Ties That Bind: How Business Connections Affect Mutual Fund Activism

• Author: KONSTANTINOS E. ZACHARIADIS,AMIL DASGUPTA,DRAGANA CVIJANOVIĆ
• DOI: http://doi.org/10.1111/jofi.12425
• Date: 01 December 2016
• Published date: 01 December 2016

THE JOURNAL OF FINANCE •VOL. LXXI, NO. 6 •DECEMBER 2016

Ties That Bind: How Business Connections Affect

Mutual Fund Activism

DRAGANA CVIJANOVI ´

C, AMIL DASGUPTA,

and KONSTANTINOS E. ZACHARIADIS∗

ABSTRACT

We investigate whether business ties with portfolio firms influence mutual funds’

proxy voting using a comprehensive data set spanning 2003 to 2011. In contrast to

prior literature, we find that business ties significantly influence promanagement

voting at the level of individual pairs of fund families and firms after controlling for

Institutional Shareholder Services (ISS) recommendations and holdings. The associa-

tion is significant only for shareholder-sponsored proposals and stronger for those that

pass or fail by relatively narrow margins. Our findings are consistent with a demand-

driven model of biased voting in which company managers use existing business ties

with funds to influence how they vote.

MUTUAL FUNDS ARE OF GREAT importance to both retail investors and corpo-

rations. They are the main investment vehicle for retail investors, not least

via their role in managing pension portfolios through 401(K) plans. They are

also highly relevant to corporate governance as they collectively own 24% of

U.S. corporate equity1and large fund families hold blocks of 10% or more in

dozens of large U.S. corporations (Davis and Yoo (2003)). Since the passage of

shareholder proposals raises firm value (Cu ˜

nat, Gine, and Guadalupe (2012)),

if mutual funds vote their proxies in a manner that enhances the value of port-

∗Cvijanovi´

c is at the Kenan-Flagler Business School, University of North Carolina at Chapel

Hill. Dasgupta is at the London School of Economics and affiliated with the CEPR and ECGI.

Zachariadis is at the London School of Economics. Weare grateful to the Editor, Kenneth Singleton,

and two anonymous referees for insightful input. We thank Miloˇ

sBo

ˇ

zovi´

c, Vicente Cu ˜

nat, Daniel

Ferreira, Nick Gantchev, Christian Julliard, Dong Lou, Gregor Matvos, Adair Morse, Abhiroop

Mukherjee, Daniel Paravisini, Michael Roberts, Breno Schmidt, Geoff Tate, Moqi Xu, Jonathan

Zinman, and audiences at the Belgrade Young Economists Conference, LSE, NBER Summer In-

stitute 2014 Law and Economics Workshop, NC State, University of Lugano, and WesternFinance

Association 2014 for helpful comments. Cvijanovi´

c thanks the Rollie and Mary Windley Tillman

Endowment Fund and the Wells Fargo Center for Corporate Finance at UNC Chapel Hill for fi-

nancial support. Dasgupta and Zachariadis thank the Paul Woolley Centre and the Department

of Finance at the LSE for financial support. Dasgupta thanks the Cambridge Endowment for Re-

search in Finance and the Faculty of Economics at Cambridge University for their hospitality. We

have read the Journal of Finance’s disclosure policy and have no conflicts of interest to disclose.

1See Investment Company Institute (for the full 2013 Investment Company Fact Book, visit

http://www.ici.org/pdf/2013_factbook.pdf).

DOI: 10.1111/jofi.12425

2933

2934 The Journal of Finance R

folio firms,2they would not only play a beneficial role in corporate governance,

but also enrich their vast base of retail clients. The proxy voting behavior of

mutual funds is thus of considerable importance.

However,mutual funds often have lucrative business relationships with port-

folio firms arising from the management of the same 401(K) plans that enhance

their importance to retail investors. On average, earnings from 401(K)-related

business equal 14% of the revenues that mutual fund families earn from their

equity funds, and such income can represent as much as 25% of fund family

revenues (Davis and Kim (2007)). Since the choice of fiduciaries for 401(K)

plans lies in the hands of firm executives who may be opposed to shareholder

activism, there has been widespread suspicion that mutual funds may vote

their proxies in a conflicted manner. For example, according to the SEC:3

. .. in some situations the interests of a mutual fund’s shareholders may

conflict with those of its investment adviser with respect to proxy voting.

This may occur, for example, when a fund’s adviser also manages. . . the

retirement plan assets of a company whose securities are held by the fund.

In these situations, a fund’s adviser may have an incentive to support

management recommendations to further its business interests.

In response to such concerns, in 2003 the SEC adopted Rule 30b1-4 of the

Investment Companies Act, which requires mutual funds to annually disclose

their votes cast for proposals arising in portfolio companies. As a result, it is

now possible to investigate the extent to which business ties with portfolio

firms affect mutual fund proxy voting. We examine this relationship over the

2003 to 2011 period.

We show that mutual funds’ voting is significantly influenced by their busi-

ness ties with portfolio firms. Since we control for unobserved heterogeneity

using a rich set of fixed effects, our result holds for given pairs of fund fami-

lies and firms, even at the level of individual proposals, and after controlling

for Institutional Shareholder Services’ (ISS) recommendations and holdings.

Our finding stands in sharp contrast to prior literature (Davis and Kim (2007),

Ashraf, Jayaraman, and Ryan (2012)), which finds that business ties with

portfolio firms do not influence voting after controlling for fund family hetero-

geneity. We discuss the connection to these papers in greater detail below.

Proposals can be sponsored either by management (insiders) or by share-

holders (outsiders). In the full sample, we find a robust incremental effect of

business ties on promanagement voting for shareholder proposals relative to

management proposals. When we split the sample by sponsorship, we find that

the association between ties and voting only obtains for shareholder propos-

2Mutual funds have a fiduciary duty to vote proxies in the best interest of their investors

(SEC Rule 206(4)-6; Securities & Exchange Commission, “Final Rule: Proxy Voting by Investment

Advisers,” available at https://www.sec.gov/rules/final/ia-2106.htm).

3Securities & Exchange Commission, “Final Rule: Disclosure of Proxy Voting Policies

and Proxy Voting Records by Registered Management Investment Companies,” available at

http://www.sec.gov/rules/final/33-8188.htm.

Ties That Bind 2935

als. Viewed in light of canonical models of corporate governance, which hold

that outsiders attempt to mitigate rent extraction by insiders, the possibility

that management may use business ties to obstruct shareholder activism is of

particular interest.

We develop a simple model that provides a theoretical foundation for the in-

cremental effect identified above. In our model, shareholder and management

proposals are distinguished by the degree to which company managers have

control: managers can withdraw or modify problematic self-sponsored propos-

als (Listokin (2008)) but have less control over the specifics of shareholder pro-

posals, which typically arise as the result of a breakdown in soft engagement

between managers and shareholders (Gantchev (2013), McCahery, Sautner,

and Starks (2016)). Managers receive information on the anticipated support

for each proposal. If a proposal is formally introduced, they have the option

to try to influence the voting of institutional blockholders. Managers are best

able to influence those institutions with which they have ongoing business

relationships, but exerting influence is costly because it requires effort and

credible threats of future punishment. Accordingly, managers exert influence

only when doing so is worth paying this cost. The model predicts that managers

influence voting more frequently for shareholder proposals than for manage-

ment proposals, thus providing an explanation for the empirical finding above:

the flexibility and control that managers have over self-sponsored proposals

raises the bar for paying the cost of exerting influence.

Our model also sheds further light on the relationship between proxy voting

and business ties. It predicts that managers are more likely to influence voting,

ex ante, for shareholder proposals that pass or fail narrowly (i.e., that are

contested), ex post. Since managers receive information on anticipated support,

the fact that these proposals pass or fail narrowly ex post implies that managers

can predict ex ante that these proposals are “in play.” It is for precisely these

proposals that it is worth paying the cost of exerting influence. We empirically

examine the incremental effect of business ties on proxy voting on shareholder

proposals that pass or fail by relatively narrow margins and find strong support

for the model’s prediction. The effect is also economically relevant: in our most

saturated specification, we find that a shift from no business ties to some

business ties leads to an increase in promanagement voting of over 12% for

proposals that pass or fail by less than 20%.

Our analysis of contested proposals is of independent interest because it al-

lows us to pinpoint the conflict of interest confronting mutual fund families

with business ties. While individual fund families cannot be pivotal for propos-

als that pass or fail by wide margins, the same may not be true for contested

proposals. Furthermore, the passage of a proposal is likely to enhance the mar-

ket value of the firm on the day of the vote precisely when the proposal is

contested because in such cases the voting outcome may not be fully antici-

pated by the market (Core, Guay,and Rusticus (2006)). Thus, it is for contested

proposals that the trade-off between value enhancement and business ties is

salient: by voting for the proposal, the fund family may be instrumental in

raising firm (and, in turn, portfolio) value but displease management and be