Is (Institutional) Shareholder Activism New? Evidence from UK Shareholder Coalitions in the Pre‐Cadbury Era

• DOI: http://doi.org/10.1111/j.1467-8683.2010.00795.x
• Author: Rafel Crespi,Luc Renneboog
• Published date: 01 July 2010
• Date: 01 July 2010

Is (Institutional) Shareholder Activism New?

Evidence from UK Shareholder Coalitions in the

Pre-Cadbury Eracorg_795274..295

Rafel Crespi and Luc Renneboog*

ABSTRACT

Manuscript Type: Empirical

Research Question/Issue: Does recent institutional shareholder activism have a precedent in the form of (ad hoc) share-

holder coalitions that force the manager to leave poorly performing companies?

Research Findings/Insights: To address the above question, we investigate whether or not shareholder voting coalitions

disciplined managers in the pre-Cadbury era. We f‌ind that the voting power held (Shapley values) by f‌inancial institutions

(mainly, the insurance companies) are positively related to executive director turnover. Coalitions of industrial and com-

mercial companies with strong relative voting power penalize poorly performing managers. In contrast, coalitions of

executive directors are able to resist board restructuring. Finally, we a strong relation between top managementsubstitution

and high debt levels, especially in companies in need of ref‌inancing.

Theoretical/Academic Implications: We use Shapley values which are power indices developed in the f‌ield of (oceanic)

game theory in order to capture the relative voting power of a shareholder in all possible shareholder coalitions. A

shareholder who is pivotal in many potential coalitions is hence powerful and received a high Shapley value. The use of

such power indices yield much stronger results in our models that any of the other measures traditionally used in corporate

governance research to measure voting power/ownership concentration. This methodology enables us to draw clear

conclusions on the existence of voting coalitions.

Practitioner/Policy Implications: This paper shows that voting coalition of shareholders can be instrumental in bringing

about change in poorly performing companies.

Keywords: Corporate Governance, Corporate Ownership and Control, Shareholder Coalitions, Shapley Values, Voting

Power

INTRODUCTION: SEPARATION OF

OWNERSHIP AND CONTROL

The process of coalition formation inevitably begins with one or

more fund managers deciding that a company is seriously

underperforming – in a way that can be changed by shareholder

intervention. [. . .] an insurer also had as many as twenty

discussions a year with one or more other institutions about the

need to “strengthen the boards” of f‌irms in its portfolio (Black

& Coffee, 1994:2047–2051).

In the traditional international corporate governance lit-

erature, governance systems are classif‌ied according to

the trade-off between liquidity and control (see, e.g., Becht &

Röell, 1999). In Anglo-American or market-based gover-

nance systems – characterized by a large number of listed

companies, strong boards, and the prevalenceof institutional

investors-individual investors and investment funds benef‌it

from high trading liquidity of shares. Still, the high free f‌loat

may create a serious agency problem, sometimes referred to

as “strong managers, weak owners” (Roe, 1994). In contrast,

the blockholder system refers to Continental European

countries and Japan that have a relatively small number

of listed companies, concentrated ownership, complex

cascade-like shareholding structures with powerful

*Address for correspondence: Tilburg University, PO Box 90153, 5000 LE Tilburg,

Netherlands. E-mail: Luc.Renneboog@uvt.nl

274

Corporate Governance: An International Review, 2010, 18(4): 274–295

© 2010 Blackwell Publishing Ltd

doi:10.1111/j.1467-8683.2010.00795.x

controlling families, industrial f‌irms, or holding companies

(Barca & Becht, 2001). In this system, share liquidity is low

and agency costs of a different type may arise, namely those

between majority and minority shareholders.

Given that controlling a f‌irm is expensive (in terms of

reduced portfolio diversif‌ication), several mechanisms have

been developed in countries with a blockholder governance

system to deviate from the one-share-one-vote rule. By uti-

lizing dual class shares, ownership pyramids, or proxyvotes

(as used by, e.g., German banks), some shareholders can

exert a higher degree of control than their cash f‌low rights

would lead one to expect. Likewise, in the market-based

governance systems, several mechanisms exist to reduce

agency problems or to avoid that the balance of power is

tilting too much towards the management. First, the lack of

shareholder monitoring may be compensated by an active

take-over market. In contrast to US evidence (e.g., Martin &

McConnell, 1990), Franks and Mayer (1996) are not con-

vinced that the market for hostile take-overs in the UK ful-

f‌ills a disciplinary role. Second, management remuneration

schemes can be designed to align managerial interests with

those of the shareholders. However, a large body of empiri-

cal evidence has shown that the compensationpackages con-

sisting of salary, bonus, long-term incentive plans, and

option schemes, are more related to corporate growth than

value creation (e.g., Bebchuk & Fried 2003; Bertrand & Mul-

lainathan 2001; Conyon & Murphy, 2000). Third, it may be

that, in spite of the diffuse nature of control concentration,

UK blockholders actively monitor f‌irms, and reduce the

problem of free-riding on control by creating voting pacts.

This paper attempts to investigate the monitoring role of

shareholder voting coalitions in the UK. Testing the impact

of such coalitions directly is not possible because their exist-

ence is usually not disclosed. The reason is that the regula-

tory authorities consider a long-term shareholder coalition

(investor group) as one shareholder who has to comply with

all regulations concerning information disclosure, manda-

tory tender offer, disclosure of strategic intent, etc. There-

fore, if coalitions are formed, they exist on ad hoc basis with

a specif‌ic aim such as the removal of underperforming man-

agement. Black and Coffee (1994:2003) state, “British institu-

tions are signif‌icantly more active than their American

counterparts. One likely reason is that American institutions

cannot act jointly and quietly, in the preferred British

pattern.” The authors provide several case studies of share-

holder coalitions that led to managerial disciplining and the

removal of the entire board (e.g., a coalition of institutions

led by Norwich Union replaced the entire board of Tace,

PLC). Therefore, we develop a formal (albeit indirect) test to

examine whether disciplining actions are the consequence

of interactions between shareholders with high relative

voting power.

To our knowledge, little empirical research has been per-

formed on shareholder voting coalitions. Interesting theo-

retical models were developed by Zwiebel (1995) and

Bennedsen and Wolfenzon (2000). The former paper devel-

ops a game theory model in which small block shareholders

join to form controlling coalitions. This deters other block

investors such that the ruling coalition will not be chal-

lenged. The latter paper shows that, in the absence of a resale

market for shares, a f‌irm’s founder may optimallychoose an

ownership structure with multiple large shareholders in

order to prevent a single shareholder from taking unilateral

actions that may hurt other shareholders and the founder.

Thus, the founder may benef‌it from the fact that coalitions

compete to seize control. Recently, Maugand Rydqvist (2004)

have developed a model in which voting is strategic in the

sense that shareholders take into account the information of

other shareholders when making their voting decisions. In

our study, we also allow insider and outsider blockholders to

compete for control. If the outsiders win, they replace man-

agement. If the management wins, they are able to impede

board restructuring by coalitions of outside shareholders,

even in the wake of poor corporate performance.

Belot (2008) mentions that the use of shareholder agree-

ments is far from anecdotal;a recent study commissioned by

the European Commission (based on ISS, Shearman, Ster-

ling, & ECGI, 2007) shows that shareholder agreements are

encountered in 12 per cent of European listed companies. It

should be noted that shareholder agreements are a much

broader concept than shareholder voting agreements as

shareholder agreements exist which do not include clauses

on voting behavior. Whereas there is no disclosure require-

ment for shareholder voting agreements in some countries

such as the UK or Spain (Gutiérrez & Tribo, 2004), share-

holder pacts (which may include voting clauses) should be

made public in Italy(Baglioni, 2008) and France (Belot, 2008).

Baglioni (2008) reports states, “On one side, a strong domi-

nant owner may be able to extract a high level of private

benef‌its of control. On the other side, a weak f‌irst share-

holder may be less effective in monitoring managers; more-

over, an evenly distributed power among several large

shareholders increases the likelihood of a gridlock in deci-

sion making. Agreements may be seen as a way to correct

such extreme situations, by manipulating the voting power

of participants.”

One may ask why shareholder coalitions are formed by

means of voting pacts as shareholders could just as well

make proxy proposals and see support at the annual

meeting.

Two recent working papers (Cziraki, Renneboog, and

Szilagyi (2009) for Europe and Renneboog and Szilagyi

(2008) for the US) focus on observable shareholder activism,

namely the voting behavior of shareholder on the annual

meeting. For UK and continental European f‌irms, Cziraki

et al. (2009) investigate the corporate governance role of

shareholder-initiated proxy proposals. These proposals are

legally binding in the UK and most of continental Europe.

Nonetheless, submissions remain relatively infrequent in

Europe. Proposals enjoy limited voting success. Hardly any

empirical research has been done on shareholder coalitions

apart from the interesting clinical paper by Becht, Franks,

Mayer, and Rossi (2007) on institutional shareholder activ-

ism. In their unique clinical study based on information on

the many possible shareholder activist actions, the authors

show that most activism is happening behind the scenes.

Such shareholder actions include contact with management

(CEO, CFO, non-executive directors, etc.), contact with other

shareholders (e.g., with the intention to build a voting coa-

lition) and other stakeholders (such as banks, brokers, etc.),

proxy proposals at the annual meeting, a press campaign,

hostile takeover attempt, litigation, or class action. So, voting

SHAREHOLDER ACTIVISM BY COALITIONS 275

Volume 18 Number 4 July 2010© 2010 Blackwell Publishing Ltd