Can Institutional Investors Improve Corporate Governance through Collective Action?

Transcription

1 Preliminary and incomplete. Comments welcome. Can Institutional Investors Improve Corporate Governance through Collective Action? Prepared for CCGG annual meeting, June 2014 Craig Doidge, University of Toronto Alexander Dyck, University of Toronto Hamed Mahmudi, University of Oklahoma Aazam Virani, University of Toronto

2 Introduction Better governance has private and social benefits: With better governance, insiders limit private benefits higher valuations and better access to external finance to fund new investment projects Academic research suggests better governance outcomes come from situations where investors have power relative to insiders (management and board). Two situations are of most focus (Shleifer/Vishny (1997)) Concentrated equity ownership Legal protections for investors Institutional investors not seen as central as lack power 2

3 Why are Institutional Investors (II) presumed to lack power?: collective action problems The income of the corporation is a collective good to the stockholders, and the stockholder who holds only a minute percentage of the total stock, like any member of a latent group, has no incentive to work in the group interest. The Free Rider Problem with collective action Collectively hold large stakes could be powerful Mancur Olson (1965) But each II holds a small stake let others do it for you, need to coordinate Additional legal risks and costs with coordination And, some investors face conflicts because of other business relationships with issuers Coordinated shareholder activism is rare. Instead, each institution acts as a lone wolf Black (1998) Almost no evidence that collective action organization of investors is important (studies of CII of US mixed results) 3

4 Can II Address Challenges of Collective Action and Generate Power to Change Governance? Use evidence from CCGG private engagements adoptions Control for fact that firms may have adopted changes without CCGG involvement, because of other activists Majority voting: plur.maj., slateind., disclose votes Say on pay: advisory vote on executive compensation Compensation structure and disclosure: capped pensions, clawbacks, performance peer groups Explore broader impact beyond engaged firms via Director interlocks Public policy initiatives Informal regulation : Develop best practices and measure firms governance against them 4

5 Contribution? Provide new evidence whether collective action organizations can be a vehicle to create investor power to change governance Quantify overall impact Provide new evidence on the importance of domestic institutional investors for governance change Important finding in literature that only what matters is level and change in foreign institutional ownership Provide new evidence on how II drive change Direct effects of engagements Indirect effects through board interlocks Interrelationships across multiple strategies 5

6 Outline 1. Context and Background 2. Data and Empirical Framework 3. Results: Engagements and Adoptions 4. Results: Spillovers 5. Discussion: Broader Influence 6. Key Choices? 7. Limits to Collective Action 8. Conclusions 6

7 CCGG: Voice of the Shareholder CCGG Members in 2005 Institution CCGG MEMBER 2005 % HELD TSX COMP TD Asset Management, Inc. YES 2.61 Fidelity Management & Research Co. NO 2.44 RBC Global Asset Management, Inc. YES 2.41 BlackRock Fund Advisors NO 2.26 BMO Asset Management, Inc. YES 1.78 Jarislowsky Fraser Ltd. YES 1.75 I. G. Investment Management Ltd. NO 1.50 Mackenzie Financial Corp. YES 1.09 Invesco Canada Ltd. NO 0.96 CIBC Global Asset Management, Inc. YES 0.76 The Caisse de depot et placement du Quebec NO 0.71 Ontario Teachers' Pension Plan Board YES 0.69 Wellington Management Co. LLP NO 0.60 Phillips, Hager & North Investment Management Ltd. YES 0.58 Pyramis Global Advisors (Canada) ULC NO 0.53 Connor, Clark & Lunn Investment Management Ltd. YES 0.47 Bissett Investment Management Ltd. NO 0.47 Canada Pension Plan Investment Board YES 0.46 Manulife Asset Management Ltd. YES 0.45 Franklin Templeton Investments Corp. YES 0.45 Mutual fund companies, asset managers, and pension funds Members have interest in L-T improvements in governance 12/20 top owners of Canadian equities Canadian-domiciled asset managers Long horizon investors with liabilities in CAD, e.g., pension plans Leadership by large pension plans 7

8 Snapshot of CCGG member influence on Issuers: Holdings in 2005 Median holding: 14% Own 20% or more of 43% of firms Bigger holdings in larger corps: median = 17% Stakes in controlled and widely held corporations Controlled corporations are important 37% by number 35% by value Non CCGG institutional owners 27% 8

9 What Strategies does the CCGG Employ to Change Governance? 1. Broadcast guidelines to the people and public companies of Canada through our website, the media, and many professional organizations 2. Engage directly with company chairs, CEOs, and secretaries to promote good governance policies and practices 3. Participate in the evolution of public policy frameworks that shape the investment industry We focus on 3 Campaigns that were the focus of CCGG engagements 1. Majority Voting: 2005/2006 (only) 2. Say on Pay: Compensation Structure and Disclosure: N.B. Don t look yet at as need 2 yrs to see if have impact 9

10 Test for Impact of Majority Voting Initiative 05/06 on Adoption 06/07 January 2005: all firms (but 1) had plurality voting, with slates of directors, no disclosure of votes CCGG developed a model majority voting policy in consultation with a leading law firm (2005) Fall 2005: CCGG contacted boards of the 5 major Canadian banks and 2 insurance companies Letter and personal All 7 firms adopted majority voting soon after Then sent letters to another 81 firms suggesting they adopt majority voting Citing adoption by the first 7 firms Later published majority voting guidelines to provide clarity and guidance over director voting practices US Studies show this can affect value: eg. Ertimur, Ferri and Oesch(2013), although not our focus 10

11 Then Test for Impact of Engagement Meetings With Firms Sought to increase investor power through direct meetings with independent directors 104 meetings over 2008 to 2011 Meetings raised concerns about Say on pay Comp. structure and disclosure (clawbacks, capped pensions, performance peer groups) N.B. CCGG controls agenda by directly contacting independent directors, sidestepping mgm t Yes No Yes% Investors and Indep. board % Long-term investor and indep. board at meeting CCGG represented by longterm investor % % Chair or lead director present % Two or more directors present at meeting % 11

12 Focus of CCGG Engagements (Table 2A) Year N Majority Voting N Engaged Say on Pay Clawbacks Capped Pension Performance Peer Groups Compensation Policies Total International Evidence shows that these issues matter: Say on Pay: eg. Ferri and Maber (2013) Clawbacks: eg. Iskandar-Datta and Jia (2013) Pension Payouts: eg. Bebchuk and Jackson (2005) Performance Peers: eg. Bettis, Bizjak, Coles and Young (2014) 12

13 Where we looked to see CCGG priorities? Engagement Reports Pre-meeting briefing book Post-meeting report 13

14 Outline 1. Context and Background 2. Data and Empirical Framework 3. Results: Engagements and Adoptions 4. Results: Spillovers 5. Discussion: Broader Influence 6. Key Choices? 7. Limits to Collective Action 8. Conclusions 14

15 Empirical Approach Regress whether adopted or not over next 2 yrs (1 yr) against variable that captures engagements, and a set of other variables Technically, a logit specification Designed to account for other potential drivers of adoption aside from CCGG specific engagements with firms Will express CCGG impact as the implied % change in probability of adoption from having been engaged by CCGG. Start with more parsimonious set of controls, then introduce more controls 15

16 What Could Drive Adoptions? Factors affecting CCGG influence? CCGG % stake Other institutional ownership Turnover Internal influences Firm performance: stock returns Controlled by insiders Governance Likelihood of needing external capital External governance influences Shareholder proposals (share.ca) Governance imported from the US: cross listings Media attention 16

17 CCGG Impact on Majority Voting Adoptions (Table 4B) Adoptions in t+1 or t+2 for year t engagements Engaged firms more likely Controlled corps less likely Good governance firms more likely Marginal effect? 30% increased probability of adoption (Col 1) (1) (2) (3) (4) (5) Engaged on Majority Voting 1.582*** 1.826*** 1.438*** 0.897** 0.842* (4.40) (4.43) (3.50) (2.07) (1.90) Controlled Corporation *** *** (-2.89) (-3.28) Good Governance 2.191*** 1.913*** (5.31) (4.46) Log(Assets) 0.714*** 0.444** (3.77) (2.42) Industry Fixed Effects YES YES YES YES YES Observations Pseudo R

18 CCGG Impact on Majority Voting Adoptions with more controls (Table 4B) Engagement has significant impact on adoptions % stake not significant In media eye, more likely Economic impact is still substantial Marginal effect = 16-22% (6) (7) Engaged on Majority Voting 1.300*** 0.992** (2.58) (2.13) Controlled Corporation ** (-2.48) Good Governance 2.044*** (4.64) CCGG % Ownership (1.14) (1.57) Institutional % Ownership ex. CCGG (0.21) (0.83) Turnover (0.12) (0.42) 1-year Stock Return (0.77) (0.18) Financing Deficit * (-1.85) (-1.40) Log(1+Firm Media Cites) 0.359* 0.354* (1.89) (1.79) Majority Voting Proposal (1.30) (1.15) Cross-Listed Firm (1.32) (0.99) Industry Fixed Effects YES YES Observations Pseudo R

19 Say on Pay Engagements (Table 6A) Engagements are driven by similar / different factors $ stake still important, but % stake not Controlled corps less likely Poorer performers more likely Similar for Comp. structure and disclosure (Table 7A) but Governance is positive and significant (1) (8) (9) (10) (11) CCGG $ Ownership 0.539*** 0.448*** 0.401** (3.54) (3.37) (2.53) CCGG % Ownership (0.08) (0.05) Institutional % Ownership ex. CCGG * (1.36) (1.78) Controlled Corporation * ** (-1.79) (-2.16) Good Governance (0.79) (0.65) 1-year Stock Return ** ** * * (-2.06) (-2.04) (-1.70) (-1.73) Cross-Listed Firm * (1.54) (1.88) (-0.02) (0.23) Log(Assets) 0.667*** 0.609*** (4.46) (3.79) Year Fixed Effects YES YES YES YES YES Industry Fixed Effects YES YES YES YES YES Observations Pseudo R

20 CCGG Impact on Say on Pay Adoptions (Table 6B) Adoptions in t+1 or t+2 for year t engagements Engaged firms more likely to adopt Controlled corps less likely Good governance firms more likely Marginal effects? 25% increased probability of adoption (col 1) (1) (2) (3) (4) (5) Engaged on Say on Pay 2.979*** 2.830*** 2.741*** 1.818*** 2.232*** (7.74) (7.31) (7.17) (4.49) (5.50) Controlled Corporation *** *** (-3.83) (-4.84) Good Governance 2.162*** 1.921*** (6.74) (5.78) Log(Assets) 0.835*** 0.474*** (6.68) (3.78) Year Fixed Effects YES YES YES YES YES Industry Fixed Effects YES YES YES YES YES Observations Pseudo R

21 CCGG Impact on Say on Pay Adoptions with more controls (Table 6B) Some controls matter: Controlled corps, Governance, Media, Proposals, Cross Listed Engagement matters, but lower magnitude Marginal effect = 10% (col 6) Engaged Firm is significant but SOP Engagement is still significant (6) (7) (8) (9) Engaged on Say on Pay 1.931*** 2.102*** 1.326*** 1.605*** (5.00) (4.81) (3.29) (3.54) Engaged Firm 1.032*** 0.828** (2.88) (2.20) Controlled Corporation *** *** (-3.77) (-3.57) Good Governance Dummy 2.019*** 1.903*** (5.91) (5.38) CCGG % Ownership (1.12) (0.80) (0.88) (0.67) Institutional % Ownership ex. CCGG (-0.06) (1.20) (0.37) (1.44) Turnover * * (1.41) (1.70) (1.38) (1.70) Log(1+Firm Media Cites) 0.538*** 0.398** 0.430** 0.308* (3.24) (2.54) (2.44) (1.88) Say on Pay Proposal 3.139*** 2.672*** 3.430*** 2.909*** (4.89) (4.49) (5.12) (4.95) Cross-Listed Firm 1.280*** 0.920** 1.143*** 0.800* (3.03) (2.24) (2.65) (1.90) Other Control Variables YES YES YES YES Year & Industry FE YES YES YES YES Observations Pseudo R

22 CCGG Impact on Compensation Policy Adoptions (Table 7B) (6) (7) (8) (9) Similar to Say on pay But shareholder proposals do not have a similar impact and Crosslisted not significant Engaged on Compensation Policies 2.273*** 2.123*** 2.464*** 2.344*** (5.60) (5.18) (5.94) (5.57) Engaged Firm (-1.31) (-1.62) Controlled Corporation *** *** (-2.71) (-2.77) Good Governance Dummy 0.753*** 0.832*** (3.31) (3.55) CCGG % Ownership (0.84) (0.47) (0.96) (0.59) Institutional % Ownership ex. CCGG (-0.77) (-0.38) (-0.88) (-0.50) Turnover (0.99) (1.21) (1.05) (1.30) Log(1+Firm Media Cites) 0.187* ** 0.172* (1.91) (1.36) (2.15) (1.75) Compensation Policy Proposal (0.62) (0.51) (0.67) (0.56) Cross-Listed Firm (0.51) (0.16) (0.67) (0.35) Other Control Variables YES YES YES YES Year & Industry FE YES YES YES YES Observations Pseudo R

23 Outline 1. Context and Background 2. Data and Empirical Framework 3. Results: Engagements and Adoptions 4. Results: Spillovers 5. Discussion: Broader Influence 6. Key Choices? 7. Limits to Collective Action 8. Conclusions 23

24 CCGG Impact Beyond Engaged Firms? Prior results show significant impact on engaged firms for all 3 campaign issues Is CCGG impact limited to engaged firms A small subset of all listed firms Spillovers from engaged firms to non-engaged firms? One channel directors of engaged firms sit on other firm s boards Influential directors (that sit on other boards), can magnify impact Initially a strategy of CCGG to focus on influential directors 24

25 Nice Experiment: Majority Voting (Table 8B) CCGG first engaged with boards of 7 major financial institutions These directors are more influential Are the other firms (non-engaged) with overlapping directors more likely to adopt? 2005 Mean Median Number of Other Directorships Big bank directors Non-big bank directors Difference 1.0*** 1.0*** Test statistic for difference Log(Market Cap of Other Directorships) Big bank directors Non-big bank directors Difference 4.4*** 1.6*** Test statistic for difference

26 Does CCGG Increase its Impact through Spillovers via Director Interlocks?(Table 9) Majority voting Say on Pay (1) (2) (3) (4) (5) (6) (7) Non-engaged firms only Engagement spillovers via interlocks Big bank directors important in majority voting Sizable impact Marginal effect = 15-18% Constant *** *** *** *** *** *** *** (-3.96) (-4.29) (-2.68) (-2.65) (-8.56) (-6.69) (-7.15) Interlock *** 1.009*** 0.679* (1.37) (2.69) (2.75) (1.89) Big Bank Interlock 1.813*** 2.435*** 2.260*** (3.15) (2.63) (2.61) Non Big Bank Interlock (0.31) (0.26) (0.28) Controlled Corporation *** (-0.60) (-3.48) Good Governance Dummy 1.640** 2.285*** (2.06) (5.68) Other Control Variables YES YES YES YES YES YES YES Year & Industry FE YES YES YES YES YES YES YES Observations Pseudo R

27 Outline 1. Context and Background 2. Data and Empirical Framework 3. Results: Engagements and Adoptions 4. Results: Spillovers 5. Discussion: Broader Influence 6. Key Choices? 7. Limits to Collective Action 8. Conclusions 27

28 Do Other CCGG Strategies Impact Governance? Policy Initiatives Very attractive potential alternative to firm-level engagements, get regulatory change imposed on all issuers with same approach Is main focus of collective action organizations in UK, US Political theories of governance however predict II will have hard time face other groups with advantages in organizing, funding and political influence, and II may have internal conflicts II do not control the agenda as is political process Recent and significant success in 2014 as the TSX made majority voting a listing requirement much credit given to CCGG Likely important that prior engagements/adoptions and measurement had laid ground work to understand costs and benefits of change 28

29 Do Other CCGG Strategies Impact Governance?: Informal Regulation Develop best practice principles and guidelines II controls agenda See much wider set of principles, and more open to change compared to regulation Not minimum standards, but aspirational goals Have to persuade firms to adopt Measurement of current practices against principles important to success CCGG shapes and/or subsidizes measurement Supports private engagements, providing data to help issuers to decide whether and how to respond. Allows for public rewards and penalties to supplement private engagements, as supports actions of others such as media. 29

30 Shaping guidelines and subsidizing measurement and subsequent changes in median Board Games Governance Scores What is measured appears to get managed. CCGG helps to pay for data collection on majority voting compliance. Shares data and some of new data incorporated into scores in following year. CCGG (and large investors)pays for data collection on compensation disclosure. Shares data and some of new data is incorporated into scores in following year. 30

31 Changing the Score Drop criteria all firms do well on, add new criteria, some of which CCGG inspired and measured 31

32 Outline 1. Context and Background 2. Data and Empirical Framework 3. Results: Engagements and Adoptions 4. Results: Spillovers 5. Discussion: Broader Influence 6. Discussion: Key Choices? 7. Limits to Collective Action 8. Conclusions 32

33 Key choices? Compare relative to investor activity in other settings Invest significantly in activities where II control the agenda i.e. Private engagements and informal regulation More limited success in regulatory changes Drive change by shaping norms and leveraging social networks and reputation rather than public confrontations Subsidized measurement important Allows others to create sanctions for bad behavior Build and maintain a broad coalition to maximize perceived impact and limit perceived distortions in agenda Funding model, Leadership choices and bylaws Narrow, rather than broad focus Very different issues than hedge fund activists Not about E &S, just traditional governance 33

34 Limits to II Collective Action? Focus on structural / process issues Requires less firm specific information Easier for the diverse group of investors to reach consensus Different from other activists such as hedge funds Focusing on norms and not regulation means less (more) success with Controlled corporations (Good governance firms) By focusing exclusively on governance and not other activities, provide limited benefits exclusive to members, raising longer-term risks Other groups e.g. ICD - have advantages by having education delivery and directors registrar producing exclusive benefits 34

35 Conclusion II can generate power through coordination and collective action to improve firms governance CCGG engagements adoptions of majority voting, say on pay, and compensation structure and disclosure Significant economic impact Impact beyond engaged firms Director interlocks Public policy and informal regulation But also limits and tradeoffs 35

36 Lessons for Institutional Investors Elsewhere? Domestic II activism can be effective - not just foreign institutional ownership Broadening Activities beyond Public Policy Worth Consideration as Mutually Reinforcing and Where Can Control Agenda Private engagements of individual issuers Increases probability of change, limits organized responses against reforms and provides info to shape norms Shaping and subsidizing public measurement Lowers costs to get issuers to respond to private and public engagements/shapes social norms/allows others to be negative and create penalties Public policy greatest success when costs and benefits of change clarified by private engagements, and public measurement. One of few ways to shift politics stacked against II 36

37 Creating a broad coalition of different investor types can yield benefits Increases perceived power, perceived legitimacy by avoiding concerns about political or other bias But not easy Need a fair funding model that captures differential interests in governance; leadership choices and bylaws which allow pension plans a leading but not dominant role; likely a limited number of investors And has limitations some issues won t be addressed because of lack of consensus, potentially slow 37

38 Thank you Comments welcome 38