Proxy voting behavior of institutional investors: Evidence from TIAA-CREF

Abstract

This thesis analyzes the voting behavior of the Teachers Insurance and Annuity As­sociation - College Retirement Equities Fund (TIAA-CREF). As one of the largest financial services companies in the United States, with over $426 billion in combined assets under management as of 31 of March 2010, the fund is using proxy voting as a tool to promote positive returns from their investments. This thesis relies on a database constructed out of SEC N-PX filings over a period of six month. The results indicate that TIAA-CREF only withholds directors their vote in a moderate amount of cases. In addition, the fund voted more often against management at proposals cast by shareholders concerning board structures and shareholder rights than at proposals concerning other corporate governance issues.

1 Introduction

Rupert Murdoch and Steve Jobs know how proxy voting works. Not only in their function as director at Fox Entertainment or Pixar, but also during the annual meetings in 2004. Instead of voting together with the management, shareholders withheld the two well-known candidates their vote. Shareholder activism is evolv- ing to become a state of the art tool of majority shareholders, especially for mutual funds. Nowadays, shareholders are making increased use of their say to tackle agency conflicts more directly, turning shareholder activism into an effective corporate gov­ernance mechanism. The Teachers Insurance and Annuity Association of America (TIAA-CREF) is one of them making use of its say. In order to manage their un- derlying assets more effectively they follow an active corporate governance strategy. Rather selling company's stock following the "Wall street rule", they are encourag- ing companies to change their policies.

While shareholder activism by institutional investors has gained increased promi- nence over the last decade, there has been limited empirical work investigating on the voting behavior of these investors. In the following analysis proxy votes from SEC N-PX filings for July 1, 2004 to December 31, 2004 by TIAA-CREF have been used in order to analyse the voting behavior. I conducted a descriptive analysis using a sample covering 574 firms and 3464 proxy votes in order to answer three questions: firstly, does TIAA-CREF use proxy votings in order to emphasize their discontent- edness with management's corporate policy? Secondly, do they withhold directors their vote in case no agreement concerning the proxy filings could be reached before the shareholders meeting? And thirdly, is management nowadays more willing to remove anti-takeover defences?

My findings suggests that TIAA-CREF is using proxy votes to put pressure on management. The results also suggest that proxy voting analyses understate the magnitude of institutions' attempts to influence management because they exclude a substantial number of cases in which management adopts the suggested changes without public knowledge. This also explains why the fund only withholds directors their vote in a moderate amount of cases. Nonetheless, the fund is also disgracing target company's management in public in case disagreement still exists after pri­vate negotiations. Finally, TIAA-CREF voted more often against management at proposals cast by shareholders concerning board structures and shareholder rights received than at proposals concerning other corporate governance issues.

The remainder of this thesis is organized as follows: The second chapter provides a brief overview of the existing literature on institutional activism. The third chapter explains the composition of the sample while the fourth chapter presents the classi- fication scheme and the descriptive analysis. Results, limitations and opportunities for further research are presented in the fifth chapter.

2 Related Literature

A considerable body of research has focused on the role and effectiveness of share- holder activism. The central idea of shareholder activism holds that shareholder proposals put pressure on managers to adopt company value increasing policies.

Indeed, Karpoff, Malatesta, and Walking (1996) as well as John and Klein (1995) found empirical results that poor performing firms are attracting shareholder pro- posals. Firm size, presence of negative net income and the percentage of outside directors with outside directorship in other S&P 500 firms are important determi- nants for casting a proxy vote by shareholders. Gillian and Starks (2000) observed an increase in the number of votes cast in favor of shareholder proposals. Whereas proposals sponsored by institutions or through coordinated activism receive signif- icantly more favorable votes than those sponsored by independent individuals or religious organizations (Gillian & Starks, 2000). Thus, shareholders have nowadays an opportunity to put pressure on management.

The question is whether this opportunity is used by institutional investors and in which way. On the one hand Parrino, Sias, and Starks (2003) states that some institutional investors voting with their feet by selling their shares when dissatisfied with corporate performance. On the other hand Cornett, Marcus, Saunders, and Tehranian (1995) confirms a positive relationship between measures of institutional investor involvement and a firms operating cash flow returns. Randall and Cotter (2007) found that boards are increasingly willing to remove important anti-takeover defenses. Due to the finding that many more proposals are receiving majority share- holder support (Randall & Cotter, 2007), pressure on management has increased over the last several years. Therefore, large institutional investors seem to encour- age companies to change their policies instead of voting with their feet.

This suggests that more and more investors are engaging in shareholder activism. Recent empirical research focuses on the motivation behind the increased share­holder activism by institutional investors and analyses voting behavior. As Cornett et al. (1995) concludes the type of investor, whether pressure-insensitive or pressure- sensitive, matters. Agrawal (2008), for example, found that the observed AFL-CIO pension funds are more likely to vote against directors of firms in which there is greater frequency of plant-level conflict between labor unions and management dur- ing collective bargaining and union member recruiting. Furthermore Ashraf, Jayara- man, and Ryan (2009) mention that mutual funds with a pension-related business tie to the firm are more likely to vote against shareholder proposals than mutual funds that do not have a pension-related business tie. They conclude that pension- related business ties influence how mutual funds vote on shareholder proposals for CEO compensation and that these ties can influence the outcome of a vote (Ashraf et al., 2009). Therefore, institutional investors have their own agendas whereas in- creasing shareholder value is only one of many aspects.

TIAA-CREF is well-known for being effective at convincing target companies to adopt changes in governance practices according to the fund's agenda. As Carleton. Nelson, and Weisbach (1998) explain TIAA-CREF activism involves several steps: Target selection, submission of a proxy statement, dialog with the management, negotiating for an agreement, and finally monitoring to ensure compliance with the agreement. The procedure underlines that large institutional investors avoid the open confrontation. The voting on proxy votes on annual meetings is hence the "ul­tima ratio" of putting pressure on management. I hypothesize that TIAA-CREF uses proxy votings in order to emphasize their discontentedness with management's corporate policy. Similarly, I hypothesize that TIAA-CREF withholds directors their vote in case no agreement concerning the proxy filings could be reached be- fore the shareholders' meeting. Finally, I hypothesize that management is nowadays more willing to remove anti-takeover defences.

3 Sample selection

In the following analysis proxy votes cast by TIAA-CREF have been used. As indicated in Agrawal's working paper concerning corporate governance objectives of labor union shareholders I collected the proxy votes from SEC N-PX filings for July 1, 2004 to December 31, 2004 (Agrawal, 2008). I have chosen the Institutional Equity Index Fund (TCIEX), because it holds a broad selection of securities. In addi­tion the investment objective of the fund seeks a long-term total return by investing primarily in a portfolio of foreign equity investments based on the MSCI EAFE index which measures the performance of the leading stocks in 21 developed coun- tries outside North America (Teachers Insurance and Annuity Association College Retirement Equities Fund, 2010). In addition shareholder activism and "Withhold voting" campaigns gained in recent years more significance (Katz, 2005). Directors are implementing more of the actions called by shareholders, due to the fact that since 2002 more proposals are receiving majority shareholder support (Randall & Cotter, 2007). Given that the Securities and Exchange Commission (SEC) adopted Rule 14a-8 in 1942 shareholders are permitted to place proposals in annual corporate proxy statements. Together with the increased regulation due to the Sarbanes-Oxley Act in 2002 and the obligation for mutual funds to report their voting behavior in 2003, shareholder activism has become a powerful tool in the Post-Enron period. Thus, I have chosen the year 2004 in order to measure if this trend could also be measured on a lasting time period. Furthermore, the second half of 2004 seems to be more interesting to analyse as I assume that more special meetings occurred, hence I reduced the sample to the second half of 2004. In this way I conducted descriptive analyses with a sample consisting out of 574 companies and 3463 proxy votes.

4 Empirical analysis

4.1 Classification

In order to classify the proposals I used the classification scheme of Karpoff et al. (1996) as an orientation. I divided the total sample into three subgroups. The first main group covers corporate governance related issues. I further divided this group into the subgroups of internal and external corporate governance. The subgroup in­ternal corporate governance takes all proposals into account which are dealing with auditors, changes in board structure, compensation, and directors. As the compen- sation subgroup is of strong interest for further analysis, I divided the subgroup into stock / equity, option, and other compensation proposals. The subgroup external corporate governance covers the proposals concerning merger decisions and restric- tions on shareholder rights. Moreover, the second main group contains all corporate social responsibility proposals. Lastly, the third main group covers all remaining proposals, subdivided for business proposals and the remainder of the sample. In this way I expect to become a clearer view on TIAA-CREF's voting behavior.

4.2 Results

Based on this classification scheme and based on the retrieved data, the sample con­tains 3245 corporate governance, 10 corporate social responsibility, and 208 other proposals, for a total of 3464 proposals over a half-year period from 2004. According to the classification scheme presented in table 1 on appendix page III with 2522 director, 344 auditor, and 208 other proposals these three categories build the vast majority of proposals. Thus, most of the proposals are submitted by management due to the pure obligation. Concerning the classification according to the caster of the proposal, management has cast 3430 or 99% of all proposals in the sample period. Only 33 or 1% of the proposals account to shareholders. Even though the number of shareholder proposals is quite small, table 2 on appendix page IV gives an interesting overview on which proposals were submitted by shareholders. Most prominent are proposals regarding corporate social responsibility with 10 proposals, or 30.3% of all shareholder proposals followed by 7 proposals relating to changes in board structures, covering 21.21%. The remaining proposals cast by shareholders deal with compensation (5, 15.15%), restrictions on shareholder rights (2, 6.06%), and other issues (9, 27.27%), respectively. These results differ from previous studies. Randall and Cotter (2007) report 27.3% compensation, 22.6% external corporate governance, 18.6% internal corporate governance, and 20.5% other social responsi­bility proposals in their main categories. Karpoff et al. (1996) reports with 8.0% for compensation, and 14.4% for corporate governance proposals numbers closer to the TIAA-CREF sample.

Moreover, I examined in a first step the voting behavior of TIAA-CREF regarding management and shareholder proposals. According to table 4 on appendix page VI the fund voted for 85.94% of all proposals. They withheld 8.61% of their votes, voted in 5.14% of all cases against a proposal, abstained 0.09% of all proposals, and voted not on 0.23% of all proposals. These numbers are mainly the same for proposals cast by management, but differ significantly for shareholder proposals. TIAA-CREF voted only for 42.42% of all shareholder proposals and voted against 48.48% of all shareholder proposals. Interestingly, the fund voted in 51.52% of all shareholder proposals with the management and 48.48% of all shareholder pro- posals against management. Overall, 477 times (13.77%) the fund voted against management. This supports the hypothesis that TIAA-CREF is an active investor regarding shareholder activism making use of the previous described activism pro- cedure by Carleton et al. (1998). Thus, the results show that the noiseless strategy in finding an agreement previous to the annual meeting pays out and dissent with management is only shown in a few cases in public.

In the following paragraph a more precise analysis examines the shareholder propos- als. In 5 (71.41%) cases TIAA-CREF voted for a shareholder proposal concerning the board structure and voted against management (Table 8 on appendix page IX). In contrast in 4 (80%) cases the fund voted against a compensation proposal cast by shareholders and voted with management (Table 9 on appendix page IX). In addition, TIAA-CREF withheld only 298 (11.82%) directors the vote (Table 13 on appendix page XI). All shareholder rights proposals cast by shareholders have been approved by the fund, while management recommended to vote against (Table 16 on appendix page XIII). Corporate social responsibility proposals are all the more interesting as all of them have been cast by shareholders. According to table 17 on appendix page XIII, TIAA-CREF voted only for 2 (20%) proposals, abstained 2 (20%) proposals, but has voted against the remaining 6 (60%) proposals. Over­all, concerning corporate social responsibility proposals, the fund voted in 7 cases with management. In the three cases TIAA-CREF voted against management, the proposal titles were "Stockholder proposal for the adoption of the China business principles for rights of workers in China.", "Shareholder proposal regarding a sus- tainability report", and "Adoption of a smoke-free policy for all company owned restaurants." These results indicate that the support of a shareholder proposals regarding corporate social responsibility is dependent on the impact on the com­pany and its underlying ethical and social guidelines. Furthermore, on a percentage basis more shareholder proposals regarding board structure and shareholder rights were approved by TIAA-CREF then other corporate governance proposals. This indicates that the fund is mainly interested in removing anti-takeover defences. A possible explanation is that TIAA-CREF could find agreements generally on corpo- rate governance issues noiselessly in advance, while management is not open-minded concerning anti-takeover defences such as staggered boards or deferred voting rights. Thus, this findings do not match with Randall and Cotter (2007) findings.

Of further interest are the results from the remaining, non-corporate governance related proposals. As table 18 on appendix page XIV indicates, TIAA-CREF voted for 126 (60.58%), against 81 (38.94%), and abstained one (0.48%) proposal. What catches one's eye is the fact that the fund voted in 82 (39.42%) cases against manage­ment, which is in relative terms in comparison to the corporate governance related numbers quite high. Table 20 on appendix page XV shows that most of these pro- posals have been non-classifiable proposals concerning corporate governance related issues. Due to the fact that these remaining proposals are not further classified, further research might also take a closer look on business-related proposals.

In the second part of the descriptive analysis I focused on the number of proposals submitted per month. In total, 77 meetings were held in July of 2004, 68 in August. 82 in September, 96 in October, 98 in November, and 61 in December. Overall. 83.97% of the meetings have been annual meetings, only 16.03% have been special or consent meetings (Table 21 on appendix page XVI). In the next step, I divided the proposals up according to the classification scheme used before. Table 23 on appendix page XVII shows that the proposals are all distributed equally over time. Thus, no certain proposal type has been submitted in one month significantly more often than other proposal types. Furthermore, I analysed how often management and shareholders cast proposals over the months. While management always sub- mits proposals due to their duty over the whole year, shareholders seem to be more active in July (Table 22 on appendix page XVI). My findings that July is a month where shareholders are more active is supported by the findings in the tables 24, 25, and 26 on appendix page XVIII. In July TIAA-CREF withheld directors 53 times their vote and voted in 103 cases against management, even though not a high num­ber of special meetings and number of proposals on special meetings occurred. Due to the small sample size concerning proposals cast by shareholders, further research could analyse whether there is time period shareholders are more active in submit- ting proposals.

Continuing with the third part, I analysed the average number of proposals per company. As shown in table 27 on appendix page XIX per company on average 5.03 internal corporate governance, 0.14 external corporate governance, 0.02 corporate social responsibility, and 0.36 other proposals were submitted, whereas table 28 on appendix page XX shows that only 0.06 shareholder proposals were submitted per company. Tables 29 and 30 on appendix page XX give a brief overview of the aver­age number of proposals concerning the voting behvior of TIAA-CREF. On average TIAA-CREF voted in favor of 5.18 proposals, withheld 0.52 director votes, and voted against 0.31 proposals. With respect to the congruency with management TIAA-CREF voted on average per company for 5.19 proposals, and against 0.83.

[...]