Does mandatory adoption of international financial reporting standards decrease the voting premium for dual-class shares: theory and evidence

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

DOES MANDATORY ADOPTION OF
INTERNATIONAL FINANCIAL REPORTING STANDARDS
DECREASE THE VOTING PREMIUM FOR DUAL-CLASS SHARES:
THEORY AND EVIDENCE

by

Hyun A Hong

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BUSINESS ADMINISTRATION)

August 2010

Copyright 2010 Hyun A Hong

ii

Epigraph

“Sunshine is the best disinfectant.” (Brandeis, 1914)

iii

Dedication

To my father (Seongkang Hong) and mother (Sunjeong Kang),
for giving me unconditional love and support.

iv

Acknowledgments

I am significantly indebted to dissertation committee members, Mark DeFond, Mingyi
Hung (co-chair), John Matsusaka, KR Subramanyam and Robert Trezevant (co-chair) for
their continuous support and guidance. I thank Carolyn M. Callahan, Hans B. Christensen,
Harry DeAngelo, Craig Doidge, Yongtae Kim, Soo Young Kwon and workshop
participants at Korea University, Queen’s University, Seoul National University, the
University of Memphis, the University of Minnesota and the University of Southern
California for helpful comments and discussions.

v

Table of Contents

Epigraph ii

Dedication iii

Acknowledgements iv

List of Tables vi

List of Figures vii

Abstract

viii
Chapter 1: Introduction

1
Chapter 2: A simple model

9
Chapter 3: Institutional background and hypothesis development
Chapter 3.1. Dual-class firms
Chapter 3.2. Legal protection of minority shareholders against private control benefits
Chapter 3.3. The impact of the IFRS mandate on the voting premium of dual-class shares

13
15
16

Chapter 4: Sample and descriptive statistics
Chapter 4.1. Sample description
Chapter 4.2. Descriptive statistics

20
22
Chapter 5: Empirical analysis of primary hypothesis
Chapter 5.1. Research design
Chapter 5.2. Regression results

32
23
Chapter 6: Additional analysis
Chapter 6.1. Improvement in country-level accounting information quality as mechanisms
behind the voting premium effects of the IFRS mandate
Chapter 6.2. The effect of the IFRS mandate on the voting premium, conditional on firm-
specific return volatility

40

52
Chapter 7: Sensitivity Test
Chapter 7.1. US GAAP and Cross-listing
Chapter 7.2. Self-selection bias for voluntary adoption of IFRS
Chapter 7.3. Different benchmark groups

58
59
60

Chapter 8: Conclusion 65

Bibliography

67
Appendices:
Appendix I: An analytical model

Appendix II: Share structure across countries

72

74
vi

List of Tables

Table 1: Sample Composition by Country

24
Table 2: Firm-Level Descriptive Statistics for Variables Used in Regression
Analyses
Panel A: Descriptive Statistics for Variables Used
in Regression Analyses
Panel B: Pearson Correlation Coefficients

30
31
Table 3: Firm-Year Regression Analysis of the Voting Premium Effects around the
IFRS Mandate

38
Table 4: Additional Analysis on Institutional Characteristics of IFRS Adoption
Countries
Panel A: Descriptive Statistics for Country-Level Institutional
Variables
Panel B: Pooled Regression

48
49
Table 5: Additional Analysis: Firm-Year Regression Analysis of the Voting
Premium Effects around the IFRS Mandate, Conditional on Stock Return Volatility
Panel A:
Descriptive Statistics for Variables Used in Regression Analyses
Panel B: Pooled Regression

57
57
Table 6: Sensitivity Test: Firm-Year Regression Analysis
of the Voting Premium Effects around the IFRS Mandate, with Different Control
Samples

63

vii

List of Figures

Figure 1: The Voting Premium Effects of the IFRS Mandate

37
Figure 2: The Voting Premium Effects of the IFRS Mandate,
Conditional on the Degree of Disclosure

42
Figure 3: The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of Differences between Local GAAP and IFRS

43
Figure 4: The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of Restrictiveness of the Local GAAPs

44
Figure 5: The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of the Volatility of Superior Voting Shares
Relative to Inferior Voting Shares
55

viii

Abstract
Proponents of International Financial Reporting Standards (IFRS) claim that one of the
consequences of switching to IFRS is greater minority shareholder protection. In this
paper I test this assertion by setting up a model of the effect of the accounting standards
on controlling parties’ expropriating behavior. I then apply this model to firms where
control and cash flow rights are separated by a dual-class structure and examine whether
mandatory adopters’ voting premium is reduced following the IFRS mandate. Based on
panel data set of 151 dual-class firms in 13 mandatory IFRS adoption countries during
the period 2002 to 2007, I find that after the mandatory introduction of IFRS reporting,
mandatory adopters’ voting premium decreases, on average, by 13.2%. This effect for
mandatory adopters is statistically significant relative to the corresponding effects for
voluntary adopters and non-IFRS adopters. In addition, the effect of mandatory IFRS
adoption on the voting premium is more pronounced for firms where minority
shareholders face higher informational risk. I further show that the decrease in the voting
premium subsequent to the IFRS mandate follows because of: (1) expanded disclosure, (2)
enhanced comparability of firms’ reported information across borders and (3) a restricted
set of measurement methods under IFRS. Taken together, these results suggest that
mandatory IFRS adoption benefits minority shareholders by providing an effective
mechanism to constrain the private benefits of control.
1

Chapter 1. Introduction
In many countries, minority shareholders are poorly protected against controlling
parties’ diversionary behavior. By having disproportionately larger voting rights than
their cash flow rights through a dual-class structure, controlling parties efficiently divert
corporate value to advance their private benefits (Berle and Means, 1932; Jensen and
Meckling, 1976; Nenova, 2002; Doidge, 2003).
1
These benefits accrue exclusively to
controlling parties. Hence, these benefits are referred to as private control benefits, and
they have been a central topic of the corporate finance literature in the past two decades
because of their linkage to shareholder value destruction.
2

An important theory in accounting is that lack of transparency allows managers
and controlling shareholders (hereafter, MCOs) to divert corporate resource for their
private benefits (see Fan and Wong, 2002; Leuz, Nanda, and Wysocki, 2003; Haw, Hu,
Hwang, and Wu, 2004; Francis, Schipper and Vincent, 2005). In the context of dual-class
firms, the divergence of control and cash flow rights associated with the dual-class
structure renders MCOs immune to capital markets’ disciplining mechanisms (e.g.,
hostile takeovers and proxy contests). Hence, the dual-class structure provides MCOs an

1
These benefits include, for example, the power to elect a family member to the board of directors or to
appoint a family member as CEO. Private benefits may also include opportunities to engage in tunneling,
self-dealing, empire building, and the appropriation of corporate growth opportunities (Grossman and Hart,
1980; Barclay and Holderness, 1989; Shleifer and Vishny, 1997; La Porta et al., 1997, 1998; Johnson et al.,
2000; Nenova, 2003; Dyck and Zingales, 2004; Djankov et al., 2008).

2
Grossman and Hart (1988) and Harris and Raviv (1988), for example, theorized that a dual-class structure
is suboptimal compared with a single-class structure. This is because the dual-class structure separates
voting and cash flow rights, which in turn allows controlling parties to perpetuate their consumption of
private control benefits to the detriment of minority shareholders. Based on these theories, corporate
finance studies have focused on finding mechanisms that restrain controlling parties from consuming
private benefits—a higher level of statutory protection of minority shareholders and a higher level of law
enforcement (La Porta et al., 1997 and 1998; Johnson et al., 2000; Nenova, 2003; Dyck and Zingales, 2004;
Djankov et al., 2008), a higher degree of newspapers’ circulation and tax compliance (Dyck and Zingales,
2004) and a cross-listing on the U.S. exchanges (Doidge, 2003).
2

opportunity and incentive to expropriate minority shareholders’ wealth. The accounting
literature suggests that to help hide their stealing, MCOs often generate financial
statements with more estimation error and noise (Fan and Wong, 2002; Leuz et al., 2003;
Haw, Hu, Hwang, and Wu, 2004; Francis, Schipper and Vincent, 2005; Pae, Thornton
and Welker, 2006). This obfuscation of accounting information increases the information
asymmetry between MCOs and minority shareholders, allowing MCOs to continuously
consume control benefits without the knowledge of minority shareholders.
3,4

Building on the preceding accounting literature, this paper examines whether
private control benefits decrease after the mandatory introduction of International
Financial Reporting Standards (IFRS), a set of new internationally recognized accounting
standards proposed by the International Accounting Standards Board (IASB). IFRS has
been increasingly accepted by a majority of countries including the constituents of the
European Union. Proponents of the IFRS mandate argue that IFRS decreases information
asymmetry between minority shareholders and controlling parties by (1) increasing
disclosure, (2) improving the comparability of reported information across borders, and
(3) restricting a set of accounting methods that controlling parties can choose to obfuscate

3
Recent accounting studies, for example, provide supporting evidence that the divergence of control and
cash flow rights leads controlling parties to generate less informative accounting earnings (Francis,
Schipper and Vincent, 2005), and to engage in a higher degree of earnings management (Fan and Wong,
2002; Leuz et al., 2003; Haw, Hu, Hwang, and Wu, 2004; Pae, Thornton and Welker, 2006).

4
This point is well illustrated by Enriques and V olpin:“Whether shareholders effectively exercise their
rights to sell, vote and sue effectively depends on their access to information. An extensive regime of
disclosure may help alleviate agency problems in listed corporations…For example, mandatory disclosure
of related-party transactions and of directors’ compensation can be an effective tool to limit self-dealing by
those in control. Disclosure of price-sensitive information helps prevent insider trading.” (Enriques and
V olpin, 2007, page 9).

3

their diversionary behavior (see, for example, EC Regulation No. 1606/2002; Ashbaugh
and Pincus, 2001; Pae, Thornton and Welker, 2006; Daske et al., 2008; Li, 2010).
5

This lower information asymmetry subsequent to the mandatory introduction of
IFRS is expected to reduce MCOs’ expropriating behavior. The decrease of information
asymmetry, for example, reduces controlling parties’ ex ante incentives to expropriate
minority shareholders because controlling parties are aware of a higher probability that
their expropriating behavior will be caught under the IFRS reporting regime; additionally,
the lower information asymmetry is anticipated to increase the ex post abilities of the
designated capital market watchdogs (e.g., auditors, institutional investors, stock
exchanges and regulators) to more effectively detect and discipline diversionary behavior
(La Porta et al., 1997 and 1998; Doidge, 2004; Pae et al., 2006). These investor
protection consequences associated with the mandatory adoption of IFRS should be thus
associated with the decrease of the size of MCOs’ private control benefits.
Critics, in contrast, assert that such effects do not necessarily materialize (Ball et
al., 2003, 2006; Burgstahler et al. 2006; Christensen, Lee, and Walker, 2007; Daske et al.
2007, 2008). They argue that investor protection will not be strengthened after a
“Cartesian top-down push for uniform accounting standards” unless there is also a change
in the underlying political and economic institutions in a given jurisdiction (Sunder, 2007,

5
Unlike the domestic GAAPs of most countries, IFRS requires firms to disclose the name of the ultimate
controlling party [IAS 24.17] and, in the event of related party transactions, the nature of the related party
relationship as well as sufficient information about the transactions and outstanding balances for
shareholders to assess the potential effect of the relationship on the firm’s financial statements [IAS 24.18-
19]. IFRS also mandates the disclosure of key managers’ compensation, both in total and for each of the
following categories: short-term employee benefits, post-employment benefits, other long-term benefits,
termination benefits, and share-based benefits [IAS 24.17].

4

page 1).
6
Extant research has little to say about whether the private benefits of control are
in fact reduced following the mandatory introduction of IFRS around the world. This
paper seeks to fill this void in the literature by investigating whether the mandatory
introduction of IFRS is correlated with the decrease of the voting premium for firms with
a dual-class share structure.
The voting premium is defined as the percentage price differential between
superior voting shares and inferior voting shares. While quantifying the magnitude of
private control benefits is empirically challenging, corporate finance studies show that
focusing on firms with a dual-class share structure allows one to estimate private control
benefits through the voting premium (Zingales, 1994; Rydqvist, 1996; Modigliani and
Perotti, 1998).
7
Since voting rights can cause an allocation of private control benefits to
shareholders owning superior voting shares, the voting premium is a proxy for the
magnitude of private control benefits. Here, given that the voting premium is determined
by generic shareholders’ trading of inferior and superior shares at a public exchange, the
voting premium represents generic shareholders’ perception on a lower bound of the size
of private control benefits (Doidge, 2003).

6
Consistent with this argument, Niemeyer states that the claim that “IFRS would lead to improved
comparability and better investor protection were among a series of widely touted myths about IFRS” (CFO
journal article, 2009).
7
The corporate finance literature estimates the size of control benefits using two metrics (Dyck and
Zingales, 2004), namely, the voting premium and the privately negotiated price of a controlling block. This
paper relies on the first metric because there is more extensive data coverage for this variable across IFRS
adoption countries. The second metric, proposed by Barclay and Holderness (1989), is based on privately
negotiated prices of controlling block transfers in publicly listed firms. The price per share paid for the
controlling block represents the value of the cash flow benefits from the acquirer’s new ownership and the
private benefits from the acquirer’s controlling power in the firm. In contrast, the market price per share
after the controlling block transaction is announced represents only the cash flow benefits that non-
controlling shareholders expect to receive under the new management. Therefore, the difference between
the price paid by the acquiring party and the prevailing market price after the announcement represents the
private benefits of control accruing to the acquirer.

5

To test for the voting premium effects of mandatory IFRS adoption, I construct a
panel data set of 151 firms with dual-class shares from 13 countries over the period from
2002 to 2007, resulting in a sample of 868 firm-year observations. I then classify firms
into two distinct groups: (1) firms that are domiciled in the 13 sample countries where
IFRS reporting became compulsory in 2005 and that first adopt IFRS in 2003 and later
(mandatory adopters) and (2) firms that voluntarily adopt IFRS before 2003 (early
voluntary adopters).
8
The first of these groups is used as the treatment sample in the
analysis. The latter group is the control sample that allows me to control for the
possibility that a decrease in voting premium in the post-IFRS period is caused by
contemporaneous macroeconomic changes rather than by the introduction of IFRS
reporting per se.
9

The analysis generates several notable findings. First, the results show that, on
average, the voting premium for the mandatory adopter decreases by 13.2% after the
mandatory introduction of IFRS reporting, while the voting premium for the voluntary
adopters does not change. The decrease in the voting premium for the mandatory
adopters in the post-IFRS period remains statistically significant after controlling for

8
Note that voluntary adopters in 2003 or 2004 are reclassified as mandatory adopters. I classified firms
based on whether firms adopt IFRS two fiscal years before the IFRS mandate because firms may adopt
IFRS in 2003 or 2004 due to their anticipation of the IFRS mandate (2005). Hence, I prevent the possibility
that classifying these firms as voluntary adopters may cause the difference-without-distinction problems.
However, when I classified firms into voluntary and mandatory adopters based on the date of the IFRS
mandate (2005), the empirical results were qualitatively the same.

9
To examine whether the empirical results in this paper are sensitive to the choice of the control group, I
employed four different sets of control groups in addition to the early voluntary adopters (see the section
6.3. of the sensitivity test): (1) firms that use domestic GAAP and domicile in countries that mandate IFRS
for their domestic firms during the sample period, (2) U.S. firms, (3) randomly-selected firms that use
domestic GAAP and domicile in countries that do not mandate IFRS for their domestic firms during the
sample period, and (4) all firms that use domestic GAAP and domicile in countries that do not mandate
IFRS for their domestic firms during the sample period.
6

firm- and country-specific attributes such as firm size, voting power, the differential
between superior versus inferior voting shares’ dividend payouts and liquidity, and
country fixed or random effects. This evidence is consistent with the view that the high-
quality accounting information associated with IFRS strengthens minority shareholder
protection.
Second, I found that the voting premium effects after mandatory IFRS adoption
appear to be concentrated among firms whose superior voting shares have, on average,
more volatile returns relative to their inferior voting shares. Prior literature demonstrates
that a higher volatility of superior voting shares’ returns signals the arrival of new private
information among traders who hold the superior voting shares (e.g., controlling owners;
Ross, 1989; Maheu and McCurdy, 2004; Chordia, Huh, and Subrahmanyam, 2007).
According to my findings, the effect of the mandatory IFRS adoption on the voting
premium is greater for firms with a higher degree of information risk (as represented by
the higher volatility of the superior voting shares relative to the inferior voting shares).
Considering this, in conjunction with the prior literature, I conclude that mandatory IFRS
adoption appears to reduce this information risk.
Finally, the voting premium effects of the IFRS mandate are observed only (1) in
countries with a larger number of rules on disclosures required under the IFRS reporting
regime but not under the local GAAP reporting regime, (2) in countries with greater
inconsistencies between local GAAP and IFRS rules, and (3) in countries where IFRS
more significantly restricts the set of accounting measurement methods relative to their
local GAAPs. This evidence supports the view that the decrease in the voting premium
7

associated with the IFRS mandate arises through the following mechanisms: (1) the
increased transparency of firms’ financial information, (2) the enhanced comparability of
firms’ financial statements across jurisdictions, and (3) the restricted set of measurement
methods that MCOs can choose to conceal their diversionary behavior.
This study makes two primary contributions to the accounting and finance
literatures. First, my study adds to the emerging body of IFRS literature by providing
empirical evidence that private control benefits are reduced after the mandatory
introduction of IFRS. A large body of the IFRS literature implies that by reducing the
private benefits of control, the IFRS mandate is associated with a higher Tobin’s Q ratio
(Pae et al., 2006; Renders and Gaeremynck, 2007; Daske et al., 2008), an increased
market liquidity (Daske et al., 2008), a positive stock market reaction for firms with a
higher degree of information asymmetry in the European Market (Armstrong et al., 2009)
and a lower cost of equity capital (Li, 2010). My study contributes to this stream of work
by providing relatively more direct evidence that the IFRS mandate is correlated with the
decrease in the mandatory adopters’ private benefits of control.
Second, this study extends the literature on the determinants of the private
benefits of control (e.g., Barclay and Holderness, 1989; Zingales, 1994; Nenova, 2002;
Doidge, 2003; Dyck and Zingales, 2004). This literature finds that while the size of
private control benefits is larger in countries with weaker legal investor protection
mechanisms, cross-listing on U.S. exchanges substitutes for the lack of legal investor
protection mechanisms by constraining controlling parties’ expropriating behavior in
countries with weak legal investor protection mechanisms (Doidge, 2004). My study
8

contributes to this research stream by providing evidence that the IFRS mandate plays a
significant role in reducing the voting premium. This evidence implies that introducing
strong accounting standards may help countries to increase the protection of their
minority shareholders – an insight of particular relevance to the U.S., where standard-
setters currently plan to converge to IFRS reporting by 2014.
The remainder of the paper is organized as follows. Section 2 of this paper
presents the model, and Section 3 formulates the primary hypotheses. Section 4 describes
the sample selection procedure and the research design and Section 5 presents the results
of primary empirical analysis. Section 6 discusses additional analyses and Section 7
presents the results of sensitivity tests. Section 8 summarizes the study.

9

Chapter 2. A simple model
This section presents a simple model of a firm fully controlled by a single controlling
shareholder through a dual-class share structure. The controlling shareholder consumes
private control benefits to the detriment of minority shareholders because this dual-class
share structure allows the controlling shareholder to have disproportionately larger voting
rights relative to his cash flow rights, and accordingly, his expropriating behavior is
relatively immune to capital markets’ disciplining mechanisms (i.e., hostile takeovers and
proxy contests). The controlling shareholder operates the firm in greater secrecy to secure
his consumption of private control benefits.
The model has two dates, date 1 and 2. At date 1 the controlling shareholder raises
external capital, I, from the debt market to finance its investment project. The interest rate
is zero. At date 2, revenue is realized from the project. For simplicity, I assume constant
returns to scale for the production function—every dollar invested in the project
generates 1+R dollars. The controlling shareholder pays back the principal, I, to
debtholders.
I further assume that the controlling shareholder has cash flow rights, s, in the firm
and the controlling shareholder is the manager. La Porta et al. (1999, 2002) show that
controlling shareholders typically serve as managers in many cases. Even when the
controlling shareholder is not the manager, the controlling shareholder can set up other
mechanisms (e.g., intermediary firm), through which he can indirectly control the firm.
Shareholders do not receive all of the profits in proportion to their fractional
ownership. As a benefit of controlling the firm, the controlling shareholder can divert a
10

share f of the profits from the firm to himself, before he distributes the rest to minority
shareholders as liquidating dividends. I assume that the quality of accounting standards
varies across countries, and f is larger in countries with a lower quality of accounting
standards. This assumption is consistent with corporate finance and accounting literatures
that lower quality accounting information allows the controlling shareholder to operate
his firm in greater secrecy and to steal corporate resources from the firm without minority
shareholders having any knowledge (La Porta et al., 1999; Fan and Wong, 2002; Leuz,
Nanda, and Wysocki, 2003; Haw, Hu, Hwang, and Wu, 2004; Francis, Schipper and
Vincent, 2005). Following the crime and punishment model by Becker (1968), I assume
that the controlling shareholder is caught with probability P 0,1. P is a function of the
quality of accounting standards, q, and the share of the profits the controlling shareholder
diverted from the firm, f. Here, I assume that a

0, b

0, c

0 and
d

0. The assumption (a) denotes that the probability of being caught is higher
in a regime with high-quality accounting standards; (b) denotes that the marginal
probability of being caught is positive; (c) denotes that the marginal probability of being
caught increases in proportion to the amount being stolen; and (d) denotes that the
marginal probability of being caught is higher when the accounting standards are of a
high quality.
If the controlling shareholder is caught, he is required to return the diverted amount to
the firm. Additionally, he should pay to the authorities a penalty of c

RI, which is a
convex function of f. In this case, the entire profit, RI, is distributed as liquidating
dividends. However, if the controlling shareholder is not caught, he keeps the entire
11

diverted amount, and the fraction of the revenue not diverted, 1 RI, is distributed
as dividends. The controlling shareholder wants to maximize his payoff at date 2, which
is given by
max

P s RI c

RI 1 P s 1 RI RI
(a)
The controlling shareholder maximizes (a) by choosing f, such that 0 1. The
problem for obtaining optimal f* is solved by differentiating the equation (1) with respect
to f

s c

P 2c

s 1 1 P 1 s 0
(b)
Equation (b) is the first order condition for this problem. By differentiating the first-order
condition with respect to q, I obtain
d

dq

Numerator
Denominator

' 0 (
where
Numerator

c

1 s

2c 1 s
0 and
Denominator

s 1

2 2c s 1

2cP '
0.
12

Equation (c) generates the main implication of this model that in countries with high-
quality accounting standards, minority shareholders are less likely to be expropriated by
MCOs.
10

10
Refer to the appendix I with respect to the intermediary calculation between equations (2) and (3).
13

Chapter 3. Institutional background and hypothesis development
Chapter 3.1. Dual-class firms
A dual-class structure of ownership exists when a firm has at least two classes of
shares with different voting rights. In firms characterized by a dual-class structure, the
separation between control and cash flow rights makes MCOs relatively immune to
hostile takeovers or proxy contests, consequently locking corporate controls in a few
MCOs. Although it is not always harmful to lock firms’ controlling power in a few MCOs
who have private information on firms’ growth opportunities, finance research highlights
that the dual-class structure entrenches poorly-performing MCOs.
11
Further, these
entrenched MCOs are often observed to extract control benefits at the expense of
minority shareholders’ welfare (Jensen and Ruback, 1983; Shleifer and Vishny, 1997;
Gompers, Ishii, and Metrick, 2009). Because of this dysfunctional effect of the dual-class
structure, this structure is correlated with lower firm value around the world (Dyck and
Zingales, 2004).
Due to the additional voting power associated with the superior voting shares, the
price of the superior voting shares in the dual-class firm is in general higher than the price
of the inferior voting shares—the voting premium. The voting premium is set by generic
shareholders’ trading of the superior and inferior voting shares in public stock exchanges.
On a regular trading day the inferior and superior voting shares are traded among generic

11
Dyck and Zingales (2004) argue that the existence of private benefits of control is sometimes the most
efficient way for the MCOs to seize some value-creating opportunities. For example, the dual-class
structure allows MCOs, who possess private information on value-creating investment opportunities, to
pursue those opportunities without incurring extra costs in communicating their information to generic
shareholders.

14

shareholders, who will never have any opportunity to consume control benefits. Thus, the
voting premium represents the expected price generic shareholders will receive when a
control contest occurs.
12
Under some reasonable assumptions on the probability of the
control contest, the voting premium reflects the size of the private benefits of control
(Zingales, 1994 and 1995).
13
For this reason, the voting premium reflects a lower bound
of the size of the private benefits of control (Doidge, 2003). In this study, following prior
voting premium literature, the voting premium is defined as the market price of a voting
right, divided by the market price of a cash flow right (Zingale, 1995; Doidge, 2003):
S I
I S
P rv P
P P
VP
* - - = , (1)
Where:

P
S
= The market closing price of a firm’s superior voting shares;
P
I
= The market closing price of a firm’s inferior voting shares; and
rv = A ratio of the number of votes of an inferior voting share to that of a superior voting
share.

12
Zingales (1995) cites three cases where there were changes in the distribution of voting power. In the
three cases the premium of the superior voting shares dramatically surge around the respective events.
These cases are the unexpected death of the largest shareholder (William Crosby) at Resorts International,
the occurrence of a conflict with the Wang family at Wang Laboratories, and the largest shareholder’s
decision to exchange his holdings for some assets because of conflicts of view with the board of directors
for Moog Inc.

13
As noted in this section, the voting premium is determined not only by the size of the control benefits,
but also by the probability of a control contest. Hence, it is possible that the observed decrease of the voting
premium subsequent to mandatory IFRS adoption may not be driven by the reduction of the control
benefits, but by the increased probability of a control contest. To address this possibility, I repeat all the
following analyses after including three control variables for the probability of the control contest,
separately or in combination: the ownership concentration index (La Porta et al., 1998), the Shapley value
(Milnor and Shapley 1978), and firm characteristics—sales growth, leverage, ROE and industry dummies.
When I repeat all the analyses after including the preceding proxies for the probability of a control contest,
untabulated results show that the change of the voting premium for the mandatory adopters before and after
mandatory IFRS adoption is significantly negative (p-value < 10%). Hence, my primary conclusion is not
sensitive to controlling for the probability of a control contest.
15

Chapter 3.2. Legal protection of minority shareholders
against private control benefits
The voting premium literature shows that the size of the voting premium is
heterogeneous across countries, depending on the extent to which minority shareholders
are legally protected. In particular, the voting premium is significant in countries with
weaker legal investor protection mechanisms. In contrast, the voting premium is marginal
in countries where investors are legally well protected from controlling parties’
expropriating behavior (Modigiliani and Perotti, 1997; Nenova, 2002; Doidge, 2003).
For example, the voting premium is documented to correspond to more than 80 percent
of firm value in Italy where legal institutions are relatively weak, yet almost zero in
Finland and 10.5% in the U.S. where legal institutions are relatively strong (Zingales,
1994; Nenova, 2002). Similarly, Nenova (2002) further reports that the average value of
control benefits is at least 25.4% of firm value in countries with French legal origin, and
only 4.5% in countries with common law origin. Extending the preceding studies, Dodige
(2004) find that while the voting premium decreases when a firm cross-lists on the U.S.
stock exchanges, these voting premium effects are concentrated among firms that
domicile in countries with weaker investor protection. These results collectively suggest
that agency costs associated with the dual-class structure are significant in countries with
weak investor protection (e.g., Nenova, 2002; Dyck and Zingales, 2004).
14

14
In countries where shareholders are legally strongly protected and thus agency costs are low, the existence
of the private benefits of control may be beneficial to generic shareholders. “First of all, private benefits
might be the most efficient way for the company to capture some of the value created. Imagine, for instance,
that a corporate executive acquires valuable information about investment opportunities in other lines of
businesses, which the company cannot or does not want to pursue. The executive could sell this
information in the interest of shareholders. But the price she will be able to fetch is probably very low.
Thus, it might be efficient that the executive exploits this opportunity on her own. Second, even if the
16

Chapter 3.3. The impact of the IFRS mandate
on the voting premium of dual-class shares
Accounting theories suggest that the size of the private benefits of control is a
function of the quality of accounting information (see Fan and Wong, 2002; Leuz, Nanda,
and Wysocki, 2003; Haw, Hu, Hwang, and Wu, 2004; Francis, Schipper and Vincent, 2005).
In the context of dual-class firms, the dual-class structure is chosen by MCOs who want
to operate their firms in greater secrecy to extract value from their firms. The divergence
of control and cash flow rights associated with the dual-class structure allows MCOs
(who have a small fraction of cash flow rights, yet a majority of voting rights) to lock in
their controlling power, thereby providing MCOs an opportunity and incentive to
expropriate minority shareholders’ wealth. To secure their consumption of such control
benefits, MCOs often generate financial statements with more estimation error and noise.
This obfuscation of accounting information, in turn, limits minority shareholders’ ability
to identify and discipline MCOs’ diversionary behavior, consequently allowing MCOs to
secure their consumption of private control benefits to the detriment of minority
shareholders. Recent accounting studies, for example, provide supporting evidence that
the divergence of control and cash flow rights leads controlling parties to generate less
informative accounting earnings (Francis, Schipper and Vincent, 2005), and to engage in
a higher degree of earnings management (Fan and Wong, 2002; Leuz et al., 2003; Haw,
Hu, Hwang, and Wu, 2004; Pae, Thornton and Welker, 2006).

extraction of private benefits generate some inefficiency, their existence might be socially beneficial,
because their presence makes value-enhancing takeovers possible (Grossman and Hart (1980)).” (Dyck and
Zingales, 2004, p. 8)
17

Building on this accounting literature, I investigate whether the IFRS mandate
will mitigate the adverse effects of this opaque accounting information at the dual-class
firms. A large body of IFRS literature suggests that mandatory adoption of IFRS, which
is characterized as having a higher degree of transparency and comparability of financial
information and a restricted set of accounting measurement choices, reduces information
asymmetry between firms and investors (see Cox (2009)). For example, Armstrong et al.
(2009) report, on average, positive abnormal returns around the announcements of the
IFRS mandate in the European stock markets. These positive returns are concentrated
among firms that have a higher degree of information asymmetry. Similarly, Daske et al.
(2008) and Li (2010) demonstrate that the high-quality accounting after the mandatory
switch to a uniform set of accounting standards allows shareholders to distinguish
between poorly-performing firms and firms that are performing well, consequently
decreasing cost of capital. Based on the preceding studies, I hypothesize that the
mandatory adoption of IFRS is correlated with the decrease of the size of private control
benefits by mitigating information asymmetry between MCOs and minority
shareholders.
15

Consistent with the preceding IFRS literature, the IFRS mandate seeks to improve
financial reporting, and thus, reduce the expropriating risk minority shareholders face

15
Consistently, Tweedie, the chairman of the International Accounting Standards Board (IASB), states that
the objective of IASB is to develop the “highest common denominator” of financial reporting standards to
generate financial statements that provide confidence to generic shareholders that financial statements
provide a complete and accurate picture of a firm’s profitability and position (Tweedie, 2002). To achieve
this objective, the IASB has worked toward accounting standards that require a higher degree of disclosure
and that produce more comparable accounting information across different jurisdictions not by eliminating,
but by reducing the differences in each jurisdiction’s generally accepted accounting standards (FASB,
1996). Consequently, mandatory IFRS adoption and the improvement of accounting information quality
brought about by the adoption are predicted to decrease information asymmetry between MCOs and
minority shareholders on MCOs’ diversionary behavior.
18

when controlling parties extract value from firms at the expense of minority shareholders.
For example, unlike the domestic GAAPs of most countries, IFRS requires firms to
disclose the name of the ultimate controlling party [IAS 24.17], denying the ultimate
controlling party the secrecy of their identity and hiding their expropriating behavior
from minority shareholders. In the event of related party transactions, controlling parties
are required to disclose the nature of the related party relationship as well as provide
sufficient information about the transactions and outstanding balances for shareholders to
assess the potential effect of the relationship on the firm’s financial statements [IAS
24.18-19]. This disclosure requirement limits the ability of controlling parties to transfer
corporate asset to themselves at a non-market price. IFRS also mandates the disclosure of
key managerial compensation, both in total and for each of the following categories:
short-term employee benefits, post-employment benefits, other long-term benefits,
termination benefits, and share-based benefits [IAS 24.17]. This disclosure requirement
restricts controlling parties’ ability to excessively distribute compensation to themselves.
The share-based payment rules reduce agency costs related to CEO stock option [IFRS 2]
and the business combination rules restricts controlling parties’ practical ability to sub-
optimally expand their firms and increase their own private control benefits [IFRS 3].
Overall, these extensive disclosures will decrease information asymmetry between
minority shareholders and controlling parties with respect to controlling parties’
diversionary behavior.
The decreased information asymmetry resulting from mandatory IFRS adoption
will have the effects of deterring the ability of MCOs to extract value from firms to the
19

detriment of minority shareholders’ wealth. If MCOs know ex ante that their
expropriation will be more visible subsequent to mandatory IFRS adoption, then they are
less likely to take actions that divert firms’ assets for their private benefits. This is
because they are aware that the increased visibility of their behavior will increase the
probability of incurring legal penalties and bearing reputational costs.
16
The increased
visibility of the MCOs’ behavior also enables minorities to ex post stem MCOs’
diversionary behavior by starting class action lawsuits and using other legal mechanisms
to discipline MCOs’ dysfunctional behavior. Based on the preceding arguments, I expect
that MCOs will decrease their diversionary behavior subsequent to the mandatory
adoption of IFRS. After observing this, rational minority shareholders will recognize that
the magnitude of control benefits is reduced, and accordingly decrease the voting
premium. Based on the preceding arguments, the primary hypothesis in this study (in
alternative form) is as follows:
H1: After the mandatory introduction of IFRS reporting, the voting premium for the
mandatory adopters will, on average, decrease.

16
Supporting this claim, Hope et al. (2006) argue that after the mandatory switch, MCOs are under the
tighter monitoring and discipline of international financial intermediaries, such as brokerage analysts,
auditors and credit rating agencies.
20

Chapter 4. Sample and descriptive statistics
Chapter 4.1. Sample description
I start the sample collection procedure by identifying all firms in countries that
require IFRS reporting for all domestic firms during the sample period and have at least
two classes of shares in accordance with the country lists of DATASTREAM. After
identifying these firms, I follow the previous voting premium studies and impose the
following sample selection criteria (see Nenova (2002) and Doidge (2003)): (1) firms
must have at least two classes of shares with distinct voting rights per cash-flow rights,
(2) both classes of shares must be publicly traded on the domestic stock exchange, (3) the
inferior voting shares cannot be convertible into the superior voting shares (however, the
opposite condition is allowed),
17
(4) neither share class can receive a fixed dividend, and
(5) neither share class can be redeemable or callable at the option of the firm at a
prearranged price.
The sample period is from 2002 through 2007. The sample period starts at 2002
in such a way that I attain a well balanced panel dataset around the mandatory
introduction of IFRS reporting in 2005. The sample period prior to the mandatory switch
allows me to address the anticipation effects of the forthcoming accounting standard
change.
18
2008 is excluded from the sample period to exclude the possibility that the

17
This criterion is imposed in order to eliminate those firms for which the two classes are interconvertible,
and thus for which the labels superior and inferior voting share may represent a “distinction without a
difference” (DeAngelo and DeAngelo, 1985).

18
These anticipation effects of the mandatory switch are documented by Pae, Thornton and Welker (2006).
Based on their analysis of the European Union, they find that Tobin’s Q ratio increases among countries
where the largest shareholder has excessive voting rights relative to cash flow rights. Their study is based
on two fiscal years (e.g., 1999 and 2003), when IFRS is not yet mandated, while my study covers pre- and
post-IFRS periods (e.g., from 2002 through 2008).
21

financial crisis in 2008 may contaminate the empirical results; prior finance studies show
that MCOs’ diversionary behavior significantly exacerbates during the financial crises
(Johnson et al., 2000; Lemon and Lins, 2001; Mitton, 2002). Finally, I require data
availability in both the pre-mandatory (2002-2004) and the post-mandatory (2005-2007)
periods.
My data collection procedures follow the guidelines established in Nenova (2002)
and Doidge (2003). First, for each share class, I collect from DATASTREAM Friday-to-
Friday weekly market closing data for the following variables used in the multivariate
regression analyses: closing stock price, market value of all equity outstanding, weekly
return, dividends paid during the week, number of shares outstanding and turnover. If the
values of the turnover variable are missing, I follow Doidge (2003) and collect the data
from Bloomberg. Second, for each fiscal year, I include a firm in the sample only if it has
at least 20 weekly observations. Third, I require data on the number of voting rights
attached to the superior- and inferior-voting shares for each firm. I hand-collect this data
from extensive documents including DATASTREAM Manuals, Moody’s International
Manuals, filings with the national stock exchange, firms’ annual reports and the sample
firm lists compiled by Doidge (2003).
19
Finally, I winsorize every variable at one
percentile on both sides of its distribution at each fiscal year to prevent the undue

19
In this project I received a significant support from Craig Doidge with respect to data collection and
project documents. If the data is still not clear or not available, I requested the data for each firm through
faxes, emails and phone calls. This data collection procedure might introduce measurement errors in the
empirical analyses.

22

influence of extreme values. After imposing these criteria, the final sample is comprised
of 868 firm-year observations from 151 firms across 13 countries.
20

Chapter 4.2. Descriptive statistics
Table 1, Panel A, presents a break-down of the number and percentage of
observations of dual-class firms in the 13 sample countries that require IFRS reporting
under the 2005 IFRS mandate for all domestic firms.
The sample firms are classified into two distinct categories: firms that (1)
domicile in countries where IFRS reporting is compulsory to all the domestic firms in
2005 and (2) first apply IFRS before two fiscal years of 2005 are classified as mandatory
adopters (treatment sample). Firms that voluntarily adopt IFRS two fiscal years before
2005 are defined as early voluntary adopters (control sample). The early voluntary
adopters are used as the control sample against which the change of the voting premium
for the treatment sample is benchmarked. The categorization of IFRS adoption type is
based on the “Accounting Standard” variable in WORLDSCOPE and is confirmed based
on hand-collected financial statements and Orbis, the data base provided by Bureau van
Dijk Electronic Publishing. Panel A illustrates that, similar to Daske et al. (2008), the
number of voluntary adopters is small, consisting of 10.3% of the sample firm-year
observations, and the adoption rates vary considerably across countries.
Also, the distribution of the dual-class firms in my sample is similar to the
findings by Doidge (2003); Germany and Sweden have the largest number of dual-class

20
My sample countries comprise the 20 countries with dual-class firms that satisfy my sample selection
criteria. Thirteen of these 20 sample countries introduce IFRS reporting for all domestic firms in 2005, and
seven countries do not introduce IFRS reporting during the sample period. These seven non-IFRS adoption
countries are used as a control group against whose voting premiums the voting premiums for mandatory
adopters in the 13 IFRS adoption countries are benchmarked.
23

firms with 32, while Austria and Venezuela have the fewest with one. One notable
divergence from the firm distribution recorded by Doidge (2003) is the number of dual-
class firms in Italy. While the number of Italian dual-class firms recorded by Doidge
(2003) is 77, the third largest in his sample distribution, the number decreases to only 30
in my sample. This is primarily caused by the recent movement of the Italian government
to unify the dual-class shares. As a consequence of this movement, a large portion of
sample firms in Italy allow the interconvertibility between inferior voting shares and
superior voting shares, thus not satisfying my sampling criteria (Aganin and Volpin,
2003).

24

Table 1. Sample Composition by Country
The full sample comprises 868 firm-year observations from 13 counties over the period of 2002 and 2007. The sample
countries consist of mandatory adopters (treatment group) and voluntary adopters (control group) with dual class stocks, which
have sufficient financial data from the DATASTREAM database. Specifically, I start to construct the sample by identifying all
firms with dual-class shares during the sample period from the country lists in the DATASTREAM database. Following the
prior voting premium literature, I impose the following sample criteria on the identified firms: (1) firms have at least two
classes of shares with different voting rights, (2) both share classes are publicly traded and listed on the domestic exchange, (3)
the inferior voting share is not convertible into the superior voting share, and (4) neither share class receives a fixed dividend
(Nenova, 2003; Doidge, 2004). For each share class, I download from the DATASTREAM database Friday day-to-Friday day
weekly market data for prices, returns, market value, dividend yields, number of shares outstanding, and turnover. In instances
in which the values of turnover variable are missing, I collect them from Bloomberg. I include a firm into the sample for a
given fiscal year if it has at least 20 weekly observations. Finally, I extensively hand collect data on the different voting rights
associated with the superior- and inferior-voting shares (data source: documentation supplied by

25
Table 1. Continued.
Datastream, Moody’s International Manuals, and firms’ annual reports among others). I exclude firm-year observations with
missing values in any variable. Consequently, I obtain 868 firm-year observations from 13 countries as the final sample. The
sample firms are classified into two groups: (1) mandatory adopters that are a subgroup of firms mandatorily adopting IFRS
from 2003 onward (a.k.a., two fiscal years before IFRS reporting is required for domestic firms) (treatment group) and (2)
the early voluntary adopters that are a subgroup of firms voluntarily adopting IFRS before 2003 (control group). Panel A
reports the number of firms, the number and percentage of firm-year observations for the sample countries. IFRS indicates
financial reports following International Financial Reporting Standards, based on the “accounting standards followed”
variable in WORLDSCOPE (WC07536) and is confirmed based on hand-collected financial statements as well as Orbis that
is the data base provided by Bureau van Dijk Electronic Publishing. Panel B reports descriptive statistics for variables used
in the main regression analysis.

26
Table 1. Continued.

Pre IFRS Post IFRS
Voluntary Mandatory

Voluntary Mandatory
Country
Firms
Firm-
Years
Firm-
Years %
Firm-
Years %
Firm-
Years %
Firm-
Years %

Australia
3 17 9 0.529 8 0.471
Austria
1 6 3 0.500 3 0.500
Denmark
8 44 3 0.068 21 0.477 1 0.023 19 0.432
Finland
15 84 3 0.036 42 0.500 3 0.036 36 0.429
France
2 12 6 0.500 6 0.500
Germany
32 186 39 0.210 57 0.306 35 0.188 55 0.296
Italy
30 172 88 0.512 84 0.488
Norway
4 23 12 0.522 11 0.478
South Africa
11 61 33 0.541 28 0.459
Sweden
32 191 96 0.503 95 0.497
Switzerland
4 22 3 0.136 9 0.409 3 0.136 7 0.318
U.K.
8 44 24 0.545 20 0.455
Venezuela
1 6 3 0.500 3 0.500

Total
151 868 48 0.055 403 0.464 42 0.048 375 0.432

27

Table 2 presents the firm-level descriptive statistics and correlation coefficients
on the variables used in the multivariate regression analyses. Panel A presents differences
in mean and median of variables between voluntary and mandatory adopters in the pre-
and post- IFRS periods. Both the mean and median of voting premium for mandatory
adopters significantly decreases subsequent to the IFRS mandate (p-value < 0.10). In
contrast, the mean and median of voting premium for voluntary adopters do not change.
These univariate comparisons support my main hypothesis that the IFRS mandate is
correlated with the decrease of mandatory adopters’ voting premium.
Table 2, Panel B presents Pearson correlation coefficients among the variables.
The panel indicates that voting premium is significantly negatively correlated with the
excess dividend payment to an inferior- over a superior-voting share, while being
significantly positively correlated with voting power. These findings are consistent with
the voting premium literature; a higher dividend payment for the inferior-voting class
relative to the superior-voting class lowers the price differencial between the two classes.
When voting power is more concentrated, the voting premium is higher.

28

Table 2. Firm-Level Descriptive Statistics for Variables Used in Regression Analyses

This table reports firm-level descriptive statistics for variables used in regression analyses. Panel A presents differences in
mean and median of variables between the early voluntary adopters and the mandatory adopters in the pre- and post- IFRS
periods. Panel B reports Pearson correlation coefficients among the variables.
Variable Definitions:
Voting premium = (PS – PI)/(PI - rv*PS), where PS and PI are the weekly market prices of the superior and inferior voting
shares. rv is the relative number of votes attached to the inferior voting shares, defined as a ratio between the number of
voting rights attached to an inferior voting share and the number of voting rights attached to a superior voting share
(between 0 and 1).
Size = The natural log of a firm’s total market capitalization, where total market capitalization is the sum of the market
capitalization of the superior voting and inferior voting share classes (in millions of US$).
Div = The ratio of the difference between the dividends of the inferior voting shares and the superior voting shares, over the
dividend of the inferior voting shares for a given fiscal year.

29
Table 2. Continued.

Turnover = The log of the ratio of the total trading volume of the inferior voting shares over that of the superior voting shares.
Voting power= The portion of total votes attached to the superior voting share class relative to the portion of total cash flow
rights attached to the superior voting share class, or (NSS + NSI)/(NSS + rv*NSI).

30
Table 2. Continued.

Panel A: Descriptive Statistics for Variables Used in Regression Analyses

IFRS
Adoption
Type
Means
Medians

Pre IFRS
Post
IFRS
Diff.
T-test Pre IFRS
Post IFRS Diff.
Wilcoxon
-test

Number of Observations Voluntary 48 42 48 42
Mandatory 403 375 403 375

Voting Premium Voluntary 0.109 0.056 -0.053 0.067 0.02 -0.047
Mandatory 0.219 0.139 -0.080 ** 0.054 0.04 -0.014 **

Size Voluntary 7.712 8.699 0.987 *** 7.57 8.904 1.334 ***
Mandatory 5.736 6.509 0.773 *** 5.863 6.911 1.048 ***

Div Voluntary 0.047 0.099 0.052 0.046 0.036 -0.010
Mandatory -0.478 -0.999 -0.521 *** 0.000 0.000 0.000

Turnover Voluntary 0.692 0.516 -0.176 0.856 0.876 0.020
Mandatory 0.359 0.589 0.230 0.449 0.588 0.139

Voting power Voluntary 1.902 1.777 -0.125 1.856 1.699 -0.157
Mandatory 2.833 2.833 0.000 1.737 1.691 -0.046

31

Table 2. Continued.

Panel B: Pearson Correlation Coefficients

Voting Premium Size Div Voting power

Size -0.053
(0.121)
Div -0.173 0.408
(0.000) (0.000)
Voting power 0.067 -0.092 -0.088
(0.049) (0.007) (0.009)
Turnover -0.035 0.063 -0.006 -0.466
(0.309) (0.063) (0.863) (0.000)

32

Chapter 5. Empirical analysis of primary hypothesis
Chapter 5.1. Research design
In this section the change in the voting premium for the mandatory adopters
subsequent to mandatory IFRS adoption is formally compared to the change in the voting
premium for the benchmark group using multivariate regression analysis. This is done by
using panel data models that control for both firm and country characteristics, while
allowing the impact of the mandatory switch to differ across firm types (i.e., mandatory
adopter and early voluntary adopter).
In the multivariate analysis the dependent variable is the voting premium as
defined in equation (1). I estimate the voting premium in equation (1) based on weekly
closing price data for each firm and take the average value each year during the sample
period (Nenova, 2002; Doidge, 2003). The independent variables of most interest in the
multivariate regression analysis are a binary indicator variable, Mandatory that takes on a
value of one if fiscal year ends on or after the IFRS adoption date (2005) for firms that
that first adopt IFRS in 2003 and later (treatment sample).
21
This variable captures the
average voting premium effects subsequent to the IFRS adoption for firms that are
mandated to adopt IFRS. I create two binary indicator variables, early voluntary and post,
for firms that report under IFRS ahead of the mandatory introduction of IFRS (control
sample). Early voluntary takes on a value of one if firms voluntarily adopt IFRS before

21
Note that voluntary adopters in 2003 or 2004 are reclassified as mandatory adopters. I classified firms
based on whether firms adopt IFRS two fiscal years before the IFRS mandate because firms may adopt
IFRS in 2003 or 2004 due to their anticipation of the IFRS mandate (2005). Hence, I prevent the possibility
that classifying these firms as voluntary adopters may cause the difference-without-distinction problems.
However, when I classified firms into voluntary and mandatory adopters based on the date of the IFRS
mandate (2005), the empirical results were qualitatively the same.

33

2003 and Post takes on a value of one if a firm’s fiscal year falls into the post-IFRS
mandate period (e.g., 2005 and afterwards). The interaction between early voluntary and
post captures the incremental voting premium effects following the IFRS mandate for
firms that voluntarily adopt IFRS.
Following previous research, the regression model includes the following set of
control variables that are expected to influence the voting premium (Zingales, 1994,
1995; Smith and Amoaku-Adu, 1995): (1) Voting power to capture the degree of
separation between cash flow rights and voting rights; (2) Div to capture dividend
differences between the share classes; (3) Turnover to control for differences in the
liquidity of the share classes; and (4) Size to capture the costs attached to holding a large
block of shares. Similar to the voting premium, the control variables are calculated based
on weekly data and averaged annually. The formal regression model is as follows:
Voting premium
it
=
β
0
+ β
1
(Early Voluntary)
it
+ β
2
(Early Voluntary * Post)
it
+ β
3
(Mandatory)
it
+ β
4
(Voting
power)
it
+ β
5
(Div dummy)
it
+ β
6
(Turnover)
it
+ β
7
(Size)
it

+ ∑ β
n
(DCountry)
i
+ ε
it
(2)
Where:

Voting premium =(P
S
- P
I
)/(P
I
- rv*P
S
), as defined in equation (1);
Early Voluntary = A dummy variable, which takes on a value of one if a firm voluntarily
adopts IFRS before 2003;
Post = A dummy variable, which takes on a value of one for firm-years ending in or after
2005;
Mandatory = A dummy variable, which takes on a value of one if a firm first adopts
IFRS in 2003 and later and applies only to firms that domicile in the countries where
IFRS is mandatorily adopted in 2005;
Voting power= The portion of total votes attached to the superior voting share class
relative to the portion of total cash flow rights attached to the superior voting share
class, or (NS
S
+ NS
I
)/(NS
S
+ rv*NS
I
);
Div = The excess dividend payment to an inferior- over a superior-voting share, divided
by the total dividend to a inferior -voting share;

34

Turnover = The log of the ratio of the total trading volume of the inferior voting shares
relative to the superior voting shares;
Size = The log of total market capitalization, where total market capitalization is the sum
of the market capitalization of the superior voting and inferior voting share classes (in
millions of USD);
DCountry = A set of dummy variables, indicating each country.

In this multivariate analysis the standard errors are heteroskedasticity robust, and
clustered by firm to account for possible serial correlation between the residuals of the
same firm.
22

The country fixed effects (DCountry) are included in equation (2) to control for
unobserved country heterogeneity in the voting premium, thus, eliminating potentially
large sources of bias in estimating equation (2). However, a problem of the fixed effects
model is that the fixed effects model completely ignores the between-country variation in
the voting premium and considers only the within-country variations. Discarding the
between-country variation can render standard errors to be higher than those generated by
models that employ both within- and between-country variation. To address this problem
of the fixed effects model, I additionally employ the random effects model, which
considers both within- and between-country variation of the voting premium. The random
effects model is an expansion of the fixed effects model that includes a country-specific
random component, ν
i
:
Voting premium
it
=
β
0
+ β
1
(Early Voluntary)
it
+ β
2
(Early Voluntary
it
* Post
it
) +β
3
(Mandatory)
it
+ β
4
(Voting
power)
it
+ β
5
(Div)
it
+ β
6
(Turnover)
it
+ β
7
(Size)
it

+ ∑ β
n
(DCountry)
i
+ ν
i
+ ε
it
(3)

22
These standard errors are comparable to GMM standard errors, where the serial correlation of the
standard errors is presumed to be of order n, where n is the number of countries (Zingales, 1995).

35

Here, ν
i
captures the between-sample variation of the voting premium. I further find that
the null hypothesis of the random effects model (σ
2
(ν
i
) = 0) is rejected by the Breusch
and Pagan Lagrange multiplier test (p-value <.001), suggesting that the random effects
model has more explanatory power relative to the fixed effects model.
23
The country
fixed or random effects are included throughout all the multivariate regression analyses to
control for country-specific shocks to the voting premium.
Chapter 5.2. Regression results
Table 3 reports the results of the estimate of equations (2) and (3). Model 1
(Model 2) of Table 3 presents the results of the effect of the IFRS mandate on the change
of firms’ voting premium by using the voluntary adopters as the benchmark group during
the full sample period (during the sample period excluding transition periods). I exclude
the transition period in model 2 to address the possibility that the regression results may
be contaminated by confounding factors during the transition period: financial
intermediaries and investors may have temporary difficulties in interpreting firms’
financial statements under the new accounting standards (see, for example, Daske et al.,
2008). Estimates of eight different specifications of equations (2) and (3) are presented in
the table 3. The base-line model reports results only with the IFRS indicator variables,
while expanded model reports results with the IFRS indicator variables after including
the firm-specific variables that are reported to influence the voting premium by prior
corporate finance literature.
Table 3 demonstrates that the voting premium for the mandatory adopters
significantly decreases after the IFRS mandatory adoption across different model

23
If σ
2
(ν
i
) = 0, then the random effects model is equivalent to the fixed effects model.

36

specifications. In regression with the full sample period, the coefficient of Mandatory, β
3
,
becomes significantly negative (p-value <.100 for fixed effects estimates; p-value <.100
for random effects estimates). The results of the mandatory adopters in model 2 (after
excluding the transition period) are identical to those in model 1 except the base-line
model that includes country-specific random-effects. However, the coefficient of
Mandatory, β
3
, is significant in the expanded model with country-specific random-effects
where firm-specific control variables are included (p-value < .001), supporting my
hypothesis that after the mandatory introduction of IFRS reporting, the voting premium
for the mandatory adopters will, on average, decrease.
Furthermore, the coefficients of the interaction between Early Voluntary and Post
are insignificant at a conventional statistical significance, indicating that the IFRS
mandate is not correlated with any change in the size of the voting premium for those
already applying IFRS to their financial statements.

37

Figure 1. The Voting Premium Effects of the IFRS Mandate

`````````

38

Table 3.
Firm-Year Regression Analysis of the Voting Premium Effects around the IFRS Mandate
a

This table reports the results of the effect of the IFRS mandate on the change of mandatory adopters’ voting premium by using
the early voluntary adopters as the control group. In particular, Panel B reports the multivariate regression coefficients, two-
sided p-values, which is calculated based on firm-cluster adjusted t-statistics of estimating equations (2) and (3). Model 1
(Model 2) of Table 3 reports the results of the effect of the IFRS mandate on the change of firms’ voting premium during the
full sample period (during the sample period excluding transition periods).

Variable Definitions:
See Table 1 for the definition of other variables.

39
Table 3. Continued.

Model 1:
Full Sample Period
(2003-2008)

Model 2:
After Excluding Transition Periods
(2003-2004,2006-2007)
Fixed Effect Estimates Random Effect Estimates Fixed Effect Estimates Random Effect Estimates
Variable
Baseline
Model
Expanded
Model
Baseline
Model
Expanded
Model
Baseline
Model
Expanded
Model
Baseline
Model
Expanded
Model

Early Voluntary -0.339 -0.358 -0.324 -0.344 -0.283 -0.312 -0.266 -0.297
(0.002) (0.001) (0.002) (0.002) (0.012) (0.007) (0.018) (0.01)
Early Voluntary*Post -0.072 -0.085 -0.07 -0.083 -0.082 -0.113 -0.078 -0.111
(0.593) (0.520) (0.604) (0.528) (0.570) (0.421) (0.590) (0.433)
Mandatory -0.077 -0.132 -0.078 -0.133 -0.079 -0.156 -0.08 -0.157
(0.093) (0.004) (0.092) (0.004) (0.096) (0.002) (0.103) (0.002)
Size -0.024 -0.024 -0.029 -0.029
(0.029) (0.027) (0.016) (0.014)
Div -0.076 -0.077 -0.069 -0.069
(<.001) (0.000) (<.001) (0.000)
Voting power 0.017 0.017 0.014 0.013
(0.041) (0.047) (0.129) (0.14)
Turnover -0.039 -0.038 -0.036 -0.035
(0.000) (0.000) (0.001) (0.001)

Country dummies Yes Yes No No Yes Yes No No

Obs. 868 868 868 868 566 566 566 566
Adj.R
2

0.11 0.16 0.11 0.16
0.12 0.14
0.12 0.14

a
Two-tailed p-values in parentheses, based on t-statistics with standard errors clustered by firm.

40

Chapter 6. Additional analysis
Chapter 6.1. Improvement in
country-level accounting information quality as mechanisms
behind the voting premium effects of the IFRS mandate
In this section I explore whether the IFRS mandate reduces the voting premium
through the following three mechanisms: (1) the increased financial disclosure, (2) the
enhanced comparability of reported information across jurisdictions that results from the
adoption of a single set of accounting standards (IFRS), and (3) the restricted set of
measurement methods controlling parties can choose relative to the local GAAPs.
As already noted, IFRS is characterized as (1) requiring a high degree of
financial disclosure relative to most local GAAPs, (2) having the ability to improve the
comparability of firms’ financial information across jurisdictions (given that IFRS is the
most commonly used GAAP world-wide) and (3) restricting controlling parties’ choices
of accounting alternatives (see Cox, 2009; Ashbaugh and Pincus, 2001; Daske et al.,
2008; Li, 2010). More specifically, the increased disclosure of financial information
under IFRS reduces information asymmetry between MCOs and minority shareholders,
thereby limiting MCOs’ capacity to operate firms in secrecy and conceal the consumption
of the private benefits of control. Similarly, the enhanced comparability of accounting
information under IFRS facilitates the ability of international financial intermediaries and
minority shareholders to identify MCOs’ diversionary behavior by reducing the costs of
comparing firms’ financial statements across different jurisdictions (Armstrong et al.,

41

2009).
24
Hence, the application of IFRS reporting enables minority shareholders to
identify firms’ atypical or deviant behavior compared to that of benchmarked firms in
other jurisdictions. Finally, IFRS generally limits MCOs’ choices of accounting
measurement methods (Ashbaugh and Pincus, 2001). With fewer measurement methods
for MCOs to choose, there are fewer ways for MCOs to disguise their diversionary
behavior. As a result, the probability that their diversionary behavior is detected and
penalized by minority shareholders will increase, which in turn renders their stealing to
be more costly.
To test the above conjecture, I employ three proxies for the increased
transparency and comparability of accounting information, and the restricted set of
accounting measurement methods in the IFRS reporting regime relative to the local
GAAP reporting regime. As for the proxy for the increased transparency of firms’
financial information, I employ the number of rules on disclosures that are required under
the IFRS reporting regime, but not under the local GAAP reporting regime.
25
If IFRS
requires relatively more additional disclosures in comparison with the local GAAP
reporting regime,

I expect that the quality of accounting information in those countries

24
Several studies demonstrate that a higher degree of comparability of financial information is associated
with the decrease of information processing costs, consequently benefiting the capital markets by
increasing U.S. mutual fund ownership among voluntary adopters and among mandatory adopters with
more transparent informational environments in the European Union countries, improving analysts’
forecast accuracy and reducing forecasts dispersion (DeFond et al., 2008; Franco, Kothari and Verdi, 2008;
Bradshaw, Miller and Serafeim, 2009).
25
The preceding proxies for the increased transparency and comparability of financial information are
collected from a survey conducted by the seven largest accounting firms in the U.S. The survey by these
accounting firms is on the differences between local GAAPs and International Accounting Standards (IAS),
which is largely similar to IFRS (GAAP, 2001).

42

subsequent to the mandatory switch will increase to a more significant extent.
26
Based on
this expectation, I predict that the reduction in the voting premium resulting from the
mandatory adoption of IFRS will be larger in countries for which IFRS requires more
additional disclosures in comparison with local GAAP.

Figure 2. The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of Disclosure

As a proxy for the increased comparability of accounting information, I employ
the number of inconsistencies between local GAAP and IFRS rules.
27
Firms domiciling
in countries with a higher number of inconsistencies between local GAAP and IFRS rules
are likely to benefit more significantly from the increased comparability of their financial
statements. As a proxy for the more restricted accounting, I employ the differences in

26
For example, while IFRS requires the disclosure of the fair values of investment properties in the firms’
annual report (IAS 40.69), the local GAAPs in all my sample countries, except those of South Africa and
the U.K., do not do so. Likewise, while IFRS mandates firms to disclose discontinued operations in a
detailed manner in the firms’ annual report (IAS 35), the GAAPs in eight of my 13 sample countries do not
mandate this or mandate very limited disclosure.

27
The preceding proxies for the increased transparency and comparability of financial information are
collected from a survey conducted by the seven largest accounting firms on the differences between local
GAAPs and IAS, which is largely similar to IFRS (GAAP, 2001).

43

measurement methods of IFRS versus the domestic GAAPs. IFRS are considered to be
more restrictive than domestic GAAPs when IFRS has fewer acceptable accounting
measurement methods for an economic event relative to the methods acceptable under
domestic GAAP reporting regimes. If firms domicile in countries with more acceptable
accounting methods under domestic GAAP reporting regimes, these firms are likely to
benefit more significantly from the restricted set of measurement methods under the IFRS
reporting regime. Based on this expectation, I predict that the reduction in the voting
premium resulting from the mandatory adoption of IFRS will be larger in countries where
there are a higher number of inconsistencies between local GAAP and IFRS rules and
where there are fewer acceptable accounting methods for an economic event under the
IFRS reporting regime relative to the domestic GAAP reporting regimes.

Figure 3. The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of Differences
between Local GAAP and IFRS

44

Figure 4. The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of Restrictiveness of the Local GAAPs

The preceding three proxies enable me to investigate whether the increased
transparency and comparability of accounting information, as well as the restricted set of
accounting measurement methods are indeed the underlying mechanisms that help to
explain the cross-sectional and time-series variability of the voting premium in the pre-
and post-IFRS reporting regime that I observe. The voting premium literature documents
that the voting premium effects of the IFRS mandate differ between countries with weak
legal investor protection mechanisms and countries with strong legal investor protection
mechanisms. Thus, I follow Daske et al. (2008), and orthogonalize the proxies for
transparency, comparability, and the restrictiveness of accounting measurement methods
to the strength of countries’ legal institutions. More specifically, I take residuals from the
regression of these proxies onto countries’ legal origin and a natural logarithm of GDP. I
then transform these residuals into binary indicator variables, depending on the sample

45

country median of a given variable. I expand the primary multivariate regression model
by including these binary indicator variables in equations (2) and (3). Specifically, the
expanded multivariate regression model is stated as follows:
Voting premium
it
=
β
0
+ β
1
(Early Voluntary)
it
+ β
2
(Early Voluntary
it
* Post
it
) + β
3
(Mandatory)
it
+ (Accting
Info)*[γ
0
+ γ
1
(Early Voluntary)
it
+ γ
2
(Early Voluntary
it
* Post
it
)

+ γ
3
(Mandatory)
it
] +
β
4
(Voting power)
it
+β
5
(Div)
it
+ β
6
(Turnover)
it
+ β
7
(Size)
it
+ ∑ β
n
(DCountry) + ε
it

(5)

Where:

Accting Infor = A binary indicator variable, which takes on a value of one if a given
measure of accounting information quality is greater than or equal to the median of
that measure for the sample countries. The measures of accounting information quality
are based on three attributes of accounting information–(1) the number of rules on
disclosures that are absent under the local GAAP regime but are required under the
IFRS reporting regime (Nobes, 2002); (2) the number of the inconsistencies between
local GAAP and IFRS rules (Nobes, 2002); and (3) the number of how IFRS restricts
the accounting measurement methods relative to the measurement methods available
under domestic GAAPs (Ashbaugh and Pincus, 2002).

Table 4, Panel A reports descriptive statistics for the indices of accounting
information quality at the country level. The descriptive statistics indicate that these
indices significantly vary across countries. The indices of additional disclosures required
subsequent to mandatory IFRS adoption are eight for Austria and Finland, seven for
Germany and Switzerland, and zero for the U.K. Similarly, the number of inconsistencies
between local GAAP and IFRS rules is, for instance, 20 for Austria and Germany, while
it is only two for South Africa. The number of measurement methods that are restricted
under IFRS, but not under local GAAPs is 17 for Switzerland, and zero for South Africa.

46

Table 4.
Additional Analysis on Institutional Characteristics of IFRS Adoption Countries

This table reports the results of examining whether the IFRS mandate reduces the voting premium through the following
mechanisms: (1) the increased financial disclosure, (2) the enhanced comparability of reported information across jurisdictions
that results from the adoption of a single set of accounting standards (IFRS), and (3) the restricted set of accounting
measurement methods controlling parties can choose to obfuscate their diversionary behavior (see Cox, 2009; Ashbaugh and
Pincus, 2001). Panel A reports descriptive statistics for the indices of accounting information quality at the country level. Panel
B reports the multivariate regression coefficients, two-sided p-values, which is calculated based on firm-cluster adjusted t-
statistics of estimating equation (4) and (5).

Variable Definitions:
Additional disclosure = Summary score of how many disclosures are additionally required by IFRS compared to domestic
GAAP from GAAP 2001 (Survey of National Accounting Rules Benchmarked against International Accounting Standards).
Higher values represent that more additional disclosures are required by IFRS relative to local GAAP.

47
Table. 4. Continued.
Difference between IFRS and local GAAP = Summary score of how domestic GAAP differs from IFAS on some 80
accounting dimensions from GAAP 2001 (Survey of National Accounting Rules Benchmarked against International
Accounting Standards). Higher values represent more discrepancies between local GAAP and IFRS.
Measurement method restrictions= Summary score of how IFRS restricts the accounting measurement methods relative to the
measurement methods available under domestic GAAPs. Comparisons are based on standards in effect as of January 1,
1993 (Ashbaugh and Pincus, 2002).

48

Table 4. Continued.

Panel A: Descriptive Statistics for Country-Level Accounting Variables

IFRS Adoption
Countries

Additional
Disclosure
Difference
between IFRS and
Local GAAP
Measurement
Method
Restrictions

Australia 3.00 13.00 6.00
Austria 8.00 20.00 8.00
Denmark 5.00 13.00 5.00
Finland 8.00 19.00 8.00
France 6.00 19.00 5.00
Germany 7.00 20.00 5.00
Italy 6.00 19.00 6.00
Norway 3.00 5.00 4.00
South Africa 1.00 2.00 0.00
Sweden 4.00 11.00 3.00
Switzerland 7.00 17.00 17.00
UK 0.00 15.00 0.00
Venezuela 4.00 8.00 14.00

Mean 4.77 13.92 6.23
Median 5.00 15.00 5.00
Std.dev. 2.55 5.98 4.83

49

Table 4. Continued.
Panel B: Pooled Regression
b

Additional Disclosure
Difference Between
IFRS and Local GAAP
Measurement
Method Restrictions
Variable

Fixed Effect
Estimates
Random
Effect
Estimates
Fixed Effect
Estimates
Random
Effect
Estimates
Fixed Effect
Estimates
Random
Effect
Estimates

Institutional factor 0.449 0.496 0.445 0.103 0.161 -0.069
(<.001) (0.012) (<.001) (0.68) (0.137) (0.291)

(a) Early Voluntary 0.000 -0.007 -0.792 -0.729 -0.206 -0.048
(0.999) (0.984) (0.385) (0.007) (0.269) (0.634)
(b) Early Voluntary -0.433 -0.412 0.506 0.442 -0.804 -0.368
* Institutional factor (0.088) (0.292) (0.58) (0.132) (0.349) (0.085)
Test of (a) + (b) = 0 [p-Value] [0.092] [<.001] [0.128] [0.017] [0.242] [0.031]

(c) Early Voluntary*Post -0.023 -0.022 -0.412 -0.379 -0.076 -0.037
(0.402) (0.975) (0.24) (0.345) (0.138) (0.647)
(d) Early Voluntary*Post -0.053 -0.052 0.348 0.314 0.031 0.002
*Institutional factor (0.383) (0.943) (0.312) (0.46) (0.554) (0.992)
Test of (c) + (d) = 0 [p-Value] [0.140] [0.578] [0.202] [0.638] [0.090] [0.862]

(e) Mandatory -0.026 -0.025 -0.034 -0.035 -0.084 -0.038
(0.579) (0.702) (0.412) (0.609) (0.163) (0.28)
(f) Mandatory -0.204 -0.205 -0.168 -0.167 -0.106 -0.105
* Institutional factor (0.002) (0.023) (0.008) (0.063) (0.087) (0.037)
Test of (e) + (f) = 0 [p-Value] [<.001] [<.001] [0.002] [<.001] [<.001] [<.001]

Control variables Yes Yes Yes Yes Yes Yes
Country dummies Yes No Yes No Yes No

Obs. 868 868 868 868 868 868
Adj. R
2
0.16 0.16 0.17 0.17 0.17 0.17
b
Two-tailed p-values in parentheses, based on t-statistics with standard errors clustered by firm.

50

Table 4, Panel B presents the results of estimating equation (5). The coefficients
of the two binary indicator variables representing the improvement of accounting
information quality (Additional Disclosure and Inconsistencies between IFRS and Local
GAAP) are significantly positive, indicating that the average voting premium is generally
higher in countries with poor accounting information environments.
28

Panel B further demonstrates that the voting premium for the mandatory adopters
significantly decreases after the mandatory adoption of IFRS only when the number of
additional disclosures required by the IFRS reporting is larger (p-value<.001 for fixed
effect estimates; p-value<.001 for random effect estimates). The patterns of the voting
premium for the mandatory adopters in figure 3 are consistent with these findings. These
results are consistent with my prediction that the reduction in the voting premium
resulting from the mandatory adoption of IFRS will be larger in countries for which IFRS
requires more additional disclosures in comparison with the local GAAP reporting
regime.
Panel B demonstrates that the voting premium for the mandatory adopters
significantly decreases after the IFRS mandatory adoption only when the number of
inconsistencies between local GAAPs and IFRS rules is greater (p-value=0.002 for fixed
effect estimates; p-value<.001 for random effect estimates). The patterns of the voting
premium for the mandatory adopters in figure 4 are consistent with these findings.
Overall, these findings suggest, as predicted, that the enhanced comparability of

28
These results are consistent with previous literature (e.g., Dyck and Zingales, 2004; Zingales, 1998).

51

accounting information subsequent to the mandatory switch across jurisdictions is one
factor influencing the post-IFRS adoption decrease in the voting premium.
Finally, Panel B demonstrates that the voting premium for the mandatory adopters
significantly decreases after the IFRS mandatory adoption only when IFRS restricts
management's choices of measurement standards compared to their local GAAP (p-
value<.001 for fixed effect estimates; p-value<.001 for random effect estimates). The
patterns of the voting premium for the mandatory adopters in figure 5 are consistent with
these findings. These findings suggest, as predicted, that the restricted set of MCOs’
measurement choices under IFRS helps explaining the mandatory adopters’ post-IFRS
adoption decrease in the voting premium.
In addition, the table 4 results show that the change in the voting premium for the
mandatory adopters is statistically insignificant between pre- and post-IFRS periods when
the number of additional disclosures required by the IFRS reporting is smaller (p-
value=0.579 for fixed effect estimates; p-value=0.702 for random effect estimates) and
when the number of inconsistencies between local GAAPs and IFRS rules is smaller (p-
value=0.412 for fixed effect estimates; p-value=0.609 for random effect estimates).
Similarly, the voting premium for mandatory adopters does not change between pre- and
post-IFRS periods when the number of accounting measurement methods that are
restricted under IFRS is smaller (p-value=0.163 for fixed effect estimates; p-value=0.280
for random effect estimates). These results suggest that, consistent with my predictions,
the voting premium effects of mandatory IFRS adoption are marginal in countries that
already have high-quality accounting standards in place.

52

Chapter 6.2. The effect of the IFRS mandate on the voting premium,
conditional on firm-specific return volatility
In this section I investigate whether the voting premium effects of the IFRS
mandate are more pronounced for firms where minority shareholders face a higher degree
of informational risk. A higher volatility of returns of the superior voting shares relative
to the inferior voting shares represents a higher degree of informational risk to minority
shareholders.
Prior finance literature demonstrates that a higher volatility of the superior voting
shares denotes that the amount of new private information that arrives among the traders
of the superior voting shares is larger ompared to the amount among the traders of the
inferior voting shares (Ross, 1989; Torben, 1996; Easley, Hvidkjaer and O'Hara, 2002;
Maheu and McCurdy, 2004; Chordia, Huh and Subrahmanyam, 2007).
29
This theory
implies that when the superior voting shares are more volatile than the inferior voting
shares, MCOs, who hold a large chunk of superior voting shares, have new private
information on firms’ value whereas minority shareholders, who hold only a small
fraction of superior voting shares, do not. Because of the increased informational
disparity between MCOs and minority shareholders subsequent to the arrival of new
private information, minority shareholders likely experience higher information gathering
costs and consequently greater adverse selection problems. For example, they may
choose firms where MCOs more significantly appropriate corporate value to the

29
“The informational interpretation of firm-specific return variation of superior voting shares, however, is
not without controversy. For example, limits to arbitrage, pricing errors, and noise also result in volatility.”
(Fernandes and Ferreira, 2007)

53

detriment of minority shareholders. As a result, minority shareholders thus face a higher
degree of informational risk.
I conjecture that ceteris paribus lower-quality accounting information will
exacerbate the informational risk that minority shareholders face. The lower-quality
accounting information indicates that minority shareholders likely have less accurate or
noisier information on firms’ value, consequently exacerbating the informational
disparity between MCOs and minority shareholders when MCOs have more private
information than minority shareholders. Hence, lower-quality accounting information is
predicted to increase the informational risk to minority shareholders when the superior
voting shares are more volatile relative to the inferior voting shares. Facing the higher
degree of informational risk, rational minority shareholders will require a higher price of
controlling parties’ voting right (e.g., a larger voting premium).
To explore whether the mandatory introduction of IFRS reporting is associated
with mitigating this informational risk to minority shareholders, I expand equations (2)
and (3) by including a binary indicator variable dichotomizing firms into two subgroups:
the first (second) group comprise firms with a higher (lower) volatility of returns of the
superior voting shares compared to the inferior voting shares. Specifically, the expanded
multivariate regression model is stated as the following:
Voting premium
it
=
β
0
+ β
1
(Mandatory)
it
+ β
2
(Post)
it
+ β
3
(Mandatory
it
* Post
it
) + Vol Dummy
it
*[ γ
0
+
γ
1
(Mandatory)
it
+ γ
2
(Post)
it
+ γ
3
(Mandatory
it
* Post
it
)]
+ β
4
(Voting power)
it
+β
5
(Div)
it
+ β
6
(Turnover)
it
+ β
7
(Size)
it

+ ∑ β
n
(DCountry) + ε
it
(6)

Where:

54

Vol dummy
it
= A binary indicator variable, which takes on a value of one if the firm’s
volatility of returns of the superior voting shares is higher than that of the inferior
voting shares for a given fiscal year.

Table 5, Panel A presents the firm-level descriptive statistics on the standard
deviation of the returns of the superior voting shares and the inferior voting shares, and
the binary indicator variable, Vol Dummy. The mean and median of the standard
deviation of the superior voting shares (the inferior voting shares) are 3.163 and 0.118
(1.960 and 0.120), suggesting that the distribution of the standard deviation of the
superior voting shares (the inferior voting shares) is right-skewed.
Table 5, Panel B presents the results of estimating equation (6). The coefficients
of the binary indicator variable, Vol Dummy, in Models 1 and 2 are significantly positive,
suggesting that minority shareholders penalize the higher volatility of the superior voting
shares by requiring a higher price of the voting right.
Table 5, Panel B demonstrates that the voting premium significantly decreases
after the IFRS mandatory adoption for the mandatory adopters that have a higher
volatility of returns of the superior voting shares relative to returns of the inferior voting
shares (p-value<.001 for fixed effect estimates; p-value=0.003 for random effect
estimates), while the voting premium does not change between pre- and post-IFRS
periods for the mandatory adopters that have a lower volatility of returns of the superior
voting shares relative to returns of the inferior voting shares (p-value=0.157 for fixed
effect estimates; p-value=0.516 for random effect estimates). These findings are

55

consistent with Figure 5, which illustrates a decreasing pattern of the voting premium for
the mandatory adopters.
Collectively, these results suggest that the decreased informational asymmetry
subsequent to the IFRS mandate reduces information risk that minority shareholders face
when the superior voting shares are relatively more volatile relative to the inferior voting
shares. As a result, rational minority shareholders will reduce the voting premium for
these mandatory adopters.

Figure 5. The Voting Premium Effects of the IFRS Mandate,
Conditional on the Level of Restrictiveness

56

Table 5.
Additional Analysis: Firm-Year Regression Analysis of the Voting Premium Effects
around the IFRS Mandate,
Conditional on Stock Return Volatility

This table reports the results of examining whether the voting premium effects of the
IFRS mandate are more significant for mandatory adopters with a higher volatility of
returns of superior voting shares relative to inferior voting shares. Panel A reports the
firm-level descriptive statistics, and Panel B reports the multivariate regression
coefficients, two-sided p-values, which is calculated based on firm-cluster adjusted t-
statistics of estimating equation (6).

Variable Definitions:
Ret std (low) = The standard deviation of returns of the superior voting shares during a
fiscal year;
Ret std (high) = The standard deviation of returns of the inferior voting shares during a
fiscal year; and
Vol dummy = A binary indicator variable, which takes on a value of one if the firm’s
volatility of returns of the superior voting shares is, on average, higher than that of the
inferior voting shares.

57

Table 5. Continued.

Panel A: Descriptive Statistics for Variables Used in Regression Analyses
Variables N Mean Std. Dev. P25 Median P75
Ret std (low) 868 1.960 11.221 0.033 0.120 0.451
Ret std (high) 868 3.163 13.893 0.033 0.118 0.446
Vol dummy 868 0.529 0.499 0.000 1.000 1.000

Panel B: Pooled Regression
c

Vol Dummy
(Higher Volatility
of Superior Voting Shares =1)

Variable
Fixed
Effect Estimates
Random
Effect Estimates

Vol dummy 0.398 0.213
(0.000) (<.001)

(a) Early Voluntary -0.263 -0.089
(0.260) (0.39)
(b) Early Voluntary * Vol dummy -0.07 -0.024
(0.693) (0.835)
Test of (a) + (b) = 0 [p-value] [0.295] [0.028]

(c) Early Voluntary*Post -0.037 -0.022
(0.451) (0.816)
(d) Early Voluntary* Post * Vol dummy -0.068 -0.023
(0.399) (0.88)
Test of (c) + (d) = 0 [p-value] [0.705] [0.609]

(e) Mandatory -0.054 -0.025
(0.157) (0.516)
(f) Mandatory* Vol dummy -0.126 -0.098
(0.052) (0.057)
Test of (e) + (f) = 0 [p-value] [<.001] [0.003]

Control variables Yes Yes
Country dummies Yes No

Obs. 868 868
Adj. R2 0.21 0.21
c
Two-tailed p-values in parentheses, based on t-statistics with standard errors clustered by firm.

58

Chapter 7. Sensitivity Test
Chapter 7.1. US GAAP and Cross-listing
Prior international studies suggest that firms that adopt U.S. GAAP or cross-list to the
U.S. stock exchanges are different from firms that do not do so. Firms choose to adopt
U.S. GAAP or cross-list to the U.S. stock exchanges with the purpose of: (1) increasing
protection of their minority shareholders (Reese and Weisbach, 2002; Pagano, Randl,
Röell and Zechner, 2001), (2) pursuing a strategy of rapidly increasing their size by
raising more equities in the U.S. stock exchanges and (3) increasing their foreign sales
(Pagano, Röell and Zechner, 2001). Thus, there may be an endogeneity problem in
studying the cross-sectional variation of voting premium between firms that adopt U.S.
GAAP or cross-list to the U.S. stock exchange and firms that do not do so. To address
this potential endogeneity problem, I implement two different research methods. First, I
excluded from my sample firms that adopt U.S. GAAP or cross-list to the U.S. stock
exchange, and repeat the multivariate regression analyses. Second, I included in the
multivariate regression analyses an additional binary indicator variable indicating firms
that adopt U.S. GAAP or cross-list to the U.S. stock exchange. Through these two
methods, I found that the voting premium effects of the IFRS mandate are qualitatively
the same. These results suggest that the potential endogeneity problem associated with
the adoption of U.S. GAAP or with the cross-listing to the U.S. stock exchange does not
influence my main results.

59

Chapter 7.2. Self-selection bias for voluntary adoption of IFRS
Prior IFRS literature illustrates that firms self-select to voluntarily adopt IFRS
because of several economic reasons: financing needs, proprietary costs, accessing
foreign stock exchange etc. (see for example, Leuz and Verrecchia, 2000; Hung and
Subramanyam, 2007). To examine whether this self-selection bias influenced the voting
premium results, I employ the Heckman (1979) two-stage regression procedure. At the
first stage, I model the voluntary adopters’ decisions on whether they will adopt IFRS by
using a probit model. In the probit model, a binary indicator variable of the IFRS
adoption type (voluntary versus mandatory adoption) is regressed onto variables that are
reported to influence firms’ decisions to voluntarily switch from their local GAAP to
IFRS:
Prob (D=1|Z)
it
= Ф(Zγ)
= γ
0
+ γ
1
LEV
it
+ γ
2
ROA
it
+ γ
3
Cross-listed
it
+ γ
4
Size
it
+ γ
5
CS
it
+ γ
6
Debt
it

+e
it
(7)
where
D = a dummy variable, which takes on a value of one for a voluntary adopter;
Z = a vector of economic factors influencing a firm’s decision to voluntarily adopt
IFRS;
Γ = a vector of coefficients;
Ф = the cumulative distribution function of the standard normal distribution;
LEV = a firm’s total liability divided by its total assets;
ROA = a firm’s net income divided by its total assets;
Cross-listed = a dummy variable, which takes on a value of one if the firm’s share is
cross-listed to the U.S. stock exchange (data=2004 J.P. Morgan ADR);
Size = the natural logarithm of market value of the sum of inferior and superior
voting shares;
CS = a dummy variable, which takes on a value of one if inferior share price at par
increase during the fiscal year; and
Debt = a dummy variable, which takes on a value of one if firm’s long-term debt
increases during the fiscal year.

60

Estimation of the preceding probit model generates results that are used to predict the
probability for each voluntary adoption of IFRS. In the second stage, I correct for self-
selection by including into my regression models a transformation of these predicted
individual probabilities (i.e., inverse mills’ ratio) as an additional independent variable.
The untabulated results show that the coefficients for the inverse mills ratios are
significantly negative (p-value < 0.10 for both fixed and random effects model) in all
regression models, indicating the potential impact of self-selection bias on the empirical
results. However, signs and significance levels of the main coefficients are qualitatively
the same. Therefore, I conclude that self-selection bias is unlikely to cause the observed
voting premium effects of the IFRS mandate.
Chapter 7.3. Different benchmark groups
In this paper I employed early voluntary adopters as a benchmark group, whose
voting premium change reflects contemporaneous macroeconomic shocks in the sample
countries around the IFRS mandate. As stated in the section 6.2., early voluntary adopters
are different from mandatory adopters in several economic dimensions including size,
profitability and corporate governance, and the number of early voluntary adopters is
small. Hence, early voluntary adopters may not be a good control group against which the
changes of the voting premium for the mandatory adopters are benchmarked. To address
this potential problem, I additionally employ four different subgroups of firms as a
control group, against which the changes of the voting premium for mandatory adopters
are benchmarked. The first set of the control groups includes firms that domicile in
countries that mandate IFRS reporting during the sample period and report under the

61

local GAAP throughout the sample period. Small or medium firms are exempt from
adopting IFRS and permitted to continuously report under the local GAAP even after
IFRS reporting is adopted by the IFRS adoption countries. The advantage of this control
group is that these firms domicile in the same capital market with the mandatory adopters,
and accordingly, the changes of the voting premium for these firms capture the
concurrent capital market shocks that influence the voting premium for the mandatory
adopters.
The second set of the control groups is the U.S. firms that have a dual-class
structure and satisfy my sample selection criteria. The U.S. market is considered to be
relatively frictionless. Accordingly, the U.S. firms have been widely used as a benchmark
in international accounting literature. The third set of the control groups comprises firms
that domicile in countries that do not adopt IFRS during the sample period. These firms
report under domestic GAAP throughout the sample period. These control samples are
composed of 1,492 firm-year observations from seven countries that allow dual-class
structure and have non-zero firms that satisfy my sample selection criteria. These seven
countries are Brazil, Canada, Chile, Columbia, Korea, Mexico and the United States.
Consistent with prior voting premium literature, I find that Brazil and Korea have the
largest number of dual-class firms with 81 and 105 among these seven countries,
respectively (Nenova, 2003; Doidge, 2004). Next, to address the possibility that countries
that have a large number of dual-class firms (e.g., Brazil and Korea) unduly influence my
empirical results, I randomly select up to 30 firms in each country as the third set of the
control groups. The final set of the control groups comprises all dual-class firms in

62

countries that do not adopt IFRS during the sample period. Here, the latter three sets of
control groups are from different capital markets compared to the mandatory adopters.
However, given that countries’ capital markets are increasingly integrated, the change of
the voting premium for these control groups is expected to capture the concurrent capital
market shocks that influence the change of the voting premium of the mandatory adopters.
Table 6 reports the results of the estimate of equations (2) and (3) based on the
preceding four sets of the control groups. Table 6 demonstrates that the mandatory
adopters’ voting premium effects of the IFRS mandate are significant across different sets
of the control groups (p-value <.100 for fixed effects model and p-value <.050 for
random effects model). Collectively, these results indicate that the observed voting
premium effects of the IFRS mandate in Table 3 are not sensitive to different sets of the
control groups.

63

Table 6. Sensitivity Test: Firm-Year Regression Analysis
of the Voting Premium Effects around the IFRS Mandate, with Different Control Groups

This table reports the results of sensitivity tests based on four different control groups. In particular, Model 1 reports the
multivariate regression coefficients and two-sided p-values, which are calculated based on firm-cluster adjusted t-statistics for
estimating equations (2) and (3) with the non-IFRS adopters in countries that adopt IFRS reporting during the sample period.
Model 2, 3 and 4 report these figures with the U.S. firms, randomly-selected and all the non-IFRS adopters in countries that do
not adopt IFRS reporting during the sample period as a benchmark group, respectively.

64
Table 6. Continued.

Model 1:
IFRS Adoption Firms
Plus
Non-IFRS adopters
in Treatment Sample
Model 2:
IFRS Adoption Firms
Plus
U.S. Firms
Model 3:
IFRS Adoption Firms
Plus
Randomly Selected
Non-IFRS adopters
in Control Sample
Model 4:
IFRS Adoption Firms
Plus
All Non-IFRS adopters
in Control Sample
Variable
Fixed Effect
Estimates
Random Effect
Estimates
Fixed Effect
Estimates
Random Effect
Estimates
Fixed Effect
Estimates
Random Effect
Estimates
Fixed Effect
Estimates
Random Effect
Estimates

Early Voluntary -0.326 -0.317 -0.333 -0.303 -0.354 -0.315 -0.421 -0.317
(0.005) (0.006) (0.028) (0.044) (0.009) (0.019) (0.001) (0.006)
Early Voluntary* Post -0.085 -0.084 -0.088 -0.084 -0.084 -0.08 -0.088 -0.084
(0.544) (0.551) (0.631) (0.647) (0.609) (0.626) (0.589) (0.551)
Mandatory -0.121 -0.122 -0.139 -0.138 -0.119 -0.11 -0.107 -0.122
(0.011) (0.011) (0.029) (0.029) (0.036) (0.050) (0.055) (0.011)
Size -0.020 -0.020 -0.028 -0.028 -0.021 -0.02 -0.022 -0.020
(0.077) (0.079) (0.028) (0.029) (0.047) (0.051) (0.004) (0.079)
Div -0.08 -0.081 -0.154 -0.154 -0.099 -0.099 -0.038 -0.081
(<.001) (<.001) (<.001) (<.001) (<.001) (<.001) (0.002) (<.001)
Voting power 0.021 0.021 0.006 0.006 0.007 0.006 -0.005 0.021
(0.017) (0.017) (0.494) (0.534) (0.272) (0.341) (0.369) (0.017)
Turnover -0.037 -0.036 -0.031 -0.029 -0.029 -0.027 -0.048 -0.036
(0.001) (0.001) (0.021) (0.028) (0.003) (0.005) (<.001) (0.001)

Country dummies Yes No Yes No Yes No Yes No

Obs. 921 921 1,099 1,099 1,403 1,403 2,413 2,413
Adj. R
2
0.26 0.26 0.10 0.10 0.08 0.08 0.06 0.06

c
Two-tailed p-values in parentheses, based on t-statistics with standard errors clustered by firm.

65

Chapter 8. Conclusion
This study examines whether and through what mechanisms the mandatory
introduction of IFRS reporting is associated with a decrease in the voting premium for
firms with a dual-class share structure. I find that after the mandatory introduction of
IFRS reporting, the voting premium for the mandatory adopters in the 13 countries
decreases by 13.2% on average. This decrease in the voting premium is statistically
significant relative to the several benchmark groups of firms—the voluntary adopters, the
U.S. firms, the non-IFRS adopters in countries that adopt IFRS reporting during the
sample period and the non-IFRS adopters in countries that do not adopt IFRS reporting
during the sample period.
Furthermore, the voting premium effects of the mandatory switch appear to be
greater when the firms have higher informational uncertainty. Additional analyses
highlight that the voting premium effects obtain through three mechanisms, namely, the
increased transparency and comparability of accounting information as well as a
restricted set of measurement methods associated with the mandatory switch to IFRS.
Taken together, these results indicate that the IFRS mandate benefits minority
shareholders by providing an effective mechanism to constrain the private benefits of
control.
The results of this study are subject to two caveats. First, the voting premium may
be a biased proxy for the size of the average firm’s control benefits. Controlling
shareholders choose the dual-class structure for a reason, for instance, to operate their
firm in greater secrecy so as to conceal their diversionary behavior from minority

66

shareholders. Hence, the size of the control benefits in dual-class firms is likely to be
biased upward in comparison to the size of the control benefits in single-class firms
(DeAngelo and DeAngelo, 1985; Grossman and Hart, 1988; Doidge, 2003). However, at
least two factors mitigate the upward bias of the voting premium as a proxy for the size
of control benefits: 1) the degree of the upward bias is alleviated by the fact that the
voting premium is set by the stock transactions of minority shareholders, who are less
likely to enjoy the control benefits — the voting premium can thus be regarded as a lower
bound of the magnitude of private control benefits; and 2) firms with a single-share
structure can also separate control and cash flow rights (like firms with the dual-class
structure) through other mechanisms such as pyramid structures and cross-holdings
(Bebchuck et al., 2000; Fan and Wong, 2002; Doidge, 2003).
The second caveat is that, while this paper provides evidence that the mandatory
switch to IFRS benefits minority shareholders by decreasing controlling shareholders’
diversionary behavior, the total costs and benefits of the mandatory switch are still an
open question. Even though the mandatory switch successfully constrains controlling
owners from extracting value from firms to the detriment of minority shareholders, there
are non-negligible direct and indirect costs associated with the mandatory switch (see Li,
2010). Further research into the relative benefits and costs of the mandatory switch to
IFRS is thus needed, especially given the possibility of mandatory IFRS adoption in the
U.S. in the near future.

67

Bibliography

Amoako-Adu, B. and B. Smith, 1995. “Outside financial directors and corporate
governance" in corporate decision making in Canada. Industry Canada, University of
Calgary Press.

Armstrong, C., M.E. Barth, A.D. Jagolinzer and E.J. Riedl, 2009. Market Reaction to the
Adoption of IFRS in Europe. The Accounting Review (Forthcoming).

Ashbaugh, H., and M. Pincus. 2001. Domestic accounting standards, International
Accounting Standards, and the predictability of earnings. Journal of Accounting
Research 39: 417-434.

Ball, R., A. Robin, and J. Wu, 2003. Incentives versus standards: Properties of
accounting income in four East Asian countries. Journal of Accounting and Economics
36: 235-270.

Ball, R., Shivakumar L., 2006. The role of accruals in asymmetrically timely gain and
loss Recognition. Journal of Accounting Research 42: 207-242.

Barclay, M. and C. Holderness, 1989. Private benefits from control of public corporations.
Journal of Financial Economics 25: 371-395.

Bebchuk, L.A., R. Kraakman and G. Trianis, 2000. Stock pyramids, cross-ownership, and
dual class equity, in: R.K. Morck, ed., Concentrated Corporate Ownership (NBER
Conference Report Series, University of Chicago Press, Chicago).

Berle, A. A., and G.C. Means, 1932. The modern corporation and private property
(Macmillian, New York).

Biddle, G., and G. Hilary, 2006. Accounting quality and firm-level capital investment.
The Accounting Review 81: 963-982.

Burgstahler, D. and M. Eames, 2006. Management of earnings and analysts' forecasts to
achieve zero and small positive earnings surprises. Journal of Business Finance &
Accounting 33: 633-652.

Chordia, T., Huh, S.W., Subrahmanyam, A., 2007. The cross-section of expected trading
activity. Review of Financial Studies 20: 709–740.
Cox, C., 2009. Speech by SEC Chairman: 'International Financial Reporting Standards:
The Promise of Transparency and Comparability for the Benefit of Investors around
the Globe' http://www.sec.gov/news/speech/2008/spch052808cc.htm

68

Christensen, H., Lee, E., Walker, M., 2007. Cross-sectional variation in the economic
consequences of international accounting harmonisation: The case of mandatory IFRS
adoption in the UK. International Journal of Accounting 42: 341-379.

Daske, H., L. Hail, C. Leuz, and R. Verdi. 2007. Adopting a label: Heterogeneity in the
economic consequences of IFRS adoptions. Working paper, Goethe University of
Frankfurt, the Wharton School, University of Chicago, and MIT.

Daske, H., L. Hail, C. Leuz, and R. Verdi. 2008. Mandatory IFRS reporting around the
world: Early evidence on the economic consequences. Journal of Accounting Research
46: 1085-1142.

DeAngelo, H., and L. DeAngelo, 1985. Managerial ownership of voting rights: A study
of public corporations with dual classes of common stock. Journal of Financial
Economics 14: 33-69.

Doidge, C., 2003. U.S. cross-listings and the private benefits of control: evidence from
dual-class firms. Journal of Financial Economics 72: 519-553.

Dyck, A., and L. Zingales, 2004. Private benefits of control: An international comparison.
Journal of Finance 59: 537-600.

Easley, D., S. Hvidkjaer and M. O'Hara, 2002. Is information risk a determinant of asset
returns? The Journal of Finance 57: 2185-2222.

Enriques, L. and P. Volpin, 2007. Corporate governance reforms in continental Europe.
Journal of Economic Perspecitves 21: 118-119.

Fan, J. P. H. and T. J. Wong, 2002. Corporate ownership structure and the
informativeness of accounting earnings in East Asia. Journal of Accounting and
Economics 33: 401-425.

Faulkender, Michael, and Rong Wang, 2006. Corporate financial policy and the value of
cash. Journal of Finance 61: 1957-1990.

Francis, J., Schipper, K., Vincent, L., 2005. Earnings and dividend informativeness when
cash flow rights and separated from voting rights. Journal of Accounting and
Economics 39: 329–360.

Financial Accounting Standards Board (FASB), 1996. The IASC-U.S. comparison
project: a report on the similarities and differences between IASC standards and U.S.
GAAP.

69

Gompers, P. A., J. Ishii, and A. Metrick, 2009. Extreme governance: An analysis of dual-
class firms in the United States. The Review of Financial Studies 118:107–55.

Grossman, S., and O. Hart, 1988. One share–one vote and the market for corporate
control. Journal of Financial Economics 20: 175-202.

Haw, I., B. Hu, L. Hwang and W. Wu. 2004. Ultimate Ownership, Income Management,
and Legal and Extra-Legal Institutions. Journal of Accounting Research 42: 423-462.

Hope, O.; J. Jin; and T. Kang, 2006. Empirical evidence on jurisdictions that adopt IFRS.
Journal of International Accounting Research 5: 1-20.

Hung, M., and K.R. Subramanyam. 2007. Financial statement effects of adopting
International Accounting Standards: The case of Germany. Review of Accounting
Studies 12: 623-657.

Jensen, M.C. and W.H. Meckling, 1976. Theory of the firm: Managerial behavior, agency
costs and ownership structure. Journal of Financial Economics 3: 305-360.

Jensen, M. C. and R. S. Ruback, 1983. The market for corporate control: The scientific
evidence. Journal of Financial Economics 11: 5-50.

Johnson, S., R. La Porta, F. Lopez-de-Silanes, and A. Shleifer, 2000. Tunnelling.
American Economic Review 90: 22-27.

Johnson, S., P. Boone, A. Breach and E. Friedman, 2000. Corproate governance in the
Asian financial crisis. Journal of Financial Economics 58: 141-186.

La Porta, R., F. Lopez-de-Silanes, A. Shleifer, R. Vishny, 1997. Legal determinants of
external finance. Journal of Finance 52: 1131–1150.

La Porta, R., F. Lopez-de-Silanes, A. Shleifer, R. Vishny, 1998. Law and finance. Journal
of Political Economy 106: 1113–1155.

Lambert, R., C. Leuz, and R. Verrecchia, 2005. Accounting information, disclosure, and
the cost of capital. Journal of Accounting Research 45: 385-420.

Lemon, M. and K. Lins, 2003. Ownership strucuture, corporate governance, and firm
value: evidence from the East Asian financial crisis. Journal of Finance 58(4): 1445-
68.

Leuz, C., D. Nanda, and P. Wysocki. 2003. Investor Protection and Earnings Management:
An International Comparison. Journal of Financial Economics 69: 505-527.

70

Leuz, C., and R. Verrecchia, 2000. The economic consequences of increased disclosure.
Journal of Accounting Research 38: 91-124.

Li, S., 2010. Does mandatory adoption of International Accounting Standards reduce the cost
of equity capital?, The Accounting Review (forthcoming).

Maheu, J. M. and T. H. McCurdy, 2004. News arrival, jump dynamics, and volatility
components for individual stock returns. Journal of Finance 59: 755-793.

Masulis, R., C. Wang and F. Xie, 2009. Agency costs at dual-class companies, Journal of
Finance 64: 1697-1727.

McNichols, M.F. and S.R. Stubben, 2008. Does Earnings Management Affect Firms’
Investment Decisions? The Accounting Review 83: 1571-1604.

Meyer, B.D., 1995. Natural and Quasi-Experiments in Economics. Journal of Business
and Economic Statistics 13: 151-162.

Mitton, T., 2002. A cross-firm analysis of the impact of corporate governance on the
East Asian financial crisis. Journal of Financial Economics 64(2): 215-241.

Modigliani, F., and E. Perrotti, 1997. Protection of minority interest and the development
of security markets. Managerial and Decision Economics 18: 519-528.

Nenova, T., 2002. The value of corporate votes and control benefits: Cross-country
analysis. Journal of Financial Economics 68: 325-351.
Nobes, C.W. 2002. An analysis of the international development of the equity method.
Abacus 38: 16 - 45.

Pae, J., D. Thornton, and M. Welker, 2007. Agency Cost Reduction Associated with EU
Financial Reporting Reform. The Journal of International Accounting Research 7: 51-
76.

Renders A, Gaeremynck A, 2007. The impact of legal and voluntary investor protection
on the early adoption of International Financial Reporting Standards (IFRS).
Economist-netherlands 155: 49 - 72.

Ross, S. 1989. Information and volatility: the no-arbitrage martingale approach to timing
and resolution irrelevancy. The Journal of Finance 44:.1-17.

Rydqvist, K., 1996. Takeover bids and the relative prices of shares that differ in their
voting rights. Journal of Banking and Finance 20: 1407-1425.

71

Shleifer, A., and RW Vishny, 1997. A survey of corporate governance. The Journal of
Finance LII: 688-726.

Sunder, S. 2007. Uniform financial reporting standards: Reconsidering the top-down
push. The CPA Journal 77: 6-9.

Tweedie, D. 2002. Statement of sir David Tweedie chairman, International Accounting
Standards Board before the committee on banking, housing and urban affairs of the
United States senate Washington, D.C. http://www.iasplus.com/resource/020214dt.pdf

Zingales, L, 1994. The value of the voting right: A study of the Milan Stock Exchange
experience. Review of Financial Studies 7: 125-148.

Zingales, L, 1995. What determines the value of corporate votes? Quarterly Journal of
Economics 110: 1047-1073.

Zingales, L., 1998. Why it’s worth being in control, in Financial Times Mastering
Finance: The Complete Finance Companion, Pitman Publishing, London, U.K.

72

Appendix I. An Analytical Model
This appendix presents the procedure, through which the solution of the optimal
f* is derived from the first-order condition of the problem (equation (2)). For a
convenience I will let the first-order condition be a function U, which is given by
U

s c

P 2c

s 1

1 P 1 s
0 6
After rearranging the equation (6), I obtain
U

c

1
2 1

1
0 7
I can now examine this first order condition to derive implications of the model with
respect to the effect of accounting standards on MCOs’ diversionary behavior.
Differentiating the function U with respect to q, I obtain
U

c

1

c

1

2

1

2
1

2
1
2

0 8
After rearranging the equation (8) with respect to

, I obtain

c

1

2 2c
s 1

2cP

c

1

2
1 (9)

73

Finally, under the assumptions on the probability of being caught, P, the equation (9)
implies
d

dq

∂

P
∂ ∂
c

1
∂P
∂
2
1 !
c

1
∂

P
∂

2 2c
s 1
∂P
∂
2cP!

" 0 10

74
Appendix II. Share Structure across Countries (Doidge, 2004)

Sample Countries Voting Arrangements
IFRS Adoption Countries:
Australia Ordinary shares have 1 vote. Preferred shares have limited voting rights in special circumstances only.
Austria Ordinary shares have 1 vote. Preferred shares are nonvoting.
Denmark ‘A’ shares have 10 votes per share. ‘B’ shares have 1 vote per share.
Finland High voting shares have 10 to 20x the voting rights of low voting shares.
France Ordinary shares have 1 vote. Preferred shares are nonvoting.
Germany Ordinary shares have 1 vote. Preferred shares are nonvoting.
Italy Ordinary shares have 1 vote. Savings shares are nonvoting.
Norway ‘A’ shares have 1 vote. ‘B’ shares are nonvoting.
South Africa Ordinary shares have voting rights from 100 - 500x the voting rights of low voting shares.
Sweden ‘A’ shares have 1 vote per share. ‘B’ shares have 0.1 votes per share.
Switzerland Ordinary voting & nonvoting shares.
UK Ordinary voting & nonvoting shares; or high voting shares have 2 - 100x the voting rights of low voting shares.
Venezuela Ordinary voting & nonvoting shares.

Non-IFRS Adoption Countries:
Brazil ON (ordinary) shares have 1 vote. PN (preference) shares are nonvoting.
Canada High voting shares have voting rights from 10 - 500x the voting rights of low voting shares.
Chile ‘A’ shares elect 6/7 or 7/8 directors. ‘B’ shares elect 1/7 or 1/7 directors.
Columbia Ordinary shares have 1 vote. Preferred shares are nonvoting.
Korea Ordinary shares have 1 vote. Preferred shares are nonvoting.
Mexico ‘A’, ‘B’, & ‘O’ shares have 1 vote. ‘C’ shares are nonvoting & ‘L’ shares are effectively non-voting.
United States Ordinary voting & nonvoting shares; or high voting shares have 1.5 - 100x the voting rights of low voting
shares.

This table reports various voting arrangements in the sample countries. The contents of this table are excerpted from Doidge (2004).

Abstract (if available)

Abstract Proponents of International Financial Reporting Standards (IFRS) claim that one of the consequences of switching to IFRS is greater minority shareholder protection. In this paper I test this assertion by setting up a model of the effect of the accounting standards on controlling parties’ expropriating behavior. I then apply this model to firms where control and cash flow rights are separated by a dual-class structure and examine whether mandatory adopters’ voting premium is reduced following the IFRS mandate. Based on panel data set of 151 dual-class firms in 13 mandatory IFRS adoption countries during the period 2002 to 2007, I find that after the mandatory introduction of IFRS reporting, mandatory adopters’ voting premium decreases, on average, by 13.2%. This effect for mandatory adopters is statistically significant relative to the corresponding effects for voluntary adopters and non-IFRS adopters. In addition, the effect of mandatory IFRS adoption on the voting premium is more pronounced for firms where minority shareholders face higher informational risk. I further show that the decrease in the voting premium subsequent to the IFRS mandate follows because of: (1) expanded disclosure, (2) enhanced comparability of firms’ reported information across borders and (3) a restricted set of measurement methods under IFRS. Taken together, these results suggest that mandatory IFRS adoption benefits minority shareholders by providing an effective mechanism to constrain the private benefits of control.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Does mandatory adoption of International Accounting Standards reduce the cost of equity capital?

Asset Metadata

Creator Hong, Hyun A. (author)

Core Title Does mandatory adoption of international financial reporting standards decrease the voting premium for dual-class shares: theory and evidence

School Leventhal School of Accounting

Degree Doctor of Philosophy

Degree Program Business Administration

Degree Conferral Date 2010-08

Publication Date 08/10/2012

Defense Date 04/16/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag dual class structure,international financial reporting standard,OAI-PMH Harvest,private control benefits

Place Name Australia (countries), Austria (countries), Denmark (countries), Finland (countries), France (countries), Germany (countries), Italy (countries), Norway (countries), South Africa (countries), Sweden (countries), Switzerland (countries), United Kingdom (countries), USA (countries), Venezuela (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Hung, Mingyi (committee chair), Trezevant, Robert (committee chair), DeFond, Mark (committee member), Matsusaka, John (committee member), Subramanyam, K. R. (committee member)

Creator Email honghyuna@gmail.com,hyun.hong.2010@marshall.usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3383

Unique identifier UC177331

Identifier etd-hong-3831 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-382100 (legacy record id),usctheses-m3383 (legacy record id)

Legacy Identifier etd-hong-3831.pdf

Dmrecord 382100

Document Type Dissertation

Rights Hong, Hyun A.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu