely that one of the options for clients under a client-directed voting system would be to choose that their shares not be voted if uninstructed; to the extent clients chose this option, they would be continuing the potential ‘distorted fail’ results. Although several of these solutions – especially client directed voting – show promise, all could only be partially effective – none can completely reduce the number of uninstructed voting shares, and therefore the problems of ‘distorted pass’ results and ‘distorted fail’ results must be dealt with in another manner.

C. Proportional Voting An alternative approach to broker voting considered by the Proxy Working Group (and advocated by several commenters on the proposal to limit uninstructed broker voting on director elections)101 is proportional voting. This would replace the current structure whereby brokers can either vote all of the shares they hold or none of the shares they hold depending on the matter, with a system whereby brokers would be required to vote in proportion to the votes of other shareholders. For instance, if 95% of other shareholders voted in favor of a proposal, brokers would vote 95% of their shares in favor of the proposal and 5% of their shares against the proposal. 100

99

Client directed voting was discussed in the See August 27, 2007 Addendum to the Report and Recommendations of the Proxy Working Group to the New York Stock Exchange Dated June 5, 2006 (“Proxy Working Group Addendum”), at 4. For further discussion of the advantages and disadvantages of client directed voting, see Frank G. Zarb and John Endean, “Restoring Balance in Proxy Voting: The Case for “Client Directed Voting””, Harvard Law School Forum on Corporate Governance and Financial Regulation, February 14, 2010, available at http://blogs.law.harvard.edu/corpgov/2010/02/14/restoring-balance-in-proxy-voting-the-casefor-client-directed-voting/; John Wilcox, “Fixing the Problems with Client Directed Voting”, Harvard Law School Forum on Corporate Governance and Financial Regulation, March 5, 2010, available at http://blogs.law.harvard.edu/corpgov/2010/03/05/fixing-the-problems-with-clientdirected-voting/; and James McRitchie, “An Open Proposal for Client Directed Voting”, Harvard Law School Forum on Corporate Governance and Financial Regulation, July 14, 2010, available at http://blogs.law.harvard.edu/corpgov/2010/07/14/an-open-proposal-for-clientdirected-voting/. 100 See the Proxy Working Group Report, supra note 25, at 16-18. 101 See 2009 SEC Proposal, supra note 36, at 31.

41

FROZEN CHARTERS Assuming the preferences of shareholders holding their shares through brokers are the same as other shareholders, this would eliminate any distortion in shareholder voting. Consequently, it would avoid both of the types of distorted outcomes discussed above. Assuming continued high levels of shareholder support for management proposals, proportional voting would undo the frozen charters caused by IM 12-4, as well as the other kinds of ‘distorted fail’ proposals – failed bylaw amendments and insider vetoes. Since broker votes would follow the votes of other shareholders, proportional voting would also avoid any ‘distorted fail’ outcomes. Proportional voting would also obviate the need for an arbitrary list of matters on which brokers could or could not vote, such as that contained in the supplementary materials to Rule 452 – instead brokers would vote proportionally on all matters. However, proportional voting would be complicated to implement.102 Brokers would need to obtain a measure of the proportion in which vote their shares. There are two separate reference groups by which the appropriate proportion could be measured – the other shareholders of the corporation, and other shareholders that have instructed a particular broker to vote. The most obvious proportion for brokers to vote would be the proportion of votes cast by all other (non-broker) shareholders of the corporation.103 This would most accurately reflect overall shareholder preferences. However, if the preferences of shareholders as a whole were different from those shareholders who do not vote, this could result in distortions of its own. Logistically, basing proportional voting on the aggregate proportion of votes cast would require either a tabulation of the shares of all of the other shareholders prior to the broker submitting their proxies, or for brokers to submit a proxy in blank to the tabulating organization or to directors, for them to vote the uninstructed shares in the proportion that is later established. Broadridge Financial Solutions, Inc. handles the overwhelming majority of proxy statement distribution and vote handling for most corporations. Broadridge could provide preliminary vote counts to the corporation for distribution to brokers, or to brokers directly. In order for proxies to be voted by brokers, this information would need to be received several days before votes were due. This problem has been overcome with respect to broker-by-broker proportional voting (described below), and although coordinating vote tallies from multiple sources would be more difficult, this is unlikely to be insurmountable. An alternative would be for brokers to give proxies to the proxy holders to vote their shares in a proportion to be determined. Most proxy cards in uncontested elections appoint selected officers or directors of the corporation as the proxy holder, and direct the proxy holder to vote in the manner specified on the proxy card. The proxy card could indicate include an option of

102

The Proxy Working Group stated that “in many ways proportional voting creates its own set of problems,” and concluded that “it was not the optimum result.” See the Proxy Working Group Report, supra note 25, at 17-18. 103 See the Proxy Working Group Report, supra note 25, at 17.

42

FROZEN CHARTERS having the proxy holder split the shares in the proportion voted by other shareholders.104 Such a system could be implemented by SEC regulation, or alternatively, by private ordering, on a company-by-company basis. However, a private ordering solution would require each company to act individually, thereby duplicating significant effort, would involve a long timeframe in getting adoption from a substantial number of companies, and would likely result in many companies not implementing the system.105 The alternative to voting on an aggregate basis would be the proportions in which brokers would vote uninstructed shares to be obtained on a broker-by-broker basis, from the instructions those brokers receive from other beneficial owners. A broker-by-broker system would be straightforward to implement. The broker, or Broadridge acting on their behalf, could tally the instructions they received from their other beneficial owners, and then split the proxies of the uninstructed shares in the same proportions. As above, the tabulation would need to be done several days prior to the votes being cast. However, as discussed in Part II above, beneficial owners holding through brokers are already required to notify the broker at least ten days before the meeting.106 As a result, the timing issue is unlikely to be a problem. Indeed, the Proxy Working Group considered that one broker, Charles Schwab, had implemented proportional voting as early as 2005.107 Following the release of the Proxy Working Group’s report, the Securities and Financial Markets Association (SIFMA) issued a “best practices” memorandum recommending that their member brokers implement proportional voting of uninstructed shares in proportion to the votes cast by the retail clients of the broker.108 According to news reports, four large brokers – Charles Schwab, Ameritrade, Morgan Stanley, Merrill Lynch and Goldman Sachs – adopted broker-by-broker proportional voting.109 In a presentation to SIFMA in 2007, Richard Daly, chief executive officer of Broadridge, outlined how they provided proportional voting services to ‘four large broker clients’ in 2007.110 The proportion was established based on shares voted by retail customers of the broker. The proportion was calculated as of two days prior to the meeting, and then recalculated the day before the meeting and the day of the meeting for newly voted shares.

104

Most proxy cards already indicate that the proxy will be voted in a particular way if no direction is made on a proxy card with respect to a particular vote, usually following directors’ recommendations. 105 For a broader discussion of the relative merits of regulatory and private ordering solutions, see Lucian Bebchuk and Scott Hirst, Private Ordering and the Proxy Access Debate, 65 Bus. Law. 329 (2009-2010). 106 See NYSE Rule 451(b)(1), supra note 21. 107 See Proxy Working Group Report, supra note 25, at 17. 108 See Proxy Working Group Addendum, at 4. 109 See Cyrus Sanati, “A Surprise at the Ballot This Proxy Season?”, New York Times DealBook, February 25, 2009 (available at http://dealbook.nytimes.com/2009/02/25/a-surprise-at-the-ballotthis-proxy-season/). 110 See also the slides from the presentation by Richard J. Daly, Chief Executive Officer of Broadridge Financial Solutions, Inc., to SIFMA Operations Conference, April 29-May 2, 2007 (available at http://archives2.sifma.org/ops2007/pdf/ProxyPanelRichardDaly.pdf ).

43

FROZEN CHARTERS Although a broker-by-broker system could be easily implemented, it could result in other distortions where only a small number of street-name holders submit instructions to a particular broker and those holders have different preferences from other shareholders. The votes of the shareholders that do vote would be ‘overweighted’ to the extent of the uninstructed shares. The larger the ratio of uninstructed shares to the shares being used to determine the proportion, the stronger this effect. The Proxy Working Group considered there to be a possibility of manipulation where a broker has a disproportionately large number of uninstructed shares. 111 However, the concerns expressed by the Proxy Working Group seem to have been assuaged where the pool of votes used to set the proportion included only retail investors.112 Henry Hu and Bernard Black also consider this approach, and conclude that there is no reason to believe that the distortion would be problematic.113 Based on the above discussion, it is clear that the logistical issues involved in a proportional voting system are surmountable, whether through regulatory action, or through private ordering by corporations or brokers. This would speak in favor of reallowing uninstructed broker voting, though encouraging a move towards proportional voting.

D. Permitting Broker Voting on Certain Charter Amendments A third set of alternatives would be to replace IM 12-4 with a rule that permits broker voting on a circumscribed set of corporate governance proposals. The set of corporate governance proposals where broker voting would be allowed would be those with the greatest likelihood of ‘distorted fail’ outcomes, and the least likelihood of ‘distorted pass’ outcomes. Such a rule could be implemented by an addition to the supplementary materials to Rule 452, indicating that a broker could not vote an uninstructed proxy on a corporate governance matter such as the kinds listed in IM 12-4, unless the proposal met certain conditions. I consider three possible alternatives for what kind of proposals would be permitted.

1. Broker voting only for removing supermajorities One set of corporate governance proposals where broker voting could be permitted is charter or bylaw amendments to remove supermajority provisions. As the distributions set out in Appendix E indicate, most frozen charters occur in companies with high supermajority requirements. This is unsurprising – in companies without 111

See Proxy Working Group Report, supra note 25, at 17. See Proxy Working Group Addendum, supra note 108, at 4 (“By limiting the vote to be considered in making proportional voting decisions to the retail vote, the Proxy Working Group thought that the potential for manipulation could be significantly reduced.”) 113 See the discussion of this point in Henry T. C. Hu and Bernard Black, Equity And Debt Decoupling And Empty Voting II: Importance And Extensions, 156 U. Penn. L. Rev. 625 (2008) at 705-6 (“This would somewhat overweight the instructions that shareholders convey, but creates no obvious incentive problems. At the margin, the prospect of overweighted voting might induce more economic owners to vote.”) 112

44

FROZEN CHARTERS supermajorities, if a proposal is overwhelmingly supported, then the possibility of 1015% of shares being uninstructed broker shares and not vote is unlikely to reduce the overall vote below 50%. However, most supermajority requirements cannot be removed, because they are themselves subject to supermajority requirements for amendment. Therefore, charter amendments to remove them are also likely to be subject to the same ‘distorted fail’ problem if broker voting is not permitted. Allowing broker votes on amendments removing supermajorities would prevent ‘distorted fail’ outcomes on these proposals, and allow the circularity problem to be broken. There are currently high levels of shareholder support for removing supermajority provisions - charter amendments to remove supermajority provisions received average of 72% of votes cast in 2013.114 As a result, it is unlikely that such a proposal would receive less than 50% support and therefore have the possibility of a ‘distorted pass’ result. Allowing broker voting on amendments to remove supermajority provisions envisages a two-step process for amendment of other charter provisions. First, the charter would be amended to remove the supermajority provision. After that amendment had been implemented, the substantive provision could be amended at a subsequent meeting. Not only would this take several years, such a private ordering solution would require each affected company to go through this process. As with other private ordering solutions, this would be more duplicative than a regulatory solution.

2. Broker voting only for proposals that shareholders generally support Permitting brokers to vote only on matters that generally receive substantial shareholder support is likely to minimize the risk of potential distortion. If broker voting is allowed on such proposals, then the possibility of a ‘distorted fail’ outcome is eliminated. And if the proposal receives majority support, then a ‘distorted pass’ outcome is also not possible. Given that frozen charters occur where proposals fail despite overwhelming shareholder support, allowing broke voting on proposals supported by shareholders would target the kinds of proposals where frozen charters and other ‘distorted fail’ outcomes are most likely to be an issue. The difficulty with this solution would lie in choosing a bright line rule for the set of corporate governance proposals on which broker voting would be permitted. One alternative would be to set out a list of topics that generally receive strong shareholder support in the supplementary materials to Rule 452, and allow broker discretionary voting on those proposals. However, this is unlikely to be workable. It would not reflect variances in voting outcomes across firms and across time, and would likely need updating on a regular basis through the lengthy SEC rulemaking procedure. A better alternative would be to permit broker voting on a proposal that previously received a strong majority (e.g., greater than 80% of votes cast) at a previous annual meeting at the

114

See Georgeson, 2013 Annual Corporate Governance Review, supra note 56, at 22.

45

FROZEN CHARTERS company. Similar to the approach in Subsection 1, above, alternative above, this would require a two-step process to amend the charter – a vote to demonstrate sufficient support at the first meeting, and then an actual amendment vote at the second meeting, likely the following year. In contrast to the approach in Subsection 1, above, the first proposal could be submitted by a shareholder, whereas a supermajority amendment proposal would have to be brought by the directors. This outcome would also reflect the general practice of many corporations, which may wait for a successful shareholder proposal to demonstrate the preferences of the company’s shareholders before putting forward a management charter amendment proposal on a particular topic.115 As a result, this approach may not require a significant divergence from current practice.

3. Broker voting only where a supermajority is required A third method set of corporate governance resolutions on which broker voting could be allowed may be those amendments for which a supermajority vote is required. This would be a broader set of amendments than that described in Subsection 1, above.116 As discussed in Subsection 1, above, proposals that require a supermajority for amendment are the very proposals that are likely to result in frozen charters and other ‘distorted fail’ proposals. They are also proposals where there is almost no likelihood of a ‘distorted pass’ outcome, since significant support for the proposal would be required for it to pass. Permitting brokers to vote only where a supermajority of outstanding shares is required would therefore reduce frozen charters and other ‘distorted fail’ outcomes, while minimizing potential ‘distorted pass’ outcomes. An addition to the supplementary materials permitting broker voting on such proposals could also be drafted in a straightforward manner, without ambiguity.

D. The Procedure of Reform As discussed in Section E of Part V above, because IM 12-4 was not a formal change to Rule 452, it avoided the considered process required for reviewing and approving such rule changes. Similarly, it could easily be undone by a similar information memorandum amending the NYSE policy on the matter. However, there may be reasons to believe that the NYSE may not take action on the matter unless forced to by the SEC. As discussed in Section D of Part V, there is some evidence that directors and managers may prefer that certain charter amendments they have approved for submission to a vote of shareholders do not actually pass. If this is the case, these directors and managers may be less concerned about the shortcomings of IM 12-4, and may wish for it to remain in effect. Since the NYSE is funded by fees from corporations and members, rather than from investors, it may have an incentive to take actions preferred by those

115

See, e.g., Mira Ganor, Why Do Managers Dismantle Staggered Boards?, 33 Del. J. Corp. L. 1 (2008). 116 This would also encompass all amendments to supermajority provisions that were themselves subject to supermajority approval requirements, thereby encapsulating the set of proposals discussed in Subsection 1, above.

46

FROZEN CHARTERS responsible for the corporation – directors and managers. As a result, it may be unrealistic to expect the NYSE to act on its own initiative to solve the problems created by IM 12-4. As a result, the most realistic path to reform may lie through the SEC. In the event that the NYSE did not take action to solve the problems created by IM 12-4, the SEC could take action. This could take two forms. The SEC may be able to take action to strike down IM 12-4, on the basis that it was a rule making, and not merely a policy change, and therefore should have been undertaken through the rule making process. Secondly, the SEC has the power to unilaterally amend the rules of the NYSE.117 A rulemaking process under the aegis of the SEC would also be preferable to a policy change through IM 12-4, as it would follow the same well-developed process discussed in Section E of Part V. However, it is likely that a considered process of this nature may take some time. Another alternative, discussed in a number of the solutions above, is action by private ordering, either by corporations, or by brokers. However, given the perspective of directors and managers on IM 12-4 described in Part V.D above, there are reasons to doubt that directors or managers would undertake action to reverse its investor-harming effect of their own volition. Instead, investors may need to engage with directors and managers to encourage them to take such action. Given the number of companies that would have to undertake individual action, this process is likely to be considerably slower than a regulatory solution. Since there are a smaller number of brokers, a broker-based private ordering solution may be more efficient. Given the time that an SEC or private ordering process is likely to take, it would be optimal for the NYSE – if necessary, at the request of the SEC – to first take action to reverse IM 12-4, so as to avoid the investor-harming effects of IM 12-4 on companies bringing charter amendments in the interim.

VII. Conclusion Broker voting rules create the possibility of two kinds of distortion. If brokers are permitted to vote, and follow management recommendations, then broker voting will positively distort vote tallies, and may result in a ‘distorted pass’ result for a proposal. This is the concern that IM 12-4 was implemented to correct. However, eliminating broker voting, as IM 12-4 did, has another distorting effect, reducing vote tallies from the value they would have if the preferences of all shareholders were considered. Where this results in a proposal failing where shareholders would prefer it passed, there will be a ‘distorted fail.’ These distortions have been the unintended consequence of IM 12-4. As a result of the rule, parts of the charters of a substantial number of corporations are frozen. The shareholders of a number of corporations are unable to amend their bylaws, and

117

See Section 19(c) of the Exchange Act.

47

FROZEN CHARTERS other corporations now permit insiders a de facto veto over charter amendments. Given current levels of support for management proposals, these effects significantly outweigh the possibility of ‘distorted pass’ outcomes that IM 12-4 was designed to address. As a result, although IM 12-4 has an implicit investor protection rationale, its effect has been the opposite. Many charter amendments that investors consider to be in their interests and value enhancing can no longer be implemented. The implicit change in the amendment term of the charter also undermines the corporate contract. I propose a number of potential solutions to these problems. At the very least, and as an interim measure, IM 12-4 should be reversed; if the NYSE is unwilling to do this, the SEC has to the power to do so and should act. In the longer term, a solution that reduces both kinds of potential distortion could be implemented. The most promising potential solutions appear to be proportional voting, or a prohibition on broker voting on corporate governance matters, except for certain circumscribed exceptions that would prevent ‘distorted fail’ outcomes while avoiding the possibility of ‘distorted pass’ outcomes. In this way, the investor protection rationale of broker voting reform could be upheld.

48

FROZEN CHARTERS

Appendix A: Certain ‘Distorted Fail’ Outcomes, 2012-2013 Panel 1: 2012 Failed Charter Amendments Receiving >90% of Votes Cast

Company

% For (αfor)

Broker Votes (αbv)

% Req’d (αreq)

% For + Broker Votes (αfor + αbv)

Alcoa Inc.

47.3%

25.6%

80.0%

72.8%

Fail

Avista Corporation

74.6%

11.8%

80.0%

86.4%

Pass

Boston Scientific Corporation

78.9%

6.6%

80.0%

85.5%

Pass

Chesapeake Energy Corporation

62.0%

19.0%

66.7%

81.0%

Pass

CIGNA Corporation

72.0%

9.3%

80.0%

81.3%

Pass

Duke Energy Corporation

52.7%

28.8%

80.0%

81.6%

Pass

Franklin Street Properties Corp.

70.6%

16.6%

80.0%

87.1%

Pass

Hercules Offshore, Inc.

67.5%

21.9%

75.0%

89.4%

Pass

Medtronic, Inc.

72.1%

10.4%

75.0%

82.5%

Pass

Piedmont Natural Gas Company, Inc.

54.5%

28.5%

80.0%

83.0%

Pass

PPG Industries, Inc.

67.1%

12.6%

80.0%

79.8%

Fail

Principal Financial Group, Inc.

58.1%

5.9%

75.0%

64.0%

Fail

Solta Medical, Inc.

61.1%

22.9%

66.7%

84.0%

Pass

St. Jude Medical, Inc.

77.8%

9.1%

80.0%

87.0%

Pass

SUPERVALU Inc.

50.5%

24.1%

75.0%

74.6%

Fail

Teradata Corporation

75.9%

7.9%

80.0%

83.8%

Pass

49

Result if BVs permitted

FROZEN CHARTERS Panel 2: 2012 Failed Charter Amendments Receiving >90% of Votes Cast

Company

% For (αfor)

Broker Votes (αbv)

% Req’d (αreq)

% For + Broker Votes (αfor + αbv)

Akamai Technologies, Inc.

73.2%

10.9%

75%

84.1%

Pass

Apache Corporation

73.9%

9.5%

80%

83.4%

Pass

Avista Corporation

73.8%

12.9%

80%

86.7%

Pass

Capital One Financial Corporation

80.0%

5.5%

80%

85.4%

Pass

Chesapeake Energy Corporation

60.1%

22.4%

67%

82.6%

Pass

Cliffs Natural Resources Inc.

46.7%

17.4%

50%

64.1%

Pass

Connecticut Water Service, Inc.

57.9%

25.3%

80%

83.2%

Pass

Emerson Electric Co.

71.2%

14.3%

85%

85.4%

Pass

Energen Corporation

78.3%

8.9%

80%

87.3%

Pass

FirstEnergy Corp.

73.3%

11.7%

80%

85.1%

Pass

L-3 Communications Holdings, Inc.

79.3%

8.8%

100%

88.1%

Fail

Marathon Petroleum Corporation

74.2%

9.2%

80%

83.4%

Pass

Masco Corporation

76.5%

5.0%

80%

81.5%

Pass

Mattersight Corporation

69.2%

19.7%

80%

88.9%

Pass

Medtronic, Inc.

74.7%

9.3%

75%

83.9%

Pass

ModusLink Global Solutions, Inc.

51.5%

0.0%

75%

51.5%

Fail

NYSE Euronext

63.8%

16.5%

80%

80.3%

Pass

OGE Energy Corp.

65.3%

16.4%

80%

81.7%

Pass

PPG Industries, Inc.

68.4%

12.3%

80%

80.7%

Pass

Principal Financial Group, Inc.

61.4%

4.0%

75%

65.4%

Fail

QEP Resources, Inc.

77.5%

9.1%

80%

86.7%

Pass

Reinsurance Group of America, Inc.

81.5%

4.4%

85%

85.9%

Pass

Southside Bancshares, Inc.

56.2%

26.2%

67%

82.4%

Pass

SPX Corporation

78.0%

6.1%

80%

84.1%

Pass

St. Jude Medical, Inc.

77.2%

8.9%

80%

86.1%

Pass

The Goodyear Tire & Rubber Co.

70.3%

14.8%

67%

85.1%

Pass

50

Result if BVs permitted

FROZEN CHARTERS Panel 3: Failed Bylaw Amendments Receiving >90% of Votes Cast, 2012-2013

Company

% For (αfor)

Broker Votes (αbv)

% Req’d (αreq)118

% For + Broker Votes (αfor + αbv)

Boston Scientific Corporation

78.9%

6.6%

80%

85.5%

Pass

Chesapeake Energy Corporation

62.0%

19.0%

67%

81.0%

Pass

CIGNA Corporation

72.0%

9.3%

80%

81.3%

Pass

CLECO Corporation

76.3%

9.2%

80%

85.6%

Pass

SUPERVALU Inc.

50.5%

24.1%

75%

74.6%

Fail

The Goodyear Tire & Rubber Company

70.4%

14.8%

67%

85.2%

Pass

118

Result if BVs permitted

Note that all of the bylaw amendments had ‘shares outstanding’ supermajority requirements. This is unsurprising – votes cast requirement are normally a supermajority, and would not have resulted in failures if supported by greater than 90% of the votes cast.

51

FROZEN CHARTERS

Appendix B: IM 12-4 and the Likelihood of Charter Amendment Failure To determine the effect of IM 12-4 on the likelihood of failure of a particular proposal, I model the relationship between whether the vote passes or fails, as the dependent variable, and whether the vote took place in 2012 or 2013 (a binary variable taking the value 1 for 2012 or 2013 and 0 for previous years). Given the importance of supermajority voting requirements, I control for the voting requirement necessary to approve the vote (αreq). I use a logistic regression model, and cluster standard errors by company. My results are set out in Table B1 below, using two separate specifications – first: f(Fail) = a + b1.Time + b2. αreq + b3. Requirement Type + e

(1)

In the second specification, I add an interaction term between the vote requirement and the time: f(Fail) = a + b1.Time + b2. αreq + b3.Time. αreq + b4. Requirement Type + e (2) Table B1: Effect of IM 12-4 on Likelihood of Proposal Failing (1) Fail

(2) Fail

2.627*** (0.470) 503.7*** (449.4)

1.256 (0.324) 0.00134*** (0.000716)

0.0447*** (0.0532) 72.59*** (74.04) 446.7*** (783.3) 1.193 (0.308) 0.00503*** (0.00296)

Observations Chi2

2,430 82.46

2,430 109.1

Difference in Odds Ratios for Time (Time = 1 – Time = 0)

0.0373

0.0820

Variables Time (2012-3) Vote Required Time x Vote Required Requirement Type Constant

As Table B1 shows, whether or not the vote took place after IM 12-4 is significant at the 1% level in determining the probability of the vote failing. In the second model, which shows a better degree of fit, the odds of a proposal failing increase 8.2% if the proposal took place after IM 12-4.

52

FROZEN CHARTERS

Appendix C: Management Proposals Receiving <50% Shareholder Support 2012-2013 Panel 1: Majority of Votes Cast Standard

Type of Proposal

% Votes Cast For (βfor)

Broker Votes (αf)

Votes For and Broker Votes as % of Votes Cast (βfor+bv)

Pacific Sunwear of California, Inc.

Other

48.4%

11.9%

55.1%

Pass

Viad Corp

Other

43.4%

4.2%

46.0%

Fail

Charter Amendment

40.4%

0.0%

40.5%

Fail

Other

36.6%

4.8%

40.0%

Fail

Fred's, Inc.

Other

14.3%

6.3%

19.9%

Fail

Benchmark Electronics, Inc.

Other

46.8%

4.2%

49.2%

Fail

Company

Green Plains Renewable Energy, Inc. Obagi Medical Products, Inc.

Result if BVs voted in favor

Panel 2: Majority of Votes Outstanding Standard Company

Cameron International Corporation

Type of Proposal

% Outstanding Cast For (αfor)

Broker Votes (αbv)

For and Broker Votes (αfor+αbv)

Result if BVs voted in favor

Charter Amendment

40.1%

2.9%

43.0%

Fail

53

FROZEN CHARTERS

Appendix D: Charter Amendments and Director Election Turnout I model the relationship on shareholder turnout of having a corporate governance proposal in a particular year. I use director election data for U.S. companies in the Russell 3000 index for the period 2010 to 2012. I used a firm fixed effects model to compare the turnout for companies that had charter amendments between 2010 and 2012 with the turnout the same companies in the years that they did not have charter amendments during that period. I use whether the company had a corporate governance proposal at the same meeting or not, as well as the voting requirement (αreq) for the corporate governance proposal, if any, as independent variables, i.e.: f(Turnout) = a + b1.CG Proposal + b2. αreq + e

(1)

I collect data from 3,213 firms over the three years, with 8,127 observations in total. I used a firm and year fixed effects model, with standard errors clustered by firm. My results are set out in Table D1 below. Table D1: Effect of Presence of Corporate Governance Proposal on Shareholder Turnout, 2010-2012 Variables

Turnout 0.0132

CG Proposal

(0.0148) -0.0147

Voting Requirement

(0.0243) 0.760***

Constant

(0.000164)

Observations

8,127

Number of Companies

3,213

R-squared

0.000

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

The results show no significant effect of a corporate governance proposal on shareholder turnout over the period.

54

FROZEN CHARTERS

Appendix E: Estimates of Companies with Potential ‘Distorted Fail’ and ‘Distorted Pass’ Results Panel 1: Frozen Charter Companies Caused by IM 12-4, by Market Capitalization and Charter Amendment Requirements Frozen Charters

% of n

Sample Comps by Vote Req

50%

25

2.5%

1,007

22

50% - 59%

0

0.0%

1

15.0%

1,074

60% - 69%

70

8.9%

790

81

7.9%

1,026

70% - 79%

53

23.5%

226

10,000 - 100,000

31

11.4%

273

80% - 89%

131

33.2%

394

>100,000

1

3.4%

29

90% 100%

4

26.7%

15

283

11.6%

2,433

Total

283

11.6%

2,433

Frozen Charters

% of n

Sample Comps by Size

0

1

11.1%

9

1 – 100

8

36.4%

100 - 1000

161

1,000 - 10,000

Market Cap. ($m)

Total

Vote Required

Panel 2: Companies with Insider Vetoes Caused by IM 12-4, by Market Capitalization and Charter Amendment Requirements Market Cap. ($m)

Insider Vetoes

% of n

n

Vote Required

0

0

0.0%

9

50%

1 – 100

2

9.1%

22

144

13.4%

1,074

70

6.8%

1,026

11

4.0%

273

0

0.0%

29

227

8.9%

2,433

100 - 1000 1,000 10,000 10,000 100,000 >100,000 Total

50% 59% 60% 69% 70% 79% 80% 89% 90% 100% Total

55

Insider Vetoes

% of n

n

57

5.7%

1,007

0

0.0%

1

84

10.6%

790

31

13.7%

226

55

14.0%

394

0

0.0%

15

227

8.9%

2,433

FROZEN CHARTERS Panel 3: Board-Only Bylaw Companies by Market Capitalization and Charter Amendment Requirements Market Cap. ($m)

Companies

% of n

n

Vote Required

0

1

11.1%

9

1 – 100

7

31.8%

100 - 1000

129

1,000 - 10,000

Companies

% of n

n

50%

23

1.6%

1,402

22

50% - 59%

0

0.0%

1

12.0%

1,074

60% - 69%

56

9.7%

579

56

5.5%

1,026

70% - 79%

42

23.5%

179

10,000 - 100,000

16

5.9%

273

80% - 89%

87

32.3%

269

>100,000

0

0.0%

29

90% - 100%

1

33.3%

3

209

8.6%

2,433

209

8.6

2,433

Total

Total

Panel 4: Estimates of Companies with Potential ‘Distorted Pass’ Results Type of Vote Requirement Outstanding Shares Votes Cast

Total

Potentially Distorted

% of n

n

354

15.3%

2,311

109

89.3%

122

463

19.0%

2,433

56

Vote Requirement

Potentially Distorted

% of n

n

460

45.6%

1,007

50% - 59%

0

0%

1

60% - 69%

3

0.3%

790

70% - 100%

0

0%

635

463

19.0%

2,433

50%

Total