Impact of Restricted Voting Share Structure on Firm Value and Performance

• Date: 01 September 2010
• Author: PengCheng Zhu,Shantanu Dutta,Vijay Jog
• Published date: 01 September 2010
• DOI: http://doi.org/10.1111/j.1467-8683.2010.00808.x

Impact of Restricted Voting Share Structure on

Firm Value and Performancecorg_808415..437

Vijay Jog, PengCheng Zhu*, and Shantanu Dutta

ABSTRACT

Manuscript Type: Empirical

Research Question/Issue: In this study, we examine whether f‌irm value, operating performance, and stock performance of

the Canadian f‌irms which have issued restricted voting shares (RVS) is different from comparable non-RVS f‌irms. We test

two competing hypotheses – the controlling shareholder expropriation hypothesis and the investor protection and substi-

tution hypothesis for the Canadian RVS f‌irms.

Research Findings/Insights: Based on a ten-year panel data sample and extensive robustness tests, we do not f‌ind that the

RVS f‌irms have lower f‌irm value, operating performance, or stock performance than the non-RVS f‌irms in Canada. There is

also no evidence of shareholder value expropriation in key f‌inancial decisions, such as mergers and acquisitions and

dividend payments. The study does not support the controlling shareholder expropriation hypothesis in the Canadian RVS

f‌irms.

Theoretical/Academic Implications: The RVSstructure has been criticized for lack of shareholder rights protection and risk

of value entrenchment in the corporate governance literature. However, the existing empirical evidence is mainly based on

the f‌irms in the United States. Using a Canadian sample, we show that these results are a function of corporate governance

and regulatory environment and the US results cannot be generalized. We f‌ind no evidence of shareholder value appro-

priation nor do we f‌ind differential value or performance implications. We believe that a blanket rejection of RVS structure

may not be warranted.

Practitioner/Policy Implications: Our study concludes that the RVS structurein Canadian f‌irms is inf‌luenced by the nature

of these f‌irms and the regulatory environment in Canada that protects minority shareholders from wealth appropriation.

Our evidence suggests that policy makers and practitioners should evaluate RVS f‌irms on their individual merit and in

individual countries – no generalization is warranted.

Keywords: Corporate Governance, Canada, Dual-class Shares, Financial Performance, Shareholder Value

INTRODUCTION

The relationship between ownership structure, voting

rights, and f‌irm performance, especially within the

context of f‌irms with restricted voting share (RVS hereafter),

has been a continued source of debate among practitioners

and academicians alike. In essence there are two schools of

thought. Some claim that a deviation from a “one share one

vote” rule allows for an entrenchment of controlling minor-

ity shareholders (CMSs) from other external and internal

governance mechanisms. This is likely to give them the

power to expropriate non-controlling shareholders and

resulting in negative consequences on shareholder wealth

(e.g., Grossman & Hart, 1988). On the other hand, propo-

nents of the RVS structure argue that such structure allows

the controlling shareholders to manage and direct f‌irms

with long-term value creation perspectives (SHARE, 2004),

and realize its growth potential (Lehn, Netter, & Poulsen,

1990). RVS structures also allow families and entrepreneurs

to continue to control f‌irms and motivate them to make f‌irm

specif‌ic investment in their human capitals (Amoako-Adu&

Smith, 2001; Bergstrom & Rydqvist, 1990; Cronqvist &

Nilsson, 2003). RVS f‌irms may increase the premium in a

takeover attempt as the controlling shareholders have an

interest in extracting a higher valuefor the shares (Comment

& Schwert, 1995; Rydqvist, 1996).

In order to resolve this debate, a number of empirical

studies have directly examined the effect of a RVS structure

on f‌irm value and performance and investigated the mecha-

nisms of expropriation with mixed results. Some studies

*Address for correspondence: Eberhardt School of Business, University of the Pacif‌ic,

3601 Pacif‌ic Avenue, Stockton, California, USA 95211. Tel: 1-209-946-3904; E-mail:

pzhu@pacif‌ic.edu

415

Corporate Governance: An International Review, 2010, 18(5): 415–437

© 2010 Blackwell Publishing Ltd

doi:10.1111/j.1467-8683.2010.00808.x

show that deviation of ownership and control leads to lower

f‌irm value (i.e., Tobin’s Q) and poorer performance (e.g.,

Claessens, Djankov, & Lang, 2000; Gompers, Ishii, &

Metrick, 2009; Lins, 2003; Yeh & Woidtke, 2005). Others f‌ind

that recapitalization of one class of common shares into

restricted voting shares is a valueenhancing initiative for the

general shareholders (Dimitrov & Jain, 2006).

Many of the studies are based on evidence of the US

market, where the number of f‌irms with restricted voting

shares is considerably low–only six per cent of the US f‌irms

have the RVS structure (Gompers et al., 2009) compared to

15 per cent in the Canadian market (SHARE, 2004). This

makes Canada an excellent research environment to inves-

tigate the RVS structure since it also has a well-developed

capital market with a strong legal environment and thus has

a high cost for reputation loss. However, we believe that

there are two key differences between Canada and the US

that have signif‌icant implications and motivation for our

study. First, Canada has a unique regulatory environment

for the RVS f‌irms. It has instituted a compulsory coattail

provision regime that prevents the expropriation of share-

holder wealth, and thus eliminating the potential wealth

appropriation from restricted voting shareholders in the

event of an acquisition where the acquirer may only choose

to acquire voting shares. Second, unlike their US counter-

parts, the Canadian f‌irms tend to have more concentrated

ownership structure and family control (Bozec, 2007). We

believe that these two conditions may lead to different out-

comes with respect to the impact of RVS structure on f‌irm

value and their operating and stock market performance.

In this study, we thus focus on publicly listed Canadian

f‌irms and examine one major research question that should

concern managers, investors, and policy makers – does RVS

structure destroy shareholder value? To answer this ques-

tion, we carry out an extensive empirical analysis of market

valuation, operating performance, and long-term stock per-

formance of RVS f‌irms. We also examine the possible private

benef‌it of expropriation by studying the M&A activities and

dividend payments undertaken by the RVS f‌irms. We care-

fully address many econometric issues in our empirical

analysis and ensure that our conclusions are robust across

various matching techniques, sample selection bias, endoge-

neity problems, and clustering effect in the panel dataset.

The study is based on a 10-year panel dataset of all RVS

f‌irms and non-RVS f‌irms listed on the Toronto Stock

Exchange between 1996 and 2005. Our results indicate that

RVS f‌irms do not have lower f‌irm value (Tobin’s Q), operat-

ing performance (ROA), or long term stock performance

than non-RVS f‌irms. We further test a possible “channel” of

controlling shareholders’ expropriation by using a compre-

hensive sample of Canadian M&A transactions and we fail

to f‌ind any evidence that the RVS f‌irms destroy more share-

holder value than the non-RVS f‌irms in these transactions.

Furthermore, we note that RVS shareholders receive similar

or even more dividend paymentthan the normal sharehold-

ers, which again leads to a rejection of the controlling share-

holder expropriation hypothesis in the Canadianmarket. We

conduct many robustness tests and f‌ind similar results.

Our study contributes to the literature in several ways.

First, contrary to the existing primary US based empirical

f‌indings and conventional wisdom, our results show that a

blanket condemnation of RVS f‌irms should be viewed with

caution. Although most of the recent US (and some interna-

tional) studies show a negative impact of RVS structure on

f‌irm value(e.g., Cronqvist & Nilsson, 2003; Gompers, Ishii, &

Metrick, 2009), our results show that in Canada, RVS struc-

ture is not detrimental to f‌irm value. While we do not ques-

tion the f‌indings of earlier studies, we argue that market

structure and regulatory environment, nature of the RVS

f‌irms, and alternative investment choices available to the

market participants are likely to play a signif‌icant role in

determining the relation between RVS structure and f‌irm

value. Second, in contrast to the studies that independently

investigate the relation between RVS structure and f‌irm

value, or channel of expropriation in RVS f‌irms, we examine

both issues together to ensure consistency in our results. This

is important,because if we do not f‌ind support for the expro-

priating behaviorof RVS f‌irms, the previous results reporting

a negative relation between RVS structure and f‌irm value

would be puzzling and questionable. We investigate this by

examining two possible channels of expropriation in RVS

f‌irms namely the M&A activities and the dividend paying

behavior of the RVS f‌irms. Third, we conduct extensive

robustness tests to ensure thatour f‌indings stand the scrutiny

of alternative methodologies and sample selections. More

specif‌ically, we use four proxies and approaches to classify

RVS structure, by using a dummy variable (to contrast the

RVS f‌irms with the non-RVS f‌irms), categorical variables (to

differentiate the different types of restricted voting shares),

and a wedge measure as well as a ratio measure to represent

the deviation between the cash f‌low right and the voting

right.We test the impact of eachof these measures on the f‌irm

performance and value. We also conduct other robustness

tests including propensity score matching, correction for

sample selection bias and endogeneity problem and adjust-

ment for both time and f‌irm clustering effect in the panel

dataset. In addition, we believe that this is one of the few

studies to employ a panel data sample to examine this issue.

As is well known, the use of paneldata providesmore reliable

evidence than a simple year cross-sectional study.

The rest of the paper is organized as follows. In the next

section we introduce the regulatory and institutional envi-

ronment for RVS f‌irms in Canada, discuss the relevant lit-

erature and develop the hypotheses. Then we introduce the

dataand the relevant methodologies followed by the empiri-

cal results; and f‌inally we summarize the paper and provide

concluding discussions.

LITERATURE REVIEW

Canadian Corporate Governance Environment

As stated earlier, one of the signif‌icant contributions of this

study is that it focuses on Canadian RVS f‌irms in contrast

with other studies that analyze the US RVS f‌irms. Since,

potential agency problems related to the RVS structure in a

f‌irm might be attributed to the corporate governance and

regulatory practices in a country, we f‌irst discuss the Cana-

dian corporate governance environment and its implica-

tions. We also highlight the differences between the

Canadian and the US corporate governance practices, to

explain the uniqueness of a Canadian study.

416 CORPORATE GOVERNANCE

Volume 18 Number 5 September 2010 © 2010 Blackwell Publishing Ltd