Corporate Ownership & Control / Volume 6, Issue 3, Spring ...

Corporate Ownership & Control / Volume 6, Issue 3, Spring 2009

3

CORPORATE

OWNERSHIP & CONTROL

Postal Address: Postal Box 36 Sumy 40014 Ukraine Tel: +380-542-611025 Fax: +380-542-611025 e-mail: [email protected] [email protected] www.virtusinterpress.org Journal Corporate Ownership & Control is published four times a year, in September-November, December-February, March-May and June-August, by Publishing House “Virtus Interpress”, Kirova Str. 146/1, office 20, Sumy, 40021, Ukraine. Information for subscribers: New orders requests should be addressed to the Editor by e-mail. See the section "Subscription details". Back issues: Single issues are available from the Editor. Details, including prices, are available upon request. Advertising: For details, please, contact the Editor of the journal. Copyright: All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by any means without the prior permission in writing of the Publisher.

Corporate Ownership & Control

ISSN 1727-9232 (printed version) 1810-0368 (CD version) 1810-3057 (online version)

Certificate № 7881

Virtus Interpress. All rights reserved.

КОРПОРАТИВНАЯ

СОБСТВЕННОСТЬ И КОНТРОЛЬ

Почтовый адрес редакции: Почтовый ящик 36 г. Сумы, 40014 Украина Тел.: 38-542-611025 Факс: 38-542-611025 эл. почта: [email protected] [email protected] www.virtusinterpress.org Журнал "Корпоративная собственность и контроль" издается четыре раза в год в сентябре-ноябре, декабре-феврале, марте-мае, июне-августе издательским домом Виртус Интерпресс, ул. Кирова 146/1, г. Сумы, 40021, Украина. Информация для подписчиков: заказ на подписку следует адресовать Редактору журнала по электронной почте. Отдельные номера: заказ на приобретение отдельных номеров следует направлять Редактору журнала. Размещение рекламы: за информацией обращайтесь к Редактору. Права на копирование и распространение:

копирование, хранение и распространение материалов журнала в любой форме возможно лишь с письменного разрешения Издательства.

Корпоративная собственность и контроль

ISSN 1727-9232 (печатная версия) 1810-0368 (версия на компакт-диске) 1810-3057 (электронная версия)

Свидетельство КВ 7881 от 11.09.2003 г.

Виртус Интерпресс. Права защищены.


4

EDITORIAL

Dear readers!

We continue to keep our hand on a pulse of the recent corporate governance concepts and practices worldwide. This issue of Corporate Ownership and Control addresses attention of the reading audience to a wide spectrum of problems in corporate governance. These problems concerns such issues as board of director practices, director remuneration, merges, financial issues in corporate governance, corporate social responsibility. Such wide spectum of issues which are considered in the recent issue of the journal allows us concluding that corporate governance as a field of research develops outside its generally accepted limits. Truly international representation of contributors support the above point of view and concludes that corporate governance gets through the borders of the science fast. Our contributors represent both developed and developing countires such us Germany, Italy, Australia, South Africa, India, Belgium, Greece, UK and Nigeria. Special section of the journal is devoted to corporate board practices. We have published papers considering various aspects of corporate board practices including director remuneration, board monitoring, etc. The section on national practices of corporate governance narrates on corporate governance in South Africa. We need to declare that we enjoyed cooperating with corporate governance experts from South Africa. During last three years we published in our journal more that 15 papers of corporate governance experts from South Africa. We need to state that their performance in researching corporate governance is very high. We hope you will enjoy reading the journal and become a loyal contributor or subscriber!


5

CORPORATE OWNERSHIP & CONTROL

Volume 6, Issue 3, Spring 2009

CONTENTS

Editorial 4

SECTION 1. ACADEMIC INVESTIGATIONS AND CONCEPTS

TAKEOVER ACTIVITIES IN THE LAST EUROPEAN MERGER WAVE: A CROSS-COUNTRY COMPARISON 8

Mirella Damiani, Fabrizio Pompei

In the present work we attempt to fill this gap, evaluating on empirical grounds not only the role of corporate governance systems and investor protection factors, but also the influence that alternative technological regimes can play on mergers and takeovers. This comprehensive analysis is another step along the lines suggested by Hall and Soskice (2001), two authors who have shown that the industrial specialization of each country may be seen in its complementarity with its institutional framework. Until now, this complementarity between production regimes and varieties of capitalism has not been fully explored in terms of the role of the market for corporate control. THE CORPORATE SOCIAL RESPONSIBILITY MANAGER MAP 26 Mario Molteni, Matteo Pedrini This article outlines the profile of ninety managers directly involved in Corporate Social Responsibility activities (CSR Managers) in Italian firms. It presents an analysis of the organisational position, educational background and activities of these professionals. The results suggest that CSR managers: 1) have a growing relevance in the firm; 2) are predominantly existing members of the organization; 3) have principally a business management educational background; 4) play a key role in supporting senior management and improving stakeholder engagement. IMPACT OF INDEX DERIVATIVES ON INDIAN STOCK MARKET VOLATILITY-AN APPLICATION OF ARCH AND GARCH MODEL 39 S. V. Ramana Rao, Naliniprava Tripathy The present study examined the impact of introduction of index futures derivative and index option derivative on Indian stock market by using ARCH and GARCH model to capture the time varying nature of volatility presence in the data period from October 1995 to July 2006. The results reported that the introduction of index futures and index options on the Nifty has produced no structural changes in the conditional volatility of Nifty but however the market efficiency has been improved after the introduction of the derivative products. CAVEAT WACC: PITFALLS IN THE USE OF THE WEIGHTED AVERAGE COST OF CAPITAL 45 Sebastian Lobe The objective of this paper is thus twofold. First, it is clarified that a constant WACC rate must fail if the implied leverage ratio is time-varying. Second, although the NLWACC approach is further amplified in


6

this paper, it must be emphasized that this approach is, even then, applicable only under specific conditions while a time-varying WACC is still able to provide reliable results.

DEBT OF FAMILY FIRMS: A COMPARISON BASED ON ACCOUNTING INDICATORS 53 Olivier Colot, Mélanie Croquet The financial behavior of family firms represents a field of research that has been little explored up to the present time. In this context, we wanted to contribute to understanding the problems linked to financing in family firms, and more specifically to family SMEs in Belgium, because they represent a major part of the Belgian economy. The results show that family SMEs seem to be more indebted. Among all of the ratios tested, six indicators were statistically significant.

AUDITOR CONSERVATISM AND EARNINGS MANAGEMENT: EVIDENCE FROM TUNISIA 61 Nadaa Hachicha Elfouzi, Mohamed Ali Zarai

The aim of this paper is to examine the effect of the type of audit opinions on the earnings management measured by the discretionary accruals in the tunisian capital market context. In particular, we investigate whether abnormal accruals are influenced by modified audit opinions. EFFECTS OF HUMAN RESOURCE PRACTICES ON FAMILY FIRMS SOCIAL PERFORMANCE 69 Olivier Colot, Claire Dupont, Mélanie Volral The aim of our research is to analyse social performance (through turnover rate) of large family owned business in relation to their human resource practices. We made multiple regressions on a sample of 60 large firms. Our global model, considering large family owned business and non-family owned business, shows that part-time contracts increase turnover significantly, while training reduces it.

SECTION 2. BOARD OF DIRECTORS

BOARD MONITORING AND FIRM PERFORMANCE: CONTROLLING FOR ENDOGENEITY AND MULTICOLLINEARITY 79 Mohammad I Azim, Dennis W Taylor In this study, panel data of the top 500 listed companies from the Australian Stock Exchange is used over three years. Results reveal that all but one of the five board characteristics and seven board committee characteristics considered in this study are significantly related to both return on assets and earnings per share in each of the three years.

EXECUTIVE BOARD MEMBERS’ REMUNERATION: A LONGITUDINAL STUDY 94 Themistokles Lazarides, Evaggelos Drimpetas, Koufopoulos Dimitrios Remuneration is considered to be closely connected with financial performance (positively), firm size (positively), the organizational structure (negatively) and corporate governance mechanisms (negatively). Furthermore, a connection of ownership structure and executives’ remuneration has been well established (theoretically and empirically) in the literature (agency theory). The paper examines if these relationships are valid in Greece. Overall the study has shown that remuneration levels in Greece are defined by a different set of factors than the ones that are prominent in an Aglo-Saxon country.


7

BAD DEBTS, OWNERSHIP CONCENTRATION, AND BOARD COMPOSITION: EVIDENCE ON THE QUALITY OF CORPORATE GOVERNANCE OUTCOMES IN CHINA’S LISTED NON FINANCIAL COMPANIES 104 Yuan George Shan, Ron P. McIver This study analyses the relationships between performance metrics and the corporate control and governance characteristics of a sample of China’s listed non financial companies in order to assess the influence of corporate governance structures on the quality and independence of corporate decision making. We use a panel data set covering the years 2001 to 2005 comprised of a stratified sample of A, AB and AH non financial companies listed on China’s Shanghai and Shenzhen stock exchanges.

AUDIT COMMITTEES IN NIGERIA 117 Ogbuagu Ekumankama, Chibuike Uche This paper reviews the law and practice of audit committees in Nigeria. It argues that for audit committees to become more useful in the Nigerian context there is need for changes to be made in both its law and practice. Key areas of concern include the need to: determine and codify the qualification for membership of the committee given its technical nature; allow appropriate remuneration for committee members, and; the determination of appropriate membership tenure for such committees.

SECTION 3. CORPORATE GOVERNANCE PRACTICES: SOUTH AFRICA

DO PORTFOLIO MANAGERS IN SOUTH AFRICA CONSIDER HUMAN BEHAVIOUR ISSUES WHEN MAKING INVESTMENT DECISIONS OR ADVISING CLIENTS? 126 Strydom JSG, Van Rooyen JH This research reviews some results from the behavioural finance and other related literature. A survey was also done to determine whether the most prominent portfolio managers in South Africa are aware of behavioural finance issues/models and consider the influence of cognitive issues when making investment decisions or giving advice to clients.

LAYER HISTOGRAM PATTERNS IN FINANCIAL TIME SERIES 137 Van Zyl-Bulitta, VH, Otte, R, Van Rooyen, JH This study aims to investigate whether the phenomena found by Shnoll et al. when applying histogram pattern analysis techniques to stochastic processes from chemistry and physics are also present in financial time series, particularly exchange rate and index data. The phenomena are related to fine structure of non-smoothed frequency distributions drawn from statistically insufficient samples of changes and their patterns in time.

THE EFFECT OF PROCUREMENT STRATEGIES OF MILLING COMPANIES ON THE PRICE OF MAIZE 147 W. Rossouw, J. Young Even though derivative instruments are available to use as counter against market fluctuations, the price risk management success of groups with a concern on SAFEX suggests that this is not achieved as yet, ultimately to the detriment of consumers. The view exists that markets are efficient and the return offered by the futures exchange cannot consistently be outperformed. This paper argues, since the use of the proposed futures/options strategies result in returns superior to that of the market. SUBSCRIPTION DETAILS 158


8

РАЗДЕЛ 1 НАУЧНЫЕ ИССЛЕДОВАНИЯ

И КОНЦЕПЦИИ

SECTION 1 ACADEMIC INVESTIGATIONS & CONCEPTS

TAKEOVER ACTIVITIES IN THE LAST EUROPEAN MERGER WAVE:

A CROSS-COUNTRY COMPARISON*

Mirella Damiani**, Fabrizio Pompei***

Abstract

The efficiency of “markets vs corporations” in reallocating assets, changing industry structure and moving capital from declining industries to emerging industries is a very important issue. The vast literature on this subject has examined the role played by corporate governance systems, technological shocks and institutional factors in triggering mergers and takeovers, but has not considered the specific influence that technological regimes of innovation can exert in reallocating assets and moving capital among sectors. In the present work we attempt to fill this gap, evaluating on empirical grounds not only the role of corporate governance systems and investor protection factors, but also the influence that alternative technological regimes can play on mergers and takeovers. This comprehensive analysis is another step along the lines suggested by Hall and Soskice (2001), two authors who have shown that the industrial specialization of each country may be seen in its complementarity with its institutional framework. Until now, this complementarity between production regimes and varieties of capitalism has not been fully explored in terms of the role of the market for corporate control. The present paper is a first attempt at filling this gap, by taking into account the European experience of the last few years (2002-2005) which seems to mark a new wave in M&A activities.

Keywords: Corporate governance, Mergers and Acquisitions, Innovation, Technological regimes

* The authors wish to thank Alberto Chilosi, Slavo Radosevic and Paolo Polinori for their useful comments. The usual disclaimers apply. **Department of Economics, Finance and Statistics, University of Perugia, e-mail: [email protected] *** Department of Economics, Finance and Statistics, University of Perugia, e-mail:: : : : [email protected]

1 Introduction

The role of takeovers, which play an important function in well-performing corporate governance systems, has been attracting increasing interest in recent years.

Many studies, like the extensive survey by Shleifer and Vishny (1997), have stressed the crucial relevance of external threats by raiders on inducing greater loyalty from management and favouring an alignment of interests with their ‘principals’,

especially when dispersed ownership impedes direct monitoring of ‘agents’.

The question of the efficiency of “markets vs

corporations” in reallocating assets, changing industry structure and moving capital from declining industries to emerging ones is also a very important issue. In this perspective, Holmstrom and Kaplan (2001) point out that, in market-based systems, like those prevailing in the US and UK, mergers may play a powerful role in reallocating assets between sectors, particularly when some significant changes such as deregulation,


9

globalisation and information technology affect the economic landscape.

Moreover, in other studies (e.g. Mitchell and Mulherin, 1996), takeover activity is one of the channels that convey industry shocks (changes in technology, deregulation, shifts in demand and supply conditions). The above authors stress that ‘takeovers and related restructuring activity are the message bearers of the more fundamental changes facing an industry’ (Mitchell and Mulherin, 1996, p. 221).

Rossi and Volpin (2004), assuming again that mergers and acquisitions permit corporate assets to be directed toward their best possible use, also identify investor protection laws and other regulatory institutions as crucial determinants which can favour merger and takeover activities.

In sum, while the role played by corporate governance systems, technology shocks and institutional factors in triggering mergers and takeovers is widely acknowledged, the prevailing literature, as pointed out in many vast surveys on M&A (Jensen, 1988; Andrade et al., 2001; Harford, 2005; Tirole, 2006), does not consider the issue of the specific influence that technological regimes of innovation can exert in reallocating assets and moving capital among sectors.

In the present work we attempt to fill this gap by evaluating empirically not only the role of corporate governance systems and investor protection factors, but also the influence that alternative technological regimes of innovation may have on mergers and takeover activities. This will be done by taking into account the European experience of the last few years (2002-2005) which seems to mark a new wave in M&A activities.

The paper is structured as follows. After reviewing the basic causes and consequences shown in the literature on takeovers and discussing our hypothesis on technological regimes (Section 2), the main studies and evidences related to M&A in European economies until 2001 will be examined (Section 3), while in Section 4 the new comprehensive framework mentioned before will be adopted to explore the European experiences of the last few years (2002-2005). Finally, some conclusions will be provided (Section 5).

2 Takeovers: causes and consequences

Since the seminal paper of Manne (1965), Mergers and Acquisitions (M&A) have been considered no longer as ambiguous economic activities, in the past forbidden by antitrust laws, but as valuable strategies that improve corporate governance.

In this context, the market for corporate control works as a cheaper alternative to bankruptcy for many corporations: it enables quick transfer of assets from failing to successful firms. M&A also trigger mechanisms of external growth which are especially notable in rapidly expanding industries, whereas

internal-scale economies supporting growth require several years to act successfully. Lastly, the takeover scheme threatens badly managed firms, ensures competitive efficiency among corporate managers, and protects the interests of large numbers of small shareholders. In other words, Manne’s contribution outlines some basic rationales underlying takeover activities that were developed in the following years by a huge theoretical and empirical literature.

Indeed, a brief overview of the main determinants of takeovers should consider their role as a basic device aimed at reaching alternative aims: i) to increase the firm’s market power; ii) to increase its efficiency (through synergistic effects operating with economies of scale or scope); iii) to mitigate the impact of transaction costs (as in vertical integration); iv) to correct mismanagement.

Clearly, except for the first of these objectives, which involves the price-cost margin, the other rationales of takeovers are related to various classes of economic costs: productive, transaction, and agency costs. This rich variety of reasons has been tested in the study by Gugler at al. (2004), in which the authors attempted to identify market power effects, efficiency impacts, and the ‘pursuit of size’ in nearly two decades of M&A activities in the US.

It is also important to stress that each takeover wave, characterising the US experience, saw the prevalence of one of the four aims mentioned above1, but it was only in the 1980s that agency costs and the correction of mismanagement begin to represent the driving force of a new merger boom. Indeed, in a system where the free-riding behaviour of small shareholders was overwhelming and impeded direct monitoring of management, hostile takeovers represented the main mechanism for reducing corporate inefficiencies and mitigating agency costs2 (Holmstrom and Kaplan, 2001).

In terms of consequences, a brief overview of the empirical literature on profitability of takeovers (see Bruner, 2004) suggests a wide consensus on the

1 For instance, the greatest level of M&A in US history,

which took place around 1900, as documented by Golbe and White (1988), was mainly motivated by the basic intent to increase firms’ market power, a way of reaching the monopolistic position stigmatised by Stigler (1950). Instead, the episodes of the 1920s and 1960s started when the new antitrust regulations made horizontal mergers more difficult. These new waves were marked, respectively, by the role of vertical integration, which minimises transaction costs - as in the famous case study of Fisher Body and General Motors- and by the emergence of conglomerates, more closely aiming at exploiting economies of scope and synergies in financial costs. 2 However, in hostile bids a free-riding problem does emerge. As Grossman and Hart (1980) write: “It is commonly thought that a widely held corporation that is not being run in the interest of its shareholders will be vulnerable to a takeover bid. We show that this is false, since shareholders can free ride on the raider's improvement of the corporation, thereby seriously limiting the raider's profit” (1980, p. 42).


10

overall wealth effect for each merger wave: target shareholders reap positive returns. On the contrary, the bidder shareholders get mixed effects: some obtain small positive premiums, while some others suffer some losses. We must be cautious about the interpretation of these findings which are still controversial, but from empirical evidence a new perspective seems to emerge. The small premium or losses of the bidder shareholders would represent a new symptom that mergers come about not only as correction of management failures, but also as the outcomes of new managerial strategies. In fact, executives choose to buy target firms in order to increase their power and to divert free cash flow, instead of returning it to investors. It is remarkable that in the 1980s, 45% of takeovers were financed by cash. This finding does not contrast with the claim that strategies adopted to enlarge managerial empires may also cause a ‘wealth destruction on a massive scale’, a thesis well documented in many studies, as in Moeller et al. (2005).

It must be re-emphasised that most of the literature signals that M&A come in waves. Furthermore, each wave seems to be triggered by some specific industry shocks. In this perspective, new contributions, starting from Mitchell and Mulherin (1996), suggest that ‘a fruitful line of research design would consider the joint effect of macroeconomic and industry-level factors in modelling the behaviour of takeovers over time’ (Mitchell and Mulherin, 1996, p. 195).

In this new field of research, which reconsiders the clustering in time of M&A and the role of industry shocks, three different perspectives seem to prevail.

The first one, a neoclassical approach, suggests that economic and technological shocks occurring at industry level cause a high degree of dispersion of the firms’ opportunities. The different Q-ratios achieved by the different firms induce the acquisition of the bad performers by the more successful ones, thus promoting an efficient selection (Jovanovic and Rousseau, 2001, 2002).

A second group of models develops a behavioural motivation by focusing on agency problems, with herding and hubris at the centre of the stage, as in Roll (1986). In this perspective, it is hypothesised that the first successful bids encourage managers of other companies to imitate the beginners. These strategies devoted to increase the firm’s size are therefore motivated by the basic aim to maintain power, which induces inefficiencies and misallocation of corporate resources.

A third group of studies signals the role of financial failures and addresses the attention to mispricing occurring in bull markets. In these episodes, management of the over-evaluated firms use their equities to buy the undervalued assets of the other ones, thus taking advantage from mispricing. At the same time, target management reaps gains by simply maximising their short run benefits. A good reason for target management’s consent, as shown in Shleifer and

Vishny (2003, p. 303), “is that the acquirer pays them for it”, for instance, “through the acceleration in the exercise of stock options (which could be very valuable if the target is overvalued)”, or “even by keeping the managers of the target in top positions (as was done, for example, in the AOL acquisition of Time Warner)”. This means that “a merger requires a coincidence of short-term objectives of the target managers with longer-run objectives of the bidders”(Shleifer and Vishny, 2003, p. 307).

In sum, as easily verified from this brief excursus, theories and evidence both suggest that takeovers represent not only an efficient way of correcting agency problems, but also manifest agency problems themselves. It is also widely admitted that industry shocks are important driving forces triggering the various merger waves.

One expected hypothesis of our analysis is that countries and sectoral patterns may also be explained by the main differences that characterise sectors in terms of innovation processes. Thus the existence of two distinct regimes, the entrepreneurial and routinised sectors, may have a significant impact on mergers.

The first regime, type of SMI, reveals the lower stability of the hierarchy of innovators, a lower concentration ratio of the more innovative firms, and a higher proportion of new innovators with respect to the old ones (Audretsch, 1996; Dosi, 1988; Malerba and Orsenigo, 1993, 1996).3

In this more turbulent environment, mergers and acquisitions are expected to be more frequent, and may be an efficient way of growing and obtaining synergies in R&D expenditure. The opposite may be true for the other, routinesed, regime, SMII. Here, innovations are incremental along existing technological trajectories and a less frequent reallocation process by acquisitions of other firms may be expected since the newly hired workforce has to spend time and effort in order to operate efficiently in specialised routines.

In the present study, one important point to be explored is whether innovation activities characterising the different technological regimes are driving factors capable of explaining the occurrence of M&A.

The challenging theme of knowledge and innovation-enhancing strategies represented by corporate acquisitions has been the focus of some recent studies. The main intent of this literature is to inquire if the innovation performance of acquiring firms is influenced not only by the technological base (measured in absolute and relative terms) of the companies involved, but also by the degree of relatedness of those knowledge bases (Cassiman and Colombo 2006).

3 Other main references are Nelson and Winter (1982), Kamien and Schwartz (1982), Malerba and Orsenigo (1997) and Breschi et al. (2000).


11

In our empirical research, more than asking what can happen ex-post, in terms of innovation capabilities, we inquire what ex ante is the actual propensity to merge associated with various innovation regimes. For instance, for chemicals, a sector characterized by large firms, continuity in innovative processes, and cumulativeness of firms’ capabilities, a lower incidence of M&A is expected. Instead, in sectors characterized by creative destruction, R&D processes and efforts are probably less serious obstacles for corporate acquisitions.

In sum, in our perspective, what is explored is the overall question of whether the creative accumulation

regime (SMII) results in the lower probability of occurrence of M&A. Here innovations are incremental along existing technological trajectory, and for this regime one can expect a less frequent reallocation process by acquisitions of other firms, since newly hired workforce has to spend time and effort to operate efficiently in specialised routines.

3. The European experience: M&A activity in the 1990s

It is well known that Continental Europe is

characterised by more concentrated ownership structures (Barca and Becht, 2001; Faccio and Lang, 2002), weaker investor protection, and less highly developed capital markets (La Porta et al., 1998), as analysed in Chilosi and Damiani (2007). In the

European scenario, there is therefore less space open to the market for corporate control, as can be seen in the remarkable gap in terms of M&A activities between the European and the Anglo-Saxon economies (see Figure 1).

The incidence of hostility, which signals the role of takeovers as a governance mechanism, is also remarkably lower in Continental Europe (see Figure 2). However, the last few years have witnessed some convergence in corporate governance systems and in Europe an increasing number of takeovers was recorded in the years 1993-2001, a period which is considered as that of the fifth wave of mergers and acquisitions. An examination of takeover activities in the 1990s, well documented in the study of Martynova and Renneboog (2006), highlights not uniform diffusion in all countries, but significant differences in geographical patterns. The resulting divergences are even greater when compared with the GDP weights of each single country. This is shown in Figure 3: the distribution of M&A activity in the 1990s is correlated not only to the total size of national economies, as measured by GDP, but also to some other factors. Additional causes are required to explain why some countries, such as Germany, France and Italy, account for a lower number of operations than those expected from the size of their economies. Italy, for instance, showed an M&A share of 6.4% but 12.7% weight in terms of GDP.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

M&A US (number of deals) M&A EU14 (number of deals) UK Total

Figure 1. M&A in EU and US: 1989-1999

Source: Thomson Financial (see Becht et al. 2003)

0

2

4

6

8

10

12

14

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Hostile Bids (%) US Hostile Bids (%) EU14 UK(%)

Figure 2. The market for corporate control: incidence of hostile bids in EU and US

Source: Becht et al. (2003)


12

Figure 3. Distribution of M&A activity and GDP between EU Member States, 1991-1999

Source: European Commission (2001) Additional information can be gathered by simply

splitting all the M&A into domestic and cross-border deals. A significant fraction, nearly one-third of the intra-European M&A of the period 1993-2001, is represented by cross-border deals. However, some important distinctions affect the patterns observed in the various countries. For instance, British and French firms were net acquirers in the European market for

corporate control, while German and Italian firms were mainly targets of cross-border deals. Moreover, Eastern European firms, except the Hungarian ones, were all ‘prey’ to intra-European M&A deals. It is shown by Figure 4, where the difference between number of all takeover announcements of each country as a bidder or target is reported (see Martynova and Renneboog, 2006, Table 4).

-40

-20

0

20

40

60

Aus

tria

Bel

gium

Bul

gari

a

Cro

atia

Cze

ch R

ep.

Den

mar

k

Finl

and

Fran

ce

Ger

man

y

Hun

gary

Irel

and

Ital

y

Net

herl

ands

Nor

way

Pola

nd

Rom

ania

Rus

sia

Slov

enia

Spai

n

Swed

en UK

Figure 4. Net bidders or net targets? Europe: 1993-2001 Source: Thomson Financial (see Martynova and Renneboog, 2006)

The prevailing literature also focuses on investor

protection factors and ownership structures as the main determinants of divergences between countries, but it does not analyse whether the different patterns of reorganisation pertaining to different sectors affect the market for corporate control.

A promising line of research, as shown below, should consider this new perspective, thus also focusing on sectoral factors and technological regimes.

4 M&A in Europe in the new merger wave (2002-2005)

4.1 Sample selection and data description Our sample of mergers and acquisitions comes from the Datastream database; additional information was taken from Lexis and Nexis. The sample comprises eight countries, whose activities represent nearly 80% of the European market for corporate control:

Source: European Economy 2001

0,0%

5,0%

10,0%

15,0%

20,0%

25,0%

30,0%

35,0%

Belgium

Den

mark

Ger

man

y

Gre

ece

Spain

Franc

e

Ireland

Italy

Luxe

mbo

urg

Net

herla

nds

Austri

a

Portu

gal

Finla

nd

Swed

en

Uni

ted

Kingd

om

Share of EU

M&A activity

(%)Share of EU

GDP (%)


13

Denmark, France, Finland, Germany, Italy, Norway, Sweden and the United Kingdom. Data on M&A were collected from the “Capital Issues and Changes” Report of Datastream; only transactions related to takeovers and involving changes in corporate control were selected. Completed and not completed takeovers, and financial and non-financial sectors were included in our sample, and total 802 deals, for the period 2002-2005.

Comparative analysis by country and sector was performed. First, we aggregated 39 four-digit sectors in 10 two-digit industries provided by Datastream4. This classification follows the Industry Benchmark Classification (IBC), a system for listed companies managed by FTSE and Dow Jones Indexes (2004).

The Datastream data by sectors (DS) were then classified and grouped into two Technological Regimes (TR): SMI and SMII. This last classification, as seen in Malerba and Orsenigo (1993, 1996, 1997), concerns the various innovation activities which are in turn described by means of patent applications. For this reason, the traditional technological classes, stemming from the International Patent Classification (IPC), are the statistical units used to classify industries within the TR context. In particular, the four-digit IPC sub-classes were adapted by Malerba and Orsenigo (1996) to obtain 49 technological classes.

Unfortunately, an official concordance table to match the two classification systems (IPC and IBC) does not exist. Nevertheless, following Van Dijk (2000), we used a concordance criterion in order to convert the traditional sector classification into Technological Regimes (Table 1).

The data obtained at aggregate level for each country are examined in Section 4.2, and an analysis by sector and technological regime is performed in Section 4.35.

4.2 The last M&A wave: varying patterns in some European countries

A convenient starting point is a comparison of the number of deals obtained in our study with the figures recorded in the previous M&A wave, fully explored in Martynova and Renneboog (2006), one of the main contributions for the European context already mentioned. Both samples refer to transactions involving changes in corporate control, but in Martynova and Renneboog only domestic and intra-

4 For lack of space, we do not show the table concerning this aggregation. It is available upon request from the authors. 5Note that, in our analysis, computers and telecommunications are separated, respectively, into Hardware (SMII) and Software Computer Industry (SMI) and Fixed (SMII) and Mobile Telecommunications (SMI), whereas these distinctions were not operated in the original study by Malerba and Orsenigo (1996). Our reclassification does not contradict new methodological refinements (see Corrocher et al., 2007).

European cross-border deals were taken into account, while our data set also includes the extra-European acquisitions.

Table 2 offers some interesting information; in particular, it displays minor changes in the distribution of M&A activity between the two periods 1993-2001 and 2002-2005, as shown by the ranking orders reported in brackets. The UK still comes top, followed at some distance by France and Germany. The Italian market for corporate control looks more active than in the past, thus reaching the ranking position that Sweden had obtained in the previous years. A small increase also affected the share of deals of Denmark.

A plausible explanation might indicate legislative changes occurring in Italy and Denmark, where in the last few years the threshold of mandatory bid rule has been lowered6

, thus increasing the ‘exit’ rights of minority shareholders. In any case, this issue is still controversial, as argued in Goergen et al. (2005). Indeed, on one hand, the mandatory bid rule mitigates the problem of expropriation of small owners; on the other, it makes transactions more expensive, since it forces the bidder to offer the same price to all shareholders, thus reducing the volume of trade in controlling stocks. Moreover, in a block holder system such as that prevailing in Italy, it is possible that domestic M&A were used as anti-takeover devices.

6 The mandatory bid rule provides minority shareholders with an opportunity to exit the company on fair terms. This rule requires that the acquirer must make a tender offer to all shareholders, once it has accumulated a certain percentage of their shares. The new norms in Italy and Denmark dictate that a tender offer needs to be made to all remaining shareholders as soon as the bidder has accumulated one-third of the company’s equities.


14

Table 1. Technological Regimes and Industries

SMI SMII Other Manufacturing

Food producers Aerospace & Defence Mining Beverages Oil and Gas Producers Forestry & Paper Tobacco Oil Equipment & Services General Industrials

Household Goods Chemicals Automobiles & Parts Personal Goods Pharmaceuticals, Biotechnology Support Services

Construction & Materials Healthcare Equipment Industrial Engineering Electronic, Electrical Equipment

Industrial Metals Fixed Line Telecomm. Industrial Transportation Technology Hardware & Equipment

Leisure Goods Mobile Telecommunications

Software & Computer Services Source: Our elaborations from FTSE and Dow Jones Indexes (2004) and from Malerba and Orsenigo (1996).

Table 2. Mergers and Acquisitions by target country in the last decade

Countries Number M&A

1993-2001 % M&A

1993-2001 Number M&A

2002-2005 % M&A 2002-

2005

United Kingdom 932 47.53 (1) 475 59.23 (1) France 308 15.71 (2) 106 13.22 (2)

Germany 269 13.72 (3) 57 7.11 (3) Sweden 150 7.65 (4) 43 5.36 (5) Norway 95 4.84 (5) 40 4.99 (6)

Italy 83 4.23 (6) 45 5.61 (4) Finland 73 3.72 (7) 11 1.37 (8)

Denmark 51 2.60 (8) 25 3.12 (7) Total 1,961 100.00 802 100.00

Sources: 1993-2001 Thomson Financial SDC; see Martynova and Renneboog (2006), 2002- 2005, our elaborations of Datastream and Lexis-Nexis data; ranking shown in brackets.

A second point to be raised is the trend observed

in the sample period which follows the stock market collapse of 2001. From 2000 (the year of the peak of the fifth European takeover wave) to 2002, M&A decreased in Europe by about 50% in volume and 70% in value (McCahery et al., 2004, p.601). Since 2002, almost all the eight countries considered in our sample have recorded a weak and unstable increase in their activities.

In 2005, a remarkable increase emerges (see Figure 5), particularly in the UK, followed at a remarkable

distance by France. Recent information on M&A suggests a new peak in 2006, since many takeovers announced but still not completed have taken place in the last few years.

Table 3 offers a better comparison of geographical patterns: for each country, the absolute number of deals was standardised on the total number of firms included in the Datastream sample. Then a binomial test was performed to evaluate the significance of the different proportions shown for each couple of countries (Table 4).


15

0

20

40

60

80

100

120

140

160

180

2002 2003 2004 2005

No.

of de

als

UK

FR

GE

IT

DK

SW

NW

FN

Figure 5. Mergers and Acquisitions by country and year Source: Datastream

Table 3. Mergers and Acquisitions by target country (2002-2005)

Absolute and relative frequencies of M&A

Countries Number M&A Total number of firms M&A / Total

firms (%)

United Kingdom (UK) 475 2,719 17.47 France (FR) 106 1,026 10.33 Germany (GE) 57 1,242 4.59 Italy (IT) 45 312 14.42 Sweden (SW) 43 436 9.86 Denmark (DK) 25 200 12.50 Norway (NW) 40 318 12.58 Finland (FN) 11 145 7.59 Total 802 6,398 12.54

Source: Datastream

Table 4. Difference between proportions and statistical significance of M&A in 2002-2005

UK FR GE IT SW DK NW UK FR 7.14* GE 12.88* 5.74* IT 3.05* -4.09* -9.83*

SW 7.61* 0.47 -5.27* 4.56*

DK 4.97* -2.17** -7.91* 1.92 -2.64

NW 4.89* -2.25** -7.99* 1.84 -2.72*** -0.08

FN 9.88* 2.74* -3.00* 6.84* 2.28*** 4.92** 4.99* Note: Reported figures are differences between column country proportion and row country proportion. Difference proportion test is based on binomial test. * Significant at 1% level ** Significant at 5% level *** Significant at 10% level

Each column in Table 4 reports the difference of each country with respect to the others (which enter distinct rows). For instance, the first column shows the positive and significant differences in the UK with respect to all the other countries, whereas the third column shows the negative and significant disparities in Germany.

This analysis confirms that the UK was the most active player in the market for corporate control. The

standardised figures show that Germany is at the bottom: only 4.59% of national companies experienced a merger in the period considered (see Table 4). In an intermediate position is a first group of countries with a relative high proportion of M&A: Italy (14.42%), Norway (12.58%) and Denmark (12.50%), whose differences are not statistically significant (see Table 3). Below them come France (10.33%) and Sweden (9.86%), which share very


16

similar takeover activity (the difference in proportion is not significant). Lastly, Finland has a low proportion of M&A (7.59%), but in any case significantly higher than that of Germany.

A preliminary interpretation of the low figures recorded in this last country should take into account the concentrated ownership structure and the role played by banks and employees, particularly important in the Renanian system. The monitoring role of insiders and their presence in the supervisory board should work as a corporate governance device, a well-functioning mechanism and a substitute for hostile bids. Instead, the dispersed ownership structure prevailing in the UK, requires the discipline exerted by raiders, thus explaining the high incidence of takeovers. However, a wider examination should focus on additional institutional features, which act as complementary factors for well-designed corporate governance. Some of these determinants, fully explored in La Porta et al. (1998) and Manchin (2004), are listed in Table 5.

In order to examine these findings, some considerations must be made. The main benefit of concentrated ownership is the possibility of overcoming the free-riding problem of dispersed ownership. Large investors are able and motivated to exercise control, since they have enough power and can obtain significant gains by monitoring activity. When cash flow rights and voting powers are aligned, large owners are in a position to deal effectively with asymmetric information ex-ante and to exert a strong bargaining power in the division of rents ex-post. However, large investors may adopt personal strategies at the expense of minority shareholders’ interests. As noted by Shleifer and Vishny (1997), large investors may obtain special dividends or use other devices, such as green mail or targeted share repurchase, to their benefit. The expropriation activity of controlling shareholders may discourage investors and obstruct hostile bids. In addition, legal shareholder protection may mitigate failures coming from ‘private benefits of control’, and better investor protection is correlated with a more active market for mergers and acquisitions, as shown in Rossi and Volpin (2004). Theories and evidence provide opposite conclusions: i) the German model of corporate governance performs well and does not need the discipline of hostile bids; ii) in such a system, the combined effects of ownership structure

and lower investor protection is a serious obstacle to the firm’s contestability.

This issue needs additional exploration, and each country tells a different story. For instance, in Italy, there is a weak governance regime, notwithstanding the legal reforms and improvements of the last few years, but there is a higher incidence of takeovers than that observed in Germany. Conversely, in the UK, good protection for minority shareholders makes control more contestable. How can these results be interpreted? By efficient deals, or hubris and managerial self-interest strategies? The evidence for the UK examined by Franks and Mayer (1996) shows that the market for corporate control ‘does not function as a disciplinary device for poorly performing companies. Rejection of bids is more consistent with opposition to the anticipated redeployment of assets by the bidder and negotiation over the terms of bids’ (Franks and Mayer, 1996, p.180).

A wide spectrum of factors, covering the structure of ownership, shareholders’ rights, the quality of the legal system, and takeover regulation may all play a significant role. Table 5 shows some selected indicators for the eight European countries.

The first columns of Table 5 show that there is distinct segmentation between ownership patterns in Continental Europe and the UK7, revealed by comparing the average ownership of the three largest shareholders, as shown by La Porta et al. (1998) for the ten largest non-financial firms of each country8. The lowest concentration figures are in the UK, and the highest in Italy and Germany. Similar results were obtained by Faccio and Lang (2002) for a different dataset, which included medium and small financial and non-financial companies and which referred to a sample of 5,232 European corporations. That study showed that widely held firms have the highest incidence in the UK (63% of firms) and the lowest in Germany (10.37%).

What about the identity of the ultimate owner? Faccio and Lang (2002) showed that, on average, a large proportion (around 44.3%) of Western European firms are family controlled (at the 20% threshold of voting rights) but, even so, the country diversities are remarkable: the phenomenon of family control is

7 See main studies by La Porta et al. (1998), Barca and Becht (2001) and Faccio and Lang (2002). 8 The study by La Porta et al. (1998) refers to a larger dataset which includes 49 countries.


17

lowest in the UK (around 24%) and Norway (39%) and highest in France and Germany (around 65%).

Table 5. Ownership, shareholders’ protection and legal system in eight European countries

Legenda: Ownership concentration: average percentage of common shares owned by three largest shareholders in 10 largest non-financial, privately owned, domestic firms of a given country (La Porta et al. 1998, Tab.7); Widely held firms: percentage of companies which do not have shareholders controlling, at least 20% of votes, in a sample of 5,232 publicly traded financial and non-financial corporations (Faccio and Lang 2002, Tab. 3); Anti-director rights: revised index by Djankov et al. (2008); see our note 8; Family control: fraction of firms which are majority controlled by wealthy families out of ten largest conglomerates (Fogel, 2006, Tab. 1); Rule of law: measures quality of legal system and takes into account judicial independence, impartiality of courts, protection of intellectual property, military interference in rule of law, and integrity of legal system (see Manchin, 2004).

New similarities and divergences across countries are shown in a recent research conducted by Fogel (2006), focusing on ownership of the largest ten conglomerates; it offers precious information, since oligarchic family groups can tunnel wealth between the firms under their control and generate great private benefit and resource misallocation, as reviewed in Morck, Wolfenzon and Yeung (2005); thus family groups may represent a serious obstacle to efficient corporate restructurings obtained via M&A deals. Fogel’s study offers a controversial finding: in Germany, control by wealthy families is lower than in the UK (and similar to the US, where families retain small, non-controlling stakes in public companies). This result contrasts with the evidence from France and Italy, since four and six groups, respectively, in these countries, are majority controlled by wealthy families (see column 4 of Table 5).

Investor protection laws and other countries’ regulatory institutions may be crucial determinants in explaining why firms are owned and financed so differently in different countries, as argued in La Porta et al. (1998). The authors make considerable efforts to elaborate accurate indicators for shareholders’ rights. A measure of the strength of minority shareholder protection is shown by the so-called anti-director rights, which considers, among other things, important areas such as voting rights in shareholders’ meetings, representation on the board of directors, and preemptive rights to subscribe to new

securities issued by the company.9 In Djankov et al. (2008), the original index for investor protection has recently revised: the new measure elaborated by the authors for 72 countries (revised anti-director rights), distinguishes better between enabling rules and mandatory or default provisions.10 The revised index (see Table 5, column 5) is also based on laws and regulations updated to May 2003 and is more useful for our purposes. Table 5 clearly shows that the index of anti-director rights is higher in the UK, and good ranking is obtained by Denmark.

Good ranking of Nordic European countries also arises according to a measure of the quality of the legal system, like that shown in the last column of Table 5. This indicator, elaborated in a European Commission study, “measures the quality of the legal system and takes into account judicial independence, impartiality of courts, protection of intellectual

9 More precisely, the index covers six areas: i) vote by mail; ii) obstacles to actual exercise of the right to vote (i.e., the requirement that shares be deposited before the shareholders’ meeting); iii) minority representation on the board of directors through cumulative voting or proportional representation; iv) an oppressed minority mechanism to seek redress in case of expropriation; v) preemptive rights to subscribe to new securities issued by the company; and vi) the right to call special shareholder meetings (see Djankov et al., 2008). 10 The authors also propose new indexes of the strength of minority shareholder protection against self-dealing by the controlling block-holder (anti-self-dealing index) for a group of 72 countries. For a methodological explanation of these new indicators, see Djankov et al. (2008, Tab. I).

Country Ownership

concentration

Widely held firms

Family Control

(at 20% threshold)

Family Control in ten largest

groups

Anti-director rights

(revised index)

Rule of

law

(1) (2) (3) (4) (5) (6) United

Kingdom 19 63.08 23.68 20 5.0 8.80

France 34 14.00 64.82 40 3.5 7.66 Germany 48 10.37 64.62 10 3.5 8.95

Italy 58 12.98 59.61 50 2.0 7.10 Sweden 28 39.18 46.94 60 3.5 8.78

Denmark 45 n.a. n.a. 10 4.0 9.08 Norway 36 36.77 38.55 50 3.5 8.86 Finland 37 28.68 48.84 30 3.5 9.16 Average 38 29.30 49.58 32 3.56 8.55


18

property, military interference in the rule of law, and the political process and integrity of the legal system” (Manchin, 2004 p.15). On the basis of this measure, as already mentioned, the superior quality of the Nordic countries’ legal system can be ascertained and, for these economies, it may be expected that the concentrated ownership structure does not obstruct a well-functioning market for corporate control.

In addition, takeover regulation may be an important determinant of transfers of control, since it affects the costs and benefits of these transfers, as extensively analysed by Bebchuk (1994). For instance, legal provision of squeeze-out rights solves free-riding problems caused by dispersed possession, which are more important in widely held firms. Indeed, each individual shareholder, anticipating that the post-takeover share price will exceed the offered price, prefers not to tender. The squeeze-out rule, giving the controlling shareholder the right to force minority shareholders to sell their shares, solves free-riding problems and thus allows raiders to make value-increasing acquisitions (Burkart and Panunzi, 2004).

EU Directive 2004/25 recently intended to harmonise EU member states’ takeover regulation, and the debate as to whether uniform national legislation produces identical effects in countries with heterogeneous corporate governance regimes is still ongoing (Goergen, Martynova and Renneboog, 2005). In any case, it should be noted that some of the member states of our database, which covers the period 2002-2005, only brought the provisions of the Directive into force in 2006, and that the Directive leaves some discretionality to national legislators.

Table 6 lists six different measures, as evaluated for the reference period 2002-2005: i) mandatory bid rule; ii) and iii) squeeze-out and sell-out rules; iv) ownership and control transparency; v) passivity rule in terms of board neutrality with respect to anti-takeover defences; vi) break-through rule.11

In a heterogeneous legislative environment like that typical of European countries, it is important to compare the different takeover rules, like those shown in Table 6. At first sight, a general evaluation of Table 6 suggests that the Nordic countries have lower ‘exit’ opportunities for minority shareholders, at least according to their takeover regulation. An instance is the mandatory bid rule, which obliges a bidder, once it has accumulated a certain percentage of stocks, to make a tender offer to all shareholders at a fair price. In Nordic countries, particularly in Finland, the percentage of shares that makes the tender offer compulsory to all shareholders was higher

However, the potential impact of this provision should be evaluated according to other governance features and, in this perspective, some trade-offs arise. On one hand, “the mandatory bid requirement may

11Additional information on the criteria adopted to obtain the index of takeover regulation used here is available upon request.

mitigate the problem of expropriation of minority shareholders”; on the other hand, “it also decreases the likelihood of value-creating restructuring. The main reason for this is that the rule makes control transactions more expensive and thereby discourages bidders from making a bid….” (Goergen et al. 2005, p.11).

An additional trade-off concerns the lower incentive for listing on the stock market which a more restrictive regulation might produce as an unintended result. This is a serious drawback for those economies, such as the Italian one, with an under-developed stock market and where too few companies go public.

Additional information was obtained by considering cross-border deals and their incidence on total M&A activities, as shown in Figure 6.

Two lines divide the diagram into four quadrants; the intercept of the horizontal line corresponds to the average value of total M&A activity (12.54 % of firms were targeted in the period 2002-2005); the vertical line corresponds to the average proportion of cross-border deals (on average, 20.37% of targeted firms were acquired by foreign investors). However, some caution should be invoked, since, in terms of M&A percentages, the proximity of Italy to the UK is biased by the low number of Italian listed companies, and it is worth noting that, in absolute terms, the number of Italian and UK target firms was, respectively, 45 and 475!

The percentage values, in any case, show that Italy is located in the bottom-right quadrant and reveals a high incidence of M&A activity, but a low fraction of cross-border deals. At the opposite, we find two Nordic countries (Finland and Sweden), open to foreign deals, but showing a lower incidence of total M&A.

A further study could focus on a comparison between Italy and Sweden, two countries whose absolute figures of M&A deals do not reveal a large gap over the period 2002-2005: 45 firms were targeted in Italy, whereas 43 mergers occurred in Sweden.

A starting point for a comparison between the two economies is evaluation of their ownership structure. Italy shows higher concentration, broader diffusion of family control (see Table 5, column 3), and the adoption of pyramids as a common control device12. However, Sweden features a higher proportion of conglomerates controlled by wealthy families, as seen in Table 5, column 4.

12 As Bianchi, Bianco and Enriques (2001, p.161) observe “… pyramidal groups have been favoured by a neutral tax policy (i.e. dividends are taxed only once, no matter how many levels the control chain has) and by the absence of any legal provisions to prevent conflicts of interest between the controlling agent and minority shareholders in the subsidiaries”.


19

Table 6. Takeover regulation - legislative framework in eight countries in period 2002-2005

Mandatory

bid rule (a)

Squeeze out rule (% of

stocks) (b)

Sell out rule

(% of stocks)

(c)

Transparency (% of stocks)

(d)

Break through

rule ( e)

Passivity rule ( f)

UK 30 90 90 3 no yes FR 33.33 95 95 5 no yes GE 30 95 95 5 no yes IT 30 98 90 2 yes yes

SW 40 90 90 5 no yes DK 33.33 90 90 5 no yes NW 40 90 90 5 no yes FN 66.67 90 90 5 no yes

Legenda: a) percentage of shares that makes tender offer compulsory to all shareholders; b) percentage of equities that gives controlling shareholder the right to force minority shareholders to sell their shares; c) threshold above which remaining shareholders have the right to sell their shares at a fair price; d) threshold above which ownership of voting rights must be disclosed; e) rule that permits a bidder to break through existing voting arrangement and to exercise control, as in a context of one share-one vote; f) rule that requires board neutrality in case of anti-takeover measures and that makes approval by shareholders’ meeting compulsory. (For legislative sources, see Appendix).

0

5

10

15

20

25

30

35

0 2 4 6 8 10 12 14 16 18 20M&As Activity

Cro

ss-B

orde

r M

&A

DK

UK

IT

FN

SWFR

GE

Closed market withhigh M&A activity

Closed market withlow M&A activity

Open market withlow M&A activity

NW

Open market withhigh M&A activity

Figure 6. Market for corporate control and cross-border M&A

Additionally, takeover regulation in Italy appears

to be more oriented at protecting minority shareholders’ interests (see Table 6), and it was adopted even before the new EU Directive was passed in 2004. In terms of investor protection, the Italian legal framework was undergone major improvements: between 1990 and 2005, significant strengthening of internal governance was enacted with various reforms, such as the Draghi Law (1998), the new Company Law (2004) and the Law on Savings (2005). All these reforms, as detailed examined by Enriques and Volpin (2007), have empowered small shareholders. For instance, legal innovations include: i) the introduction of the mandatory provision that at least one director and one board of auditor-members must be elected by minority owners; ii) the increasing

disclosure and procedural requirements on related-party transactions; iii) the ban of voting caps; iv) the limits for pacts among block-holders. But what is remarkable is that an indicator, as Rule of Law which captures the quality of the legal system, as shown in Table 5 column 6, still reveals Italy’s poor performance with respect to Sweden. The same problem of related-party transactions, which in Italy are at the origin of many common self-dealing deals, was faced by insufficient private enforcement of legal provisions (Enriques and Volpin, 2007, p. 138).

In sum, in principle, according to legislative innovations, Italian companies should have become more market-oriented and more exposed to the market for corporate control. But most of the changes concern only the instruments used to exert control,


20

which remains stable and concentrated in the hands of wealthy families, at least until now. For instance, the lower concentration of ownership and the less extensive use of pyramidal structures have been accompanied by a higher percentage of companies

controlled through coalitions. Paradoxically, the extensive reforms undertaken between 1990 and 2005, were followed by strengthening of family coalitions, as shown in Figure 7.

0

5

10

15

20

25

30

35

Coalition of

listed firms

Cooperatives Familiy

coalition

Family

coalition with

other

members

Total

1990

2001

2005

Figure 7. Listed companies controlled by coalitions: Italy, 1990, 2001, 2005

Source: Bianchi and Bianco (2006, Tab.16)

These findings explain why the market for corporate control in Italy is still confined within its national borders. Analogous evidence was gathered for non-listed companies, since, over the last ten years, half of the changes in control (amounting to 3% every year) occurred within the family. These changes reveal not ‘actual contestability’, but only “infra-generational transfers (possibly inducing the inefficiencies that the literature suggests are associated with 2nd, 3rd and so forth generation family control)”(Bianchi and Bianco, 2006, p. 5).

In Sweden, the corporate governance system has been described as “promoting strong private owners with a long term investment horizon and a far reaching social responsibility towards employees and society in general” (Agnblad et al., 2001, p.251). A distinctive feature of its market for corporate control is that the bidder retains a substantial long-term toehold, and its voting power is on average 31%. This may explain why corporate acquisitions are mainly friendly and negotiated between the two parties (Agnblad et al. 2001, p. 247).

Some concluding remarks, to compare Italy and Sweden better, concern a disaggregate analysis by sector. Italy shows high concentrations of M&A in the financial sector, since 25% of total deals involve banks. In this sector, the privatisation process of the last few years, besides the need to face stronger international competition, probably triggered a M&A wave, but it is remarkable that, in the period 2002-2005, no cross-border deals can be detected on the eleven deals which targeted banking companies. One potential explanation is the attitude of the national authorities, more oriented to the protection of domestic companies and not encouraging foreign acquisitions of national ‘champions’.

Instead, Sweden is not only open to foreign deals, but shows a wider sectoral diffusion of the market for corporate control. But sectoral aspects are a matter for exploration in the next section.

4.3 The last M&A wave: patterns by sector and the role of technological regimes

The structural barriers to takeover activities may have a complementary explanation. Technological factors in different industries may constitute another determinant of takeovers. As already seen, there is a research line in this field, which maintains that industries also matter in driving the occurrence of M&A (Mitchell and Mulherin, 1996; Andrade and Stafford, 2004). In particular, some authors have stressed that once a technological, regulatory or economic shock to an industry’s environment occurs, industry assets can be reallocated through mergers (Harford, 2005). Sometimes technology itself boosts institutional and regulatory changes (Jovanovic and Rousseau, 2001, 2002). In addition, when radical innovations such as microprocessors and computers emerged thirty years ago, deregulation and friendly antitrust policies encouraged mergers in telecommunications, airlines and other hi-tech sectors (Jovanovic and Rousseau, 2001).

In our analysis, as already mentioned, we group sectors by Technological Regime (TR). As discussed in Section 2, TR may constitute a context in which systematic differences in takeover frequencies can be observed. In particular, the specific knowledge-based system characterising the SMII regime, centred on higher investments on Research and Development, may raise structural barriers and limit the market for corporate control.


21

Table 7 shows the sectoral dimension of our Datastream sample for the eight pooled countries. Data by industry, provided by IBC in Datastream, are also grouped by technological regime: SMI and SMII13.

First, the performance of the Telecommunications sector over the period 2002-2005 is worth noting. Despite the low number of deals and the high concentration at industry level, the highest relative frequency of takeovers occurred in this sector. Indeed, a first round of consolidation in this sector occurred in the 1990s and was driven by the need to compete with American providers (OECD, 2001). Probably, the rapid technological change of the last few years, coupled with drastic regulatory reforms and the need to combine content and delivery, drove a further surge of mergers.

Takeover activity in Telecommunications was been notable not only in the number of deals, but also in value: the share market fluctuated from 7% to 11% (Thomson Financial 2002; 2003; 2004; 2005). In 2003, Olivetti’s acquisition of the remaining 46% interest in Telecom Italia for $28 billion was by far the largest deal in Europe. And Telefonica’s planned tender offer for O2 (UK), for $31.8 billion, was the second largest deal announced for 2005 (Thomson Financial 2003; 2005).

Consumer Services and Utilities were the second and third most active markets for corporate control: 18.87% and 15.20% of companies were targets of M&A, respectively. The growing importance of service sectors in advanced economies, combined with the success of Information and Communication Technologies, the introduction of the Euro, and the interest of manufacturing firms also for services such as retail and wholesale trade, probably boosted merger activity in these industries. Unlike Utilities, deals in Consumer Services were larger in terms of number (177 was the largest absolute number of deals occurred in industries for the period 2002-2005) but not in value (Thomson Financial 2002, 2003, 2004, 2005).

However, manufacturing sectors also have played an important role. The relative frequencies of takeovers in Industrial and Consumer Goods were slightly below the sample average: 12.49% and 12.34%, respectively (see Table 7). In terms of value, for 2002-2005 Thomson Financial estimated a share market ranging from 7% to 11% for Industrial and from 3% to 6% for Consumer Goods.

It is worth noting that in the 1990s, a considerable number of deals in industries such as plastics, metals, machinery, food, textiles and chemicals was detected by the European Commission (2001). In particular, this number of deals was more stable in the last decade, responding less to the evolution of the economic cycle, in both its upswings and downturns (European Commission, 2001).

13 Seven manufacturing industries are included in SMI, SMII and the residual Other Manufacturers. Oil&Gas, Basic Materials, Industrials, Consumer Goods, Healthcare, Telecommunications and Technology.

These last considerations appear to support our view that structural characteristics related to sectors may affect takeover activity. Indeed, if we reorganise manufacturing sectors by technological regime (SMI and SMII), a considerable difference emerges.

The last three rows of Table 7 show that the relative frequency of takeovers within manufacturing sectors included in SMI (13.10%) was significantly higher than in SMII (9.77%).


22

Table 7. M&A in 8 countries analysed by industry and Technological Regime (2002-2005)

Absolute and relative frequencies of

M&A

Industries No. Target

firms No. Total

Firms % target

firms

Oil & Gas 18 154 11.69 Basic Materials 31 318 9.75

Industrials 166 1,329 12.49 Consumer Goods 89 721 12.34

Healthcare 40 358 11.17 Consumer Services 177 938 18.87

Telecomm. 21 98 21.43 Utilities 19 125 15.20 Financial 155 1,480 10.47

Technology 86 876 9.82 Total 802 6,398 12.54

Technological Regimes

No. Target firms No. Total Firms % target firms

SMI 264 2,015 13.10 SMII 109 1,116 9.77

Difference: SMI - SMII

3.33***

Note: values in last two columns represent difference of SMI, SMII proportions, and other sectors

Difference proportion test is based on binomial test * Significant at 1% level

** Significant at 5% level ***Significant at 10% level

These findings, related to the aggregate level,

confirm the assumption formulated in Section 2. Except for Telecommunications, large companies included in SMII sectors (Chemicals, Pharmaceuticals, Electronics, Computers) are less frequently targeted by other firms, probably because the costs to reorganise the specific assets and knowledge accumulated inside the firm are too high,

and the need to maintain continuity in specific lines of research and development is too expensive for the bidder, who generally expects short-term returns from the deal.

Some interesting results may be obtained from an analysis by country and sector. For instance, Italy and Sweden offer an interesting comparison, as shown in Table 8.

Table 8. Mergers and Acquisitions in Italy and Sweden by industry and Technological

Regime (2002-2005)

Italy Absolute and relative frequencies of M&A Difference between

proportions


firms No. Total

Firms % target

firms SMI SMII

Oil & Gas 0 4 0.00 13,79* 14,63* Basic Materials 1 11 9.09 4,7*** 5,54 Industrials 8 62 12.90 0,89 1,73 Consumer Goods 4 54 7.41 6,38** 7,22 Healthcare 2 6 33.33 -19,54* -18,7** Consumer Serv. 3 33 9.09 4,7*** 5,54 Telecom. 4 8 50.00 -36,21* -35,37* Utilities 5 17 29.41 -15,62* -14,78** Financial 14 87 16.09 -2,3 -1,46 Technology 4 30 13.33 0,46 1,3 Total 45 312 14.42 -0,63 0,21


23

Table 8 continued

Tech. Regimes

SMI 16 116 13.79 - 0,84

SMII 6 41 14.63 -0,84

Other Manufact. 1 18 5.56 8,23 9,07

Sweden Absolute and relative frequencies of M&A Difference between

proportions


firms No. Total

Firms % target

firms SMI SMII

Oil & Gas 0 9 0.00 12,99* 5,45* Basic Materials 4 25 16.00 -3,01 -10,55* Industrials 10 114 8.77 4,22*** -3,32 Consumer Goods 4 44 9.09 3,9*** -3,64 Healthcare 3 44 6.82 6,17* -1,37 Consumer Serv. 3 41 7.32 5,67** -1,87 Telecom. 4 11 36.36 -23,37* -30,91* Utilities 1 4 25.00 -12,01* -19,55* Financial 10 64 15.63 -2,64 -10,18* Technology 4 79 5.06 7,93* 0,39

Total 43 435 9.89 3,1 -4,44*** Tech. Regimes

SMI 20 154 12.99 - -7,54*

SMII 11 178 5.45 7,54*

Other Manufact. 4 62 6.45* 6,54 -1

* Significant at 1% level ** Significant at 5% level *** Significant at 10% level

Clearly, the weak and good performances of

corporate governance of, respectively, Italy and Sweden (see also Table 5) probably interact with the role played by technological regime, thus producing opposite effects. In other words, institutional factors and a poor performing ‘variety of capitalism’ may interfere with technological determinants, giving rise to too low levels of M&A in those sectors characterized by ‘creative destruction’ regimes of innovation.

These results bring a wider set of factors into the analysis to evaluate the market for corporate control. The low incidence of M&A activity in sectors featuring technological regime SMI and the low incidence of cross-border deals, as shown for Italy, confirm the low capabilities of Italian firms to remain competitive. Conversely, the results for M&A obtained for Nordic economies show opposite trends. 5. Conclusions

The prevailing literature on mergers and acquisitions focuses on investor protection and ownership structure as the main determinants of divergences between countries, but does not analyse whether the different patterns of reorganisation activities (pertaining to different sectors and different innovation patterns) affect the market of corporate control.

The new approach adopted in this study promises a more comprehensive perspective. Indeed, our results

for the European experience of the last merger wave suggest remarkable divergences in sectors characterised by different technological regimes. At the same time, the role of institutional determinants turns out not to be secondary, but as a driving force of the market for corporate control. These considerations may qualify and render more controversial the relation between technological features and takeover activities which may actually be discovered in reality.

For instance, in countries like those of Nordic Europe, featuring good quality and enforcement of their governance mechanisms, companies are exposed to changes in control and international contestability, Sweden is a good case in point, since takeovers are more frequent in ‘entrepreneurial’ sectors where radical, investment projects are short-lived, capital depreciation is rapid, and knowledge and competences are general, i.e., in SMI sectors. By contrast, in the same countries the incidence of M&A deals is lower in routinised sectors, featuring higher stability of the hierarchy of innovators, lower concentration ratio of the more innovative firms, and a higher portion of new innovators with respect to the old ones.

These findings are reversed in a country like Italy, where cross-border deals are less frequent, one quarter of all transaction involves national deals targeted to banks, merger activities seem to be unrelated to innovation sectoral patterns. This should be interpreted as confirmation that not all mergers are efficient devices to remedy faulty governance and/or


24

to exploit innovative synergies in entrepreneurial industries (marked as SMI sectors).

This motivates us to explore further various inefficiencies of takeover activities found in a vast theoretical and empirical literature.

A plausible explanation takes sectoral and innovative dimensions into explicit account. In this more comprehensive perspective, a further step should be overall evaluation of the main consequences of takeover activity, in order to detect in which corporate governance systems and in which sectoral regimes takeovers really do improve shareholders’ returns rather than managerial or block-holder private benefits.

References 1. AGNBLAD J. et al. (2001), “Ownership and Control in

Sweden: Strong Owners, Weak Minorities, and Social Control”, in Barca F. and Becht M. (2001), pp.228-258.

2. ANDRADE, G., MITCHELL M. and STAFFORD, E. (2001) “New Evidence and Perspectives on Mergers”, Journal of Economic Perspectives, 15, pp. 103-120.

3. ANDRADE, G. and STAFFORD, E. (2004) “Investigating the economic role of Mergers”, Journal

of Corporate Finance, 10, pp. 1-36. 4. AUDRETSCH D.B.(1996), Innovation and industry

evolution, The MIT Press, Cambridge. 5. BARCA, F. and BECHT, M. (2001) The Control of

Corporate Europe, Oxford University Press, Oxford. 6. BEBCHUK L. A. (1994), “Efficient and Inefficient

Sides of Corporate Control”, Quarterly Journal of

Economics, 109, pp. 957-93. 7. BECHT, M., BOLTON, P. and ROELL A.A. (2003)

“Corporate Governance and Control”, in CONSTANTINIDES, G., HARRIS, M., STULZ, R. (eds.) Handbook of the Economics of Finance, Amsterdam, North Holland, pp.4-109.

8. BIANCHI M. and BIANCO M. (2006), “Italian Corporate Governance in the Last 15 Years: from Pyramids to Coalitions?”, ECGI Working Paper Series in Finance, n. 144

9. BIANCHI M., BIANCO M. and ENRIQUES L. (2001), “Pyramidal Groups and the Separation between Ownership and Control in Italy”, in Barca F. and Becht M. (2001), pp. 154-187.

10. BRUNER, R.F. (2004), Applied Mergers and

Acquisitions, Wiley & Sons, New York. 11. BURKART M. and PANUNZI F.(2004), “Mandatory

Bids, Squeeze-out, Sell-out and the Dynamics of the Tender Offer Process”, in FERRARINI G., HOPT K. J., WINTER J. and WYMEERSCH E. (eds.), Reforming Company and Takeover Law in Europe, Oxford University Press, Oxford, pp. 737-765.

12. CASSIMAN B. and COLOMBO M.G. (eds.) (2006), Mergers and Acquisitions. The Innovation Impact, Edward Elgar, Cheltenham, UK.

13. CHILOSI A. and DAMIANI M, “Stakeholders vs. shareholders in corporate governance”, Icfai Journal of

Corporate Governance, 2007, 6, pp. 7-45. 14. CORROCHER, N., MALERBA F. and MONTOBBIO

F. (2007), ‘Schumpeterian Patterns of Innovative Activity in the ICT Field’, Research Policy, 36, pp. 418-432.

15. Djankov, S., La Porta R., López-de-Silanes F. and Shleifer A. (2008), The Law and Economics of Self-Dealing’, Journal of Financial Economics, 88, pp. 430-465.

16. DOSI, G. (1988) “Sources, procedures and microeconomic effects of innovation”, Journal of

Economic Literature, 26, pp. 1120-1171. 17. ENRIQUES, L. and VOLPIN P. (2007), ‘Corporate

Governance Reforms in Continental Europe’, Journal

of Economic Perspectives, 21, pp.117-140. 18. EUROPEAN COMMISSION (1999) “Mergers and

Acquisitions”, Supplement Economic Trends, No.2. 19. EUROPEAN COMMISSION (2000) “Mergers and

Acquisitions”, Supplement Economic Trends, No.5/6. 20. EUROPEAN COMMISSION (2001) “Mergers and

Acquisitions”, Supplement Economic Trends, No.12. 21. FACCIO, M. and LANG, L.H.P. (2002) “The Ultimate

Ownership of Western European Corporations”, Journal of Financial Economics, 65, pp. 365-395.

22. FACCIO, M., STOLIN, D. (2003) “Expropriation vs. Proportional Sharing in Corporate Acquisitions”, Journal of Business, 79, pp. 1413-1444.

23. FOGEL (2006), ‘Oligarchic Family Control, Social Economic Outcomes and the Quality of Government’, Journal of International Business Studies, 37, pp. 603–622.

24. FRANKS J. and MAYER C. (1996) “Do Hostile Takeovers Improve Performance?”, Business Strategy

Review, 7, pp. 1-6. 25. FTSE - DOW JONES INDEXES (2004) “Rules for the

management of the Industry Classification Benchmark”, FTSE Index Company, London.

26. GOERGEN, M. and RENNEBOOG, L., (2004) “Shareholder wealth effects of European domestic and cross-border takeover bids”, European Financial

Management , 10, pp.9-45. 27. GOERGEN M. and MARTYNOVA M. and

RENNEBOOG L. (2005) “Corporate Governance Convergence: Evidence from Takeover Regulation Reforms in Europe”, Oxford Review of Economic

Policy, 21, pp. 243-268. 28. GOLBE, D.L. and WHITE, L.J. (1988) “A time series

Analysis of Mergers and Acquisitions in the U.S. Economy”, in Auerbach, A.J. (eds) Corporate

Takeovers: Causes and Consequences, Chicago, University of Chicago Press.

29. GROSSMAN, S.J. and HART O.D. (1980) “Takeover Bids, the Free Rider Problem, and the Theory of the Corporation”, Bell Journal of Economics, 11, pp. 42-64.

30. GUGLER K. and MUELLER D.C. and YURTOGLU B.B. (2004), “Corporate Governance and the Returns on Investment”, Journal of Law and Economics, XLVII, pp. 589-633.

31. HARFORD J. (2005) “What driver mergers waves?”, Journal of Financial Economics, 77, pp. 529-560.

32. HOLMSTROM, B. and KAPLAN, S.N. (2001) “Corporate Governance and Merger Activity in the US: Making Sense of the 1980s and 1990s”, Journal of

Economic Perspectives, 15, pp. 121-145. 33. JENSEN M.C. (1988) “Takeovers: Their Causes and

Consequences”, Journal of Economic Perspectives, 2, pp. 21-49.

34. JOVANOVIC, B. and ROUSSEAU, P. (2001) “Merger and Technological Change”: 1885-1998”, Working Paper No. 01-W16, Vanderbilt University.


25

35. JOVANOVIC, B. and ROUSSEAU, P. (2002) “The Q-theory of mergers”, American Economic Review, 92, pp.198-204.

36. KAPLAN, S.N., WEISBACH, M.S. (1992), “The Success of Acquisitions: Evidence from Divestitures”, Journal of Finance, 47, pp. 107-138.

37. LA PORTA R. ,SHLEIFER A., VISHNY R. and DE SILANES F. (1998) “Law and Finance”, Journal of

Political Economy, 106, pp. 1113-1155. 38. MALERBA, F and ORSENIGO, L. (1993)

“Technological regimes and firm behaviour”, Industrial

and Corporate Change, 2, pp. 45-74. 39. MALERBA, F. and ORSENIGO, L. (1996)

“Schumpeterian patterns of innovation are technology specific”, Research Policy, 25, pp. 451-478.

40. MALERBA, F. and ORSENIGO, L. (1997) “Technological regimes and sectoral pattern of innovative activities”, Industrial and Corporate

Change, 6, pp. 173-194. 41. MANCHIN, M. (2004) “Determinants of European

cross-border mergers and acquisitions”, Economic Papers No. 212, European Commission.

42. MANNE, H.G. (1965) “Mergers and the market for corporate control”, Journal of Political Economy, 73, 110-120.

43. MARTYNOVA M. and RENNEBOOG L. (2006) “Mergers and acquisitions in Europe”, in L. Renneboog Advances in Corporate Finance and Asset

Pricing, ch.2, pp. 15-76. Elsevier, Amsterdam. 44. McCAHERY, J. - RENNEBOOG, L., RITTER, P. and

HALLER, S. (2004) “The Economics of the Proposed European Takeover Directive”, in FERRARINI, G., HOPT, K.J., WINTER, J., WYMEERSCH, E. (eds) Reforming Company and Takeover law in Europe, Oxford, Oxford University Press, pp.575-646.

45. MITCHELL, M.L. and MULHERIN, H.J. (1996) “The impact of industry shocks on takeover and restructuring activity”, Journal of Financial Economics, 41, pp. 193-229.

46. OECD (2001), “New Patterns of Industrial Globalisation”, OECD Report, Paris.

47. MOELLER S.B. - SCHLINGEMANN, F.P. and STULZ, R.M. (2005), “Wealth Destruction on a Massive Scale? A study of Acquiring-Firm Returns in the Recent Merger Wave”, Journal of Finance, 60, pp. 757-782.

48. MORCK R., WOLFENZON, D. and YEUNG B. (2005), “Corporate Governance, Economic Entrenchment, and Growth”, Journal of Economic

Literature, 43, pp. 655-720. 49. ROLL, R. (1986) “The Hubris Hypothesis for

Takeovers”, Journal of Business, 59, pp. 197-216. 50. ROSSI, S. and VOLPIN, P. (2004) “Cross-Country

Determinants of Mergers and Acquisitions”, Journal of

Financial Economics, 74, pp. 277-304. 51. SHLEIFER, A. and VISHNY R. (1997) “A Survey of

Corporate Governance”, The Journal of Finance, 52, pp. 737-783.

52. SHLEIFER, A. and VISHNY R. (2003) “Stock market driven acquisitions”, Journal of Financial Economics, 70, pp. 295-311.

53. STIGLER, G. (1950) “Monopoly and Oligopoly by Merger”, American Economic Review, 40, pp. 23-34.

54. SUDARSANAM, S. (2003) Creating Value from

Mergers and Acquisitions: the Challenges, FT Prentice Hall.

55. THOMSON FINANCIAL (2002) European Legal

Advisors Press Release, League Tables. 56. THOMSON FINANCIAL (2003) European Legal



Advisors Press Release, League Tables. 59. TIROLE J. (2006) The Theory of Corporate Finance,

Princeton University Press, Princeton. 60. VAN DIJK, M. (2000) “Technological Regimes and

Industrial Dynamics: the Evidence from Dutch Manufacturing”, Industrial and Corporate Change, 6, pp. 83-117.

Appendix

Table A1. Sources for takeover regulation

UK The Panel on Takeovers and Mergers (2002), The Takeover Code

Sweden The Swedish Financial Supervisory Authority’s Regulations Governing Rules of Conduct on the Securities Market (2002) Tude B., “Swedish Securities Council Issues Statement on Mandatory Bids”, International Financial Law Review, 2000

Finland Himonas D., “The Financial Supervision Authority Imposes New Guidelines”, International Financial Law Review, 2000 Securities Market Act 26.5.1989/495

Norway Act on Securities Trading, Act no. 79 , 19th June 1997

Germany Roos M. – Cornett C. (2002), Takeover season in Germany, AltAssets Schmid F.A. – Wahrenburg M. (2002), Mergers and Acquisition in Germany, The Federal Reserve Bank of St. Louis, Working Paper Series 2002- 027A

France Décret n. 2003-1109 du 21 Novembre 2003 Relatif à l'Autorité des Marchés Financiers Règlement général de l’Autorité des marchés financiers, 2006

Italy Testo Unico Finanziario, Decreto Legislativo 24 febbraio 1998, n. 58 , “Testo Unico delle Disposizioni in Materia di Intermediazione Finanziaria, ai sensi degli Articoli 8 e 21 della Legge 6 febbraio 1996, n. 52”


26

THE CORPORATE SOCIAL RESPONSIBILITY MANAGER MAP

Mario Molteni*, Matteo Pedrini**

Abstract

This article outlines the profile of ninety managers directly involved in Corporate Social Responsibility activities (CSR Managers) in Italian firms. It presents an analysis of the organisational position, educational background and activities of these professionals. The results suggest that CSR managers: 1) have a growing relevance in the firm; 2) are predominantly existing members of the organization; 3) have principally a business management educational background; 4) play a key role in supporting senior management and improving stakeholder engagement. It emerges that CSR managers are supporting senior management in different manners. The “CSR Manager Map” allows for the identification of four types of CSR manager: (1) Specialist; (2) Generalist; (3) Process oriented; (4) External oriented.

Keywords: Corporate Social Responsibility; Manager; Stakeholder Engagement; Stakeholder view. *Full Professor of Corporate Strategy, Catholic University of the Sacred Heart – Milan, Director of ALTIS – Postgraduate School of Business and Society, [email protected] **Researcher of Altis, Catholic University of the Sacred Heart – Milan, Research Fellow at ALTIS – Postgraduate School of Business and Society, [email protected]

1.Introduction The proposition that firms are responsible for the effects produced by their activities on stakeholders and society is becoming extensive year by year. The Corporate Social Responsibility (CSR) represents a new strategic issue for executives and academics. There is a widespread acknowledgement that CSR can take a variety of forms, and significant efforts has been devoted to developing theoretical frameworks vis-à-vis CSR issues and practices Currently thought this effort is surrounded by much the same ambiguity as it was 30 years ago (Sethi, 1979, Waddock, 2004). Together with the normative CSR theory development, academics in recent years have also manifested a greater interest in implementation issues. This new interest calls for a better understanding about the mode, tools and people engaged in the integration of CSR in strategy (Verschoor, 2006). Although the interest is intensified, there is a lack of research which describes the features of those functionaries who support the introduction of CSR into the corporate management structure.

Interest in CSR has spawned a new generation of manager: the “Corporate Responsibility Managers” (CSR managers). They support senior management in implementing the new conception of the firm, and they facilitate leadership’s reorientation of corporate culture, values, strategy, systems and tools in accordance with stakeholder view.

This research analyzes 90 Italian CSR managers to identify their characteristics and the activities they perform. By utilizing a cluster analysis based on the type and number of CSR issues that a given manager

implements it proposes the “CSR Manager Map”, a typology of the different CSR managers.

2.Theoretical framework In recent years we have witnessed in literature the idea that stakeholders are relevant to business competitiveness (Van De Ven, Jeurissen, 2005), and have seen the stakeholder view become part of the wider theory of the firms (McWilliams, Siegel, 2001, Post, Preston & Sauter-Sachs, 2002). There is an acknowledgement that the quality of its relations determines the firms’ power to generate long term wealth (Nelson, 2002). The normative reasons for CSR (Reynolds, Schultz & Hekman, 2006) and this relation between stakeholder management and competitiveness (De Man, 2005, Gardberg, Fombrun, 2006) have stimulated broad corporate engagement in CSR practices. The choice to manage a firm in accordance with the stakeholder view transforms the business concept and leads to changes in company decision-making, processes and activities (Robins, 2006).

2.1.The CSR manager’s functions in CSR implementation CSR represents a renovation, more or less profound, in the firm’s strategy, processes and activities. This change is based on the direct engagement of senior managers (Reynolds, Schultz & Hekman, 2006) in the implementation of CSR in the firm (Smith, 2003). It thus becomes necessary to identify a leader of the team that supports the senior managers who conduct this change. The CSR manager fulfils this role. He or


27

she has the responsibility to support the executives’ new management conception that can potentially impact on all firm’s elements. The amplitude of the effects of CSR implementation requires that the CSR manager cover three roles (Figure 1):

(1) the sensor of social and environmental changes; (2) the integrator of those engaged in the CSR implementation team; (3) the expert in CSR issues and practices.

Sensor of social and

environmental changes

Integrator of CR team

Expert in CR issues and practices

CRMANAGER

Offering to executives a critical synthesis of social and

environmental trends to be

considered in strategy

(Molteni, 2004)

Assuring the different members’ actions

contribute to a unique

project

(Panwar, 2006)

Support executives in the implementing CR practices

(Hess, 2002)

Figure 1. The CSR manager’s critical features

2.1.1 The sensor of social and environmental changes The CSR manager weighs sustainability issues in decision making and aids strategy makers in thinking about their industries ongoing social and environmental trends. The manager has to collaborate with the board and CEO in strategy development (Molteni, 2006). His contribution consists in offering a synthetic judgment of those social and environmental trends to be considered among the strategic variables. To monitor these social and environmental trends the CSR manager has to be involved in engagement activities and posses a proficiency in conducting a continuous dialogue with stakeholders (O'Dwyer, 2005).

2.1.2 The integrator of the CSR implementation team The CSR manager provides the cohesion between the multiple internal actors involved in CSR implementation. He or she assures that diverse members of the firm contribute to a unique strategic plan (Panwar et al., 2006). The implementation path requires a team composed of experts from each of the firm’s functions that work in close contact with the senior management. The CSR manager assures that the implementation process is efficient and in line with the executives’ strategy (Elkington, Emerson & Beloe, 2006). To do this the CSR manager collaborates with people from a large range of functions and with different educational background, and he or she needs to maintain cohesion between a large ranges of people work in the firm.

2.1.3 The expert in CSR practices The CSR manager needs to be an expert in those practices which translate into expressions of responsibility toward stakeholders and community (Hess, Rogovsky & Dunfee, 2002). He or she has the competence to manage multiple CSR issues or practices. The issues dealt with depend upon individual corporate strategy. They are analyzed in the next paragraph.

2.2.The CSR manager’s tasks A large range of new tools and practices have emerged as direct expressions of CSR. The presentation of an exhaustive list of practices regarding the responsibility of a corporation in society appears a difficult charge. Every single CSR implementation case has to fit the firm’s strategy and nexus of stakeholders. Consequently these practices are often creative and comprise a unique solution. A limited list seems inadequate to define the CSR manager’s activities, but a classification of the CSR manager’s tasks by universally recognized categories of CSR issues is possible. (Thorne McAlister, Ferrell & Ferrell, 2005). Those CSR issues are: (1) the integration of CSR in strategy and decision-making; (2) the extension of corporate governance; (3) responsible supply chain management; (4) social accountability; (5) socially responsible investing; (6) philanthropy and business in the community; (7) environmental management; (8) corporate welfare.


28

2.2.1. The integration of CSR in strategy and decision-making The CSR manager’s primary task is to assist senior management in integrating CSR into corporate strategy, and to present an assessment of stakeholder claims in decision-making. CSR consists in a new notion of business that introduces the adherence to a strategy based on the equilibrium between economic, social and environmental performance (Elkington, Emerson & Beloe, 2006). This adherence influences the firm’s competitiveness and can lead to the reconsidering of corporate mission, values and the entire strategy framework. The CSR manager assists the senior management in the definition of a strategy aimed at bridging the existent gap between stakeholders’ expectations and the organization (Wartick, Mahon, 1994) and at facilitating corporate response to threats or opportunities. He or she contributes to the development of a new approach, and moving toward a synthesis in stakeholder satisfaction where engagement is the essential condition in the development of a new, stakeholder inclusive strategy. The continual affirmation of the unavoidability of stakeholder engagement is at the fulcrum of the CSR manager’s job. From a normative point of view stakeholder engagement is indispensable because of those legitimate stakeholders’ interests that give them the moral right to managerial attention (Post, Preston & Sauter-Sachs, 2002). This means that the CSR manager has to favour the weighting of decisions according to both the stakeholders’ salience while also taking into consideration the interests of multiple stakeholders (Mitchell, Agle & Wood, 1997). In accord with the instrumental approach, the CSR manager introduces a stakeholder engagement based appoint the impact they have on firm’s success (Frooman, 1999). 2.2.2 The extension of corporate governance The redefinition of the firm through stakeholder engagement calls for an extension in the interests considered by corporate governance. The change in corporate fundamentals renders obsolete the traditional corporate governance principle: to assure shareholders (principal) a sufficient guaranty on their investments by managers (agent) (Shleifer, Vishny, 1997). The introduction of CSR forces the company to recognize that the stakeholders for whom it is responsible can be more extensive. (Hemphill, 2004). This extension of responsibility redefines corporate governance, moving it from the traditional mechanism of the principal-agent relationship to the new rules of the firm-stakeholder relationship. In accordance with legal frameworks the CSR manager’s task is to facilitate the design of an “owner model” of enlarged corporate governance. This is done by integrating different stakeholders’ claims both in a representative democracy – the election of members of the board of directors– and a participative democracy – a management in which stakeholders are involved in

decision making. The CSR manager cans introduce tools like the ethical code, social internal auditing, an ethical committee, an ethical officer or he or she can conceive new corporate governance tools. 2.2.3 Social accountability Social accountability involves the collection and disclosure of independent measures of social, environmental and economic performance (Elkington, 1997). Accountability is not wholly coherent area or activity, but could be understood as “the preparation and publication of an account about an organisation’s social, environmental, employee, community, customer and other stakehoder interactions and activities, and, where possible, the consequences of those interaction and activities” (Gray, 2000).

The CSR manager helps senior managers to interpret social accountability as a new form of transparency which educates the stakeholder and, in this way, enables a more participative democratic form of governance (Owen et al., 2000). The CSR manager contributes to the design of a system of key performance indicators, but principally he or she structures a dialogic process that empowers stakeholders to hold organisations accountable for actions impacting directly on their lives (O'Dwyer, 2005).

The CSR manager’s accountability task is to introduce a formal mechanism through which stakeholders might be assured a voice that could influence decision making (Kerr, 2004). At the same time the CSR manager has to recognize the difficulties in assuring adequate stakeholder representation and try to reduce the heterogeneity of expectations. He or she has to develop meaningful participation (Unerman, Bennett, 2004), and move from mono-directional communication to an accountability that fosters the interaction between stakeholder groups (Thomson, Bebbington, 2005). 2.2.4 Socially responsible investing We can define SRI as “the process of integrating personal values and societal concerns into investment decision making” (Schueth, 2003). In the ’90s socially responsible investing (SRI) became a style of investment that involved a relevant number of firms, persons and capital. It represents an opportunity for profitable investment (Sethi, 2005). SRI basically consists in the facing of two problems: (1) a financing problem consisting in the necessity of mirroring personal values and societal concerns to attract the socially responsible investment (Schueth, 2003); (2) an investment problem consisting in the assurance that the investment portfolio of a financial service firm respects a set of social criteria and maximizes profitability in turbulent economic times.

The CSR manager’s task is to address the financing problem, creating consistency between the CSR implementation process and the SRI criteria and


29

thus assuring the capacity to attract capital of ethical investments. In the financial services firms the CSR manager faces the investment problem. He or she has to suggests and support the introduction of policies that insure that the firm’s investments are limited to responsible and sustainable companies. The CSR manager also faces the sometimes ambiguous relationship between profitability and ethical investments. He or she has to continuously demonstrate the benefits of SRI (Kurtz, 1997, Statman, 2000). 2.2.5 Philanthropy and business in the community Philanthropic activities represent a discretionary manifestation of CSR that it is not an explicit exchange of value between two parties but, rather, a transfer of wealth from one party to another. The CSR manager’s task is to guide the philanthropic activities toward a compromise between disinterested giving and the natural profit interest of business. He or she has to support the implementation of a strategic philanthropy (Post, Waddock, 1995). It means that a CSR manager collaborates with senior management to define a specific strategy to gradually orient these nonreciprocal transfers toward the interest of the firm (Margolis, Walsh, 2003) and to pursuit a win-win strategy also in corporate giving. To achieve this objective he or she can suggest actions such as a cause related advertising program to boost a company’s reputation (Fombrun, 2005); the use of philanthropy to create differentiation and thus competitive advantage (Porter, Kramer, 2002) or the creation of new business opportunities through the development of innovative services and products (Smith, 1994). 2.2.6 Environmental management The recent debate on the relationship between environment and business suggests that there are extensive opportunities for business to benefit from environmental investment in regards to profitability (Johnson, Greening, 1999) and the development of competitive advantages. These benefits can be seen as low cost, differentiation or heterogeneous resources (Porter, Van Der Linde, 1995). The firm’s assumption of environmental responsibility facilitates the design of an Environmental Strategic Management System. Such a system takes into account the firm’s assumption of responsibilities about the externalities the environmental consequences of its activities as well as the growing attention of customers to these issues (Orsato, 2006). The firm’s final objective is to satisfy the customers’ environmental expectations and obtain a differentiation advantage which can be imitated only with difficulty (Reinhardt, 1999). The firm thus gains those customers that are willing to pay the cost of a product’s positive environmental standing.

The CSR manager contributes to ecological sustainability of a business by supporting the

implementation of a management system that combines positive environmental performance and financial performance (Christmann, 2000). The CSR manager is the link between the firm and the professionals and organizations with expertise in environmental management issues. In the case that management decides to access to specific external competences, the CSR manager is the person that coordinates this external network. 2.2.7 Corporate welfare Corporate welfare is a structured system of actions aimed to improve the general welfare of the workforce. The employee welfare programs haven’t a legislative mandate and are aimed at offering a comprehensive health service for employees. Issues such as weight, fitness, nutrition, smoking, dealing with stress, and drug awareness are covered. These programs’ benefits consist in the opportunity to reduce the cost of medical insurance, lower the absentee rate due to illness, and lessen turnover (Drennan, Ramsay & Richey, 2006). They offer a corresponding reduction in recruitment and training cost. In addition these programs can contribute to increased productivity (Mitchell, Davis, 2006). The CSR manager is responsible of the coherence between the CSR strategic objective and workforce management. In this role he or she collaborates with the Human Resources Function. In particular the CSR manager assures that the model developed conforms simultaneously to a number of factors (Davis, Gibson, 1994): the firm’s size, administrative structure and history, the availability of services, the rate of current health coverage, the degree of unionization, the physical layout of the plant or office, and the existence of policies and procedures that facilitate or obstruct personal problem identification or resolution. 2.2.8 Responsible supply chain management Today supply chain management must respond to an array of social and environmental pressures including regulations, consumer demands, and limited resource availability, while at the same time efficiently delivering “the right product at the right time” (Jamison, Murdoch, 2004). This involves the development of distinct operating models, objectives, and new supply chain processes that expand the scope of management to operate in a responsible way (Mamic, 2005).

Along these lines the CSR manager conducts a supply chain review to satisfy stakeholders’ environmental and social expectations. He or she has to respect the growing consumer demand for more environmentally friendly and socially responsible products. Those consumers rarely accept the same products with inferior performance, and are, albeit within limits, willing to pay a premium price for social and environmental friendly attributes (Mohr, Webb & Harris, 2001). The CSR manager supports the process of redefining supplier identification


30

criteria and has to arrive at a model that allows for the coexistence of cost efficiency and responsible supplier behaviour. One the great challenge the CSR manager faces is to help the supply chain designer manage the tension between the best business practices and environmental, social and labour standards (Jamison, Murdoch, 2004).

The role assigned to the CSR manager consists in supporting the correct implementation of CSR in a firm. Though the relevance of such an activity stimulates interest in this new career, CSR literature has yet to study the features and roles of this emerging profession. This research aims at recognizing the principal features of CSR managers in medium and large Italian firms to better understand their contributions.

3.Methodology 3.1.Sample and data collection To identify the sample for this study, the Italian Firms that in the recent years have demonstrated efforts in

CSR issues were considered the sample was identified on a non-random basis. A firm to be included in the sample had to meet one of the following three criteria: (1) nominated to the Sodalitas Social Awards – the most important CSR award in Italy – during the years 2003 or 2004; (2) published a Social or Sustainability Report in 2003 and/or 2004; (3) cited in the principal Italian newspaper (Corriere della Sera) or in the principal Italian financial newspaper (Il sole 24 ore) for a CSR best practice in the years 2003 or 2004.

These criteria allowed for the identification of 354 firms that were successively contacted to verify the existence of a CSR manager. Of the 354, 90 took part in the research (25.4%). The CSR managers identified work prevalently in firms that are not publicly traded (72.2%) and that have more than 999 employees (42.2%) (Table 1).

Table 1. The sample

Employees Listed <50 50-250 251-999 >999

Total

Listed 1 2 2 22 27 Not listed 13 16 17 17 63 Total 14 18 19 39 90

3.2.Variable description and measurement

Data on CSR issues were obtained from a survey directly addressed to each CSR manager. The variables used in the research were the following. − official assignment: presence or absence in the

organization of an expressed assignment of responsibility for CSR activities or practices;

− functional dependence: the function/person to which the CSR manager reports. The possible function of the direct superior of the CSR manager proposed in the interview were: CEO, General Manager, CFO, Public Relations Director (P.R. Director), Legal Affair Director, Human Resource Director (H.R. Director), Planning and Control Director and Internal Audit Director.

− functional collaboration: the corporate functions which collaborate with the CSR manager in the implementation process of CSR . The alternatives proposed were identical to those proposed regarding functional dependence, but for the exception of the CEO, the collaboration with whom is at the essence of the CSR manager’s task.

− educational background: the question proposes four types of education: business administration, law, humanities, and a residual category (other).

− official CSR documentation: the existence of official documents that express a direct reference to CSR issues. The accuracy of this data was verified directly by the researcher. Documents in question were: mission and vision statements, strategic plans, ethical codes, presentations to financial analysts, marketing and investment policy papers.

− intensity of stakeholder engagement: the intensity in a Linkert 5 points scale of the commitment to engage direct stakeholders, associations and experts. Collaborators, customers, institutional investors, shareholders and suppliers were considered direct stakeholders. The category of associations included the community, consumer groups, environmental unions, foundations, human right defence unions, industrial unions, NGOs, political parties, religious organizations and trade unions. The engagement of experts was explored in relation to academics, consultants and professionals.

− CSR issues and practices realized: the involvement of the CSR manager in those engagement issues proposed in the previous paragraph. This reveals a profile of the CSR manager according to a wide range of activities. In addition to the main features of the CSR

manager’s the research considered firm related variables to explore the dependence that exists between the firm’s characteristics and the CSR


31

manager’s features: the firm’s dimension (in accordance with the European Community Recommendation 2003/301/EC of 6th may 2003) and listing or not on the Italian exchange.

3.3.Data analysis The multi dimensional data obtained by the survey was synthesized in a two-dimensional output space. The map obtained represents the similarity between the vectors of data collected about CSR issues in terms of distance between the respective units located in a fix grid. The “CSR Manager Map” is based on the standard architecture and learning procedure of Kohonen’s Self-Organizing Map (SOM) with rumours. The SOM helps us to label the multidimensional data in a two dimensional “virtual” output space, by an activation function based on a city block distance between a weight vector.

The weight vector used to calculate the distances includes four variables: the number of CSR issues in which CSR manager is present in at least one activity or practice (weight 0.40), the engagement in each single CSR issue (weight = 0.10), the number of CSR practices and activities realized by the firm (weight = 0.02) and the engagement in every single CSR practice and activity (0.01). The weight assignation was used to evaluate the CSR managers’ engagement in CSR issues, and in this sense, the number and amplitude of CSR issues in which he or she is engaged was considered more relevant than the involvement in individual practices. At the same time intensity of engagement, measured by the number of CSR issues and the number of CSR practices and activities in which the managers is involved and the amplitude of engagement, measured by distance in

each single CSR issue and single CSR practice or activity was considered more relevant.

The SOM establishes the planar coordinates for each firm that, successively, was the basis for the clusters identification. The cluster analysis is based on a four clusters K-means analysis. The clusters identified by this methodology were tested with a hierarchical clustering and the results were consistence with those obtained with K-means.

4.The CSR manager role, an empirical analysis

The research allows for the identification of the trends that define the current features of the CSR manager. The evidence suggests that the modern day CSR manager is a person that: (1) has a growing relevance in the firm; (2) is identified in an existing member of organization; (3) has principally a business and management educational background; (4) supports the senior management in coordinating CSR integration in company strategy and in the consequent CSR implementation process; (5) promotes stakeholder engagement activities.

The growing relevance In last five years the number of firms that have decided to have a CSR manager has steadily increased (Figure 2). This trend emerges when observing the increases in the number of firms in the sample that have assigned CSR responsibility to a specific manager. This number passes from seven official assignments in 2000 to thirty-seven in 2005. In particular, in the years from 2002 to 2004, the number of official CSR managers grew with a rate superior to 30% a year.

1111

1 1 3 58 9

2621

1611

8

42110

5

10

15

20

25

30

35

40

1997 1998 1999 2000 2001 2002 2003 2004 2005

SMEs Large

Figure 2. The official CSR managers’ assignments by firms’ dimension (%) An existing member of the organization

The CSR manager is in prevalently identified from among the people that work in the firm at the time that management decides to implement CSR: 88.2% of the firms analyzed have chosen the CSR manager among the existing members of the organization (Figure 3). This trend is common to both large firms and SMEs. Both prefer an internal CSR manager

rather than recruiting the CSR professional from the outside (90.0% large firms; 87.5% SMEs). This clear preference emerged may be for three different reasons. First, the majority of CSR managers spend only a part of their work time on CSR issues (Figure 4) and thus it is simpler to assign this position to a person that can spend the work time remaining in another way for the company. Second, senior


32

management consider it important that the CSR manager have a broad understanding of a firm’s process, organization and activities. and in this sense is easy to attribute the position to a person that has a

deeper knowledge of organization. Third, currently it exists a restricted number of professionals in the labour market with job experience and specific competences in CSR issues.

12,5%

10,0%

11,8%

87,5%

88,2%

90,0%

Large

SMEs

Total

Professional from outside Existing member of organization

Figure 3. The official CSR managers’ assignments by firms’ dimension (%)

20,5%

11,1%

20,5% 31,8%

18,5%

27,3%

59,3%11,1%

Unofficial

Official

Full time > 50% 25% - 50% < 25%

Figure 4. The CSR manager work time spent for CSR issues by firms’ dimension (%)

A managerial background

The CSR manager when integrating CSR into company practice is more apt to work from a business (65.1% large firms; 64.3% SMEs) rather than a humanistic educational background (7.0% large firms; 25.0% SMEs) (Figure 5).

This broad diffusion of management background doesn’t mean that these managers are approaching CSR implementation exclusively from an instrumental point of view. On the contrary, is possible to affirm that that the wide diffusion of formal business education leads to greater attention to the company benefits inherent in CSR.

Business & administration

Law

Humanistic

Other

Large SMEs Total

65,1

18,6

9,3

7,0

64,3

7,1

25,0

3,6

64,8

14,1

15,5

5,6

Figure 5. The CSR manager educational background by firms’ dimension (%)

A support to senior management in

coordinating activities and functions CSR managers are significantly involved with the executive staff in the CSR implementation process coordinating the various corporate functions necessary to realize new practices. The research shows that the majority of the CSR managers report directly to a top management figure (Figure 6), 40.6% directly to the CEO (39.5% large firms; 42.3% SMEs), while another 29% to the General Manager (23.3% large firms; 38.5% SMEs). In particular it

emerged that the CSR managers are intensely collaborating with: the Human Resource Function (71.1% large firms; 48.3% SMEs), supporting their efforts to realize a employee welfare management; the Public Relations Function (48.9% large firms; 24.1% SMEs), implementing social accounting practices and developing engagement practices with external stakeholder; the General Manager (55.6% large firms; 65.5% SMEs), redefining internal process and introducing supply chain or environmental management (Figure 7).


33

Large SMEs Total

39,5

23,3

11,6

11,6

9,3

2,3

2,3

42,3

38,5

3,8

-

-

7,7

7,7

40,6

29,0

8,7

7,2

5,8

4,3

4,3

CEO

General Manager

CFO

P.R. Director

Legal Affair Director

H.R. Director

Other

Figure 6. The CSR managers’ functional dependence by firms’ dimension (%)

Large SMEs Total

71,1

55,6

48,9

24,4

17,8

13,3

13,3

48,3

65,5

24,1

17,2

24,1

17,2

10,3

62,2

59,5

39,2

21,6

20,3

14,9

12,2

H.R.

General Manager

P.R.

Health and safety

Legal Affair Director

Plan and Control

Internal Audit

Figure 7. The function with CSR managers collaborate by firms’ dimension (%)

A promoter of stakeholder engagement

The CSR manager with an official assignment works closely with the CEO to increase stakeholder engagement, and thus augments the intensity of the dialogue. The research requested that the CSR manager evaluate the intensity of their engagement activities as related to eighteen potential stakeholders. The mean intensity of effort realized by the CSR manager with an official assignment is stronger for each category of stakeholder (Figure 8). The one exception is the engagement with collaborators, which is almost equal among those with official and unofficial assignments. In particular the official CSR manager enhances the intensity of collaboration with customers, institutional investors, shareholder, suppliers, NGOs, academics and trade unions.

The official CSR managers are introducing a particular attention toward stakeholder dialogue and

are increasing engagement activities in their firms. In this sense the official CSR manager is more likely to monitor social and environmental trends and support top management by synthesising the claims and trends that emerge as a result of stakeholder engagement activities.

5.The CSR Manager Map The research presents those trends that define the main features of the CSR manager features, but a detailed analysis of the multiple activities for which they are responsible demonstrates that CSR managers play varying roles. The relation between the number and the type of issues managed suggests that there exist four types of CSR manager. Each type correspond to a particular clusters of a matrix that we call the “CSR Manager Map” (Figure 9 and Table 2).


34

4,24

3,76

1,86

2,14

3,16

4,19

4,14

2,78

2,68

3,65

3,81

1,65

1,76

1,49

1,27

2,51

1,38

1,38

1,65

2,57

3,54

2,57

2,41

1,86

1,81

2,84

2,30

1,65

1,70

3,59

1,76

1,11

2,35

2,03

Collaborators

Customers

Institutional Investor

Shareholders

Suppliers

Community

Consumer Unions

Environmental Unions

Foundations

Human Right Defense Unions

Industrial Unions

NGOs

Politics

Religious organizations

Trade Unions

Academics

Consultants and professionals

Unofficial Official

DIRECT

ASSOCIATIONS

EXPERTS

Figure 8. The intensity of stakeholder engagement by assignment (1-5 scale) (%)

Philanthropy and

Social Responsible Investing

Environmental and

Ethical Supply Chain Management

One Area

ProfessionalCR

Professional

Process Oriented

External Oriented

Generalist

Focalized

Figure 9. The CSR Managers Map


35

Table 2. The centres of clusters and number of cases

Centres Clusters

Number of areas Type of areas Number of cases

Specialist ,25 ,53 4 Generalist ,81 ,37 5

Process Oriented ,59 ,84 11 External Oriented ,53 ,27 12

This tool identifies the typology of each CSR

manager, the role they play and the mandate they have obtained from senior management. The typologies of CSR manager defined in the map are:

(1) specialist;

(2) generalist; (3) process oriented; (4) external oriented. The table 3 synthesizes the main characteristics

of the typologies of CSR managers.

Table 3. The main characteristics of CSR managers

CSR Manager Profiles Characteristics

Focalized Generalist Process oriented External oriented

Executive’s mandate

Manage one or two CSR issues

Manage a relevant number of CSR issues

Direct the CSR practices: − Regarding internal

stakeholders − With impact on

internal processes

Direct the CSR practices: − Regarding external

stakeholder − With impact on

reputation Skills & competences requested

Technicality of practices

General knowledge of CSR issues

− CR internal impact − Technicality of

management systems

− Reputational management

− Communication

Functional collaboration

Handy to CSR practice managed

High level Limited to internal of function impacted (i.e. environmental or supply chain management)

− Public relation − Communication − Financial services

5.1 Specialist CSR Manager The Specialist CSR manager has the responsibility of one or two issues related to CSR implementation. This is the most operative way to interpret the role. The CSR manager technically addresses the practices to implement, and possesses a profound expertise in the issues he or she has to manage. The operational task is principally the development and coordination of the collaboration between corporate functions. The collaborations he or she manages are limited to the emerging needs of the operation of which he or she is in charge. 5.2 Generalist CSR Manager The Generalist CSR Manager manages a significant number of CSR issues and in some cases has to supervise all the practices related to CSR. The generalist has a broad responsibility and as a consequence must work in close contact with senior management. The principal job of this CSR manager is the coordination of the CSR practices implemented by the specific corporate functions involved in the changes generated by the stakeholder approach. This manager has a general knowledge of CSR issues and

his/her work is assessed on the capacity to assure the fit of CSR activities within the corporate strategic framework. 5.3 Process Oriented CSR Manager The Process Oriented CSR manager is engaged in CSR practices that have an impact on the firm’s processes and internal stakeholders. His/her principal task consists in supporting the functions involved in the reengineering of one or more internal process in accordance with CSR. The task of this CSR manager regards typically the implementation of an environmental management system or the redefinition of the supply chain. He or she focuses on the technicalities of the management systems. This CSR manager is evaluated based on specific technical skills and on the capacity to collaborate with the managers responsible for the functions impacted by the new management systems. In this way he or she insures the coherence of the new systems and processes with CSR principles.


36

5.4 External Oriented CSR Manager The External Oriented CSR Manager has the responsibility to manage the implementation of those CSR practices which regard the external stakeholders and have an impact on corporate reputation. Their mandate consists predominantly in directing those activities which most effectively enhance the ethical reputation of a firm, namely philanthropy and socially responsible investing. This type of CSR manager pays close attention to the impact that the CSR activities have on stakeholder opinion. This doesn’t mean that this CSR manager’s job does not affect the firm’s processes, but that these effects are rather consequences of the attention paid to society and external stakeholders and. In some cases the consequent internal change process will be part of another manager’s job. The External Oriented CSR Manager possesses competencies in public relations and communication. The criteria used to judge this manager is the ability to suggest efficient and economic ways to reduce the legitimacy gap, the capacity to build a good corporate reputation and the ability to support senior management when facing reputational crisis.

Discussion and conclusion This article represents a first step in the comprehension of the role that the CSR manager plays in supporting senior management’s decision making and implementation process related to CSR issues. It emerges that the CSR manager is a new professional that has the basic function to support the management on CSR issues, but the mandate received varies according to his/her sphere of responsibility. The CSR manager can play diverse roles in CSR implementation. In some cases they are directly in charge of the management of one or more CSR practices and assume the role of professional. In other cases they coordinate the activities related to CSR issues that have an impact on internal processes or on external stakeholder opinion. Although the role may vary, it is clear that his job determines the efficiency of CSR implementation. The CSR manager’s role is fundamental for transforming executive strategy into operational activities, and, in other words, in establishing a new stakeholder culture in the firm. Future research though is desirable. In particular the future fields of research that we consider relevant are: the influence that different CSR approaches (a natural approach; a crisis reaction; a strategy redefinition) have on the CSR managers’ tasks; the connection between CSR manager type and the effectiveness of CSR strategy; the role of the CSR manager in strategy deliberation and in reducing gap between deliberate and emergent strategy; and the criteria used to evaluate the efficiency and proficiency of CSR manager activities.

References 1. ADAMS, R., (1997), "Linking Financial and

Environmental Performance", Environmental Accounting and Auditing Reporter, 2 (10), pp. 4-7.

2. AHRENS, D. (2004), Infesting in Vice: The Recession-Proof Portfolio of Booze, Bets, Bombs, and Butts, St. Martin's Press, New York, NY.

3. BALL, A. (2004), "A Sustainability Accounting Project for the UK Local Government sector? Testing the Social Theory Mapping Process and Locating a Frame of Reference", Critical Perspectives on Accounting, 15(8), pp. 1009-1035.

4. BEAUCHAMP, T.L. and N.E. BOWIE, (2001), Ethical Theory and Business, VI edn, Prentice Hall, Englewood Cliffs, NJ.

5. BERMAN, S.L., A.C. WICKS, S. KOTHA and T.M. JONES, (1999), "Does Stakeholder Orientation Matter? The Relationship Between Stakeholder Management Models and Firm Financial Performance", Academy of Management Journal, 42(5), pp. 488-506.

6. BURNES, B. (1992), Managing Change, Pitman, London, UK.

7. CARROLL, A.B. (1979), "A Three Dimensional Conceptual Model of Corporate Social Performance", Academy of Management Review, 4(4), pp. 497-505.

8. CHRISTMANN, P. (2000), "Effects of «Best Practices» of Environmental Management on Cost Advantage: The Role of Complementary Assets", Academy of Management Journal, 43(4), pp. 663-680.

9. CLARKE, R.A., R.N. STAVINS, J.L. GREENO, J.L. BAVARIA, F. CAIRNCROSS, D.C. ESTY, B. SMART, J. PIET, R.P. WELLS, R. GRAY, K. FISCHER and J. SCHOT, (1994), "The Challenge of Going Green", Harvard Business Review, 72(4), pp. 37-49.

10. CLARKSON, M.B.E. (1995), "A Stakeholder Framework for Analysing and Evaluating Corporate Social Performance", Academy of Management Review, 20(1), pp. 92-117.

11. COHEN, G.S., L.H. GARD and W.R. HEFFERNAN, (1998), "Employee Assistance Programs: A Preventive, Cost-effective Benefit", Journal of Health Care Finance, 24(3), pp. 45-53.

12. Cohen, M.A., S.A. FENN and J. NAIMON, (2000), Environmental and Financial Performance: Are They Related? Owen Graduate School of Management, Nashville.

13. DAVIS, A. and L. GIBSON (1994), "Designing Employee Welfare Provision", Personnel Review, 23(7), pp. 33-45.

14. DE MAN, F. 2005, "Corporate Social Responsibility and Its Impact on Corporate Reputation", Brand Strategy, 195, pp. 40-41.

15. DONALDSON, T. and L.E. PRESTON (1995), "The Stakeholder Theory of the Corporation: Concepts, Evidence, and Implications", Academy of Management Review, 20(1), pp. 65-91.

16. DOWELL, G., S.L. HART and B. YEUNG (2000), "Do Corporate Environmental Standards Create or Destroy Market Value?", Management Science, 46(8), pp. 1059-1074.

17. DRENNAN, F.S., J.D. RAMSAY and D. RICHEY, (2006), "Integrating Employee Safety and Fitness", Professional Safety, 51(1), pp. 26-35.


37

18. ELKINGTON, J. (1997), Cannibals with Forks. The Triple Bottom Line of 21st Century Business, Capstone Publishing, Oxford, UK.

19. ELKINGTON, J. (1994), "Towards the Sustainable Corporation: Win-Win-Win Business Strategies for Sustainable Development", California Management Review, Winter, pp. 90-100.

20. ELKINGTON, J., J. EMERSON, and S. BELOE, (2006), "The Value Palette: a Tool for Full Spectrum Strategy", California Management Review, 48(2), pp. 6-28.

21. FOMBRUN, C.J. (2005), "Building Corporate Reputation Through CSR Initiatives: Evolving Standards", Corporate Reputation Review, 8(1), pp. 7-11.

22. FROOMAN, J. (1999), "Stakeholder Influence Strategies", Academy of Management Review, 24(2), pp. 191-205.

23. FRY, L.W., G.D. KEIM and R.E. MEINERS, (1982), "Corporate Contributions: Altruistic or For-Profit?", Academy of Management Journal, 25(1), pp. 94-106.

24. GARDBERG, N.A. and C.J. FOMBRUN (2006), "Corporate Citizenship: Creating Intangible Assets Across Institutional Environments", Academy of Management Review, 31(2), pp. 329-346.

25. GRAY, R.H. (2000), "Current Developments and Trends in Social and Environmental Auditing, Reporting and Attestation: a Review and Comment", Internationa Journal of Auditing, 4(3), pp. 247-268.

26. GRAY, R., D. OWEN and K. MAUNDERS, (1996), Account and Accountability. Changes and Challenges in Corporate Social and Environmental Reporting, Prentice Hall Europe, UK.

27. HART, S.L. and G. AHUJA, (1996), "Does It Pay To Be Green? An Empirical Examination of the Relationship Between Emission Reduction and Firm Performance", Business, Strategy and the Environment, 5, pp. 30-37.

28. HASNAS, J. (1998), "The Normative Theories Of Business Ethics: a Guide for The Perplexed", Business Ethics Quarterly, 8(1), pp. 19-42.

29. HEMPHILL, T.A. (2004), "Corporate Citizenship: The Case for a New Corporate Governance Model", Business and Society Review, 109(3), pp. 339-361.

30. HESS, D., N. ROGOVSKY, and T.W. DUNFEE (2002), "The Next Wave of Corporate Community Involvement: Corporate Social Initiatives", California Management Review, 44(2), pp. 110-125.

31. HIGGINS, C.P. (2002), "Triple Bottom Line Reporting: The Importance of Reputation and Trust", New Academy Review, 1(2), pp. 91-112.

32. HOFFMAN, A.J. (1997), From Heresy to Dogma: an Institutional History of Corporate Environmentalism, 1st edn, New Lexington Press, San Francisco, CA.

33. JAMISON, L. and H. MURDOCH (2004), Taking the Temperature: Etichal Supply Chain Management? The Institute of Business Ethics, UK.

34. JAWAHAR, I.M. and G.L. MCLAUGHLIN, (2001), "Toward a Descriptive Stakeholder Theory: an Organizational Life Cycle Approach", Academy of Management Review, 26(3), pp. 397-414.

35. JOHNSON, R.A. and D.W. GREENING (1999), "The Effects of Corporate Governance and Institutional Ownership Types on Corporate Social Performance", Academy of Management Journal, 42(5), pp. 564-576.

36. KERR, J.L. (2004), "The Limits of Organizational Democracy", Academy of Management Executive, 18(3), pp. 81-95.

37. KLONOSKI, R.J. (1991), "Foundational Considerations In the Corporate Social Responsibility", Business Horizons, 34(4), pp. 9-18.

38. KOTLER, P., N. ROBERTO and N. LEE (2002), Social Marketing: Improving the Quality of Life, Sage Publications, Thousand Oaks, CA.

39. KURTZ, L. (1997), "No Effect, or No Net Effect? Studies on Socially Responsible Investing", Journal of Investing, 6(4), pp. 37.

40. LEVY, R. (1999), Give and Take: A Candid Account of Corporate Philantrophy, Harvard Business School Press, Boston, MA.

41. MAMIC, I. (2005), "Managing Global Supply Chain: The Sports Footwear, Apparel and Retail Sectors", Journal of Business Ethics, 59(1/2), pp. 81-100.

42. MARGOLIS, J.D. and J.P. WALSH (2003), "Misery Loves Companies: Rethinking Social Initiatives by Business", Administrative Science Quarterly, 48(2), pp. 268-305.

43. MCWILLIAMS, A. and D. SIEGEL, (2001), "Corporate Social Responsibility: a Theory of Firm Perspective", Academy of Management Review, 26(1), pp. 117-145.

44. MINTZBERG, H. (1983), Power in and around organizations, Prentice-Hall, Englewood Cliffs, NJ.

45. MITCHELL, K. and C. DAVIS, (2006), "Reducing Lost Time: The Correlation Between Family-medical Leave and Short-term Disability", Employee Benefit News, 20(6), pp. 16-18.

46. MITCHELL, R.K., B.R. AGLE and D.J. WOOD, (1997), "Toward a Theory of Stakeholder Identification and Salience: Defining The Principle of Who and What Really Counts", Academy of Management Review, 22(4), pp. 853-886.

47. MOHR, L.A., D.J. WEBB and K.E. HARRIS, (2001), "Do Consumers Expect Companies to Be Socially Responsible? The Impact of Corporate Social Responsibility on Buying Behavior", Journal of Consumer Affairs, 35(1), pp. 45-72.

48. MOLTENI, M. (2006), "The Social-Competitive Innovation Pyramid", Corporate Governance: an International Review, 6(4), pp. 516-526.

49. NELSON, J. (2002), Building Competitiveness and Communities: How World Class Companies Are Creating Shareholder Value and Societal Value, Prince of Wales Business Leaders Forum, London, UK.

50. O'DWYER, B. (2005), "Stakeholder Democracy: Challenges and Contributions from Social Accounting", Business Ethics: A European Review, 14(1), pp. 28-41.

51. ORSATO, R.J. (2006), "Competitive Environmental Strategies: When Does it Pay to Be Green?", California Management Review, 48(2), pp. 127-143.

52. OWEN, D.L., T.A. SWIFT, C. HUMPHREY and M. BOWERMAN (2000), "The New Social Audits: Accountability, Management Capture or the Agenda of Social Champions?", European Accounting Review, 9(1).

53. PANWAR, R., T. RINNE, E.HANSEN, and H. JUSLIN (2006), "Corporate Responsibility. (Cover story)", Forest Products Journal, 56(2), pp. 4-12.

54. PERRINI, F. (2006), "The Practitioner's Perspective on Non-Financial Reporting", California Management Review, 48(2), pp. 73-103.

55. PORTER, M.E. (2003), "The Corporation and Society: The Role of Corporate Philanthropy", Second EABIS Colloquium 2003, ed. EABIS, EABIS, .


38

56. PORTER, M.E. and M.R. KRAMER, (2002), "The Competitive Advantage of Corporate Philanthropy", Harvard Business Review, 80(12), pp. 56-68.

57. PORTER, M.E. and M.R.KRAMER, (1999), "Philanthropy’s New Agenda: Creating Value", Harvard Business Review, 77(6), pp. 121-130.

58. PORTER, M.E. and C. VAN DER LINDE, (1995), "Green and Competitive: Ending the Stalemate", Harvard Business Review, 73(5), pp. 120-133.

59. POST, J.E. and S.A. WADDOCK, (1995), "Strategic Philanthropy and Partnerships for Economic Progress" in Philanthropy and economic development, ed. R.F. America, Greenwood press, Westport, CO.

60. POST, J.E., L.E. PRESTON, and S. SAUTER-SACHS, (2002A), "Managing the Extended Enterprise: The New Stakeholder View", California Management Review, 45(1), pp. 6-28.

61. POST, J.E., L.E. PRESTON, and S. SAUTER-SACHS, (2002B), Redefining the Corporation: Stakeholder Management and Organizational Wealth, Stanford University Press, Stanford, CA.

62. PRESTON, L.E. and H.J. SAPIENZA, (1990), "Stakeholder Management and Corporate Performance", Journal of Behavioural Economics, 19(4), pp. 361-375.

63. REINHARDT, F.L. (1999), "Bringing the Environment Down to Earth", Harvard Business Review, 77(4), pp. 149.

64. REYNOLDS, S.J., F.C. SCHULTZ and D.R. HEKMAN, (2006), "Stakeholder Theory and Managerial Decision-Making: Constraints and Implications of Balancing Stakeholder Interests", Journal of Business Ethics, 64(3), pp. 285-301.

65. ROBINS, F. (2006), "The Challenge of TBL: A Responsibility to Whom?", Business and Society Review, 111(1), pp. 1-14.

66. ROCHLIN, S.A. and J. BOGUSLAW, (2001), Business and Community Development, The Center for Corporate Citizenship at Boston College, Newton, MA.

67. SCHNIETZ, K.E. and M.J. EPSTEIN, (2005), "Exploring the Financial Value of a Reputation for Corporate Social Responsibility During a Crisis", Corporate Reputation Review, 7(4), pp. 327-345.

68. SCHUETH, S. (2003), "Socially Responsible Investing in the United States", Journal of Business Ethics, 43(3), pp. 189-194.

69. SETHI, S.P. (2005), "Investing in Socially Responsible Companies is a Must for Public Pension Funds - Because There is no Better Alternative", Journal of Business Ethics, 56(2), pp. 99-129.

70. SETHI, S.P. (1979), "A Conceptual Framework for Environmental Analysis of Social Issues and Evaluation of Business Response Patterns", Academy of Management Review, 4(1), pp. 63.

71. SHLEIFER, A. and R.W. VISHNY, (1997), "A Survey of Corporate Governance", Journal of Finance, 52(2), pp. 737-783.

72. SMITH, C. (1994), "The New Corporate Philanthropy", Harvard Business Review, 72(3), pp. 105-116.

73. SMITH, N.C. (2003), "Corporate Social Responsibility: Whether or how?", California Management Review, 45(4), pp. 52-76.

74. STATMAN, M. (2000), "Socially Responsible Mutual Funds", Financial Analysts Journal, 56(3), pp. 30-39.

75. STONE, B.A. (2000), "Social responsibility and institutional investment: an empirical analysis of the environmental screen", The Journal of Investing, 9(3), pp. 81-94.

76. THOMSON, I. and J. BEBBINGTON, (2005), "Social and environmental reporting in the UK: a pedagogic evaluation", Critical Perspectives on Accounting, 16(5), pp. 507-533.

77. THORNE MCALISTER, D., O.C. FERRELL, and L. FERRELL, (2005), Business and Society. A strategic Approach to Social Responsibility, 2nd edn, Houghton Mifflin Company, Boston, MA.

78. UNERMAN, J. and M. BENNETT, (2004), "Increased Stakeholder Dialogue and the Internet: Towards Greater Corporate Accountability or Reinforcing Capitalist Hegemony?", Accounting, Organizations and Society, 29(7), pp. 685-707.

79. VAN DE VEN, B. and R. JEURISSEN, (2005), "Competing Responsibly", Business Ethics Quarterly, 15(2), pp. 299-317.

80. VERSCHOOR, C.C. (2006), "Good corporate citizenship is a fundamental business practice", Strategic Finance, 87(9), pp. 21-22.

81. WADDOCK, S.A. (2004), "Parallel Universes: Companies, Academics, and the Progress of Corporate Citizenship", Business and Society Review, 109(1), pp. 5-42.

82. WARTICK, S.L. and J.F. MAHON, (1994), "Toward a Substantive Definition of the Corporate Issue Construct", Business and Society, 33(3), pp. 293.

83. WERHANE, P.H. (1999), Moral Imagination and Management Decision-making, Oxford University Press, New York, NY.

84. WINN, M.I. (2001), "Building Stakeholder Theory With a Decision Modelling Methodology", Business and Society, 40(2), pp. 133.

85. WOOD, D.J. and R.E. JONES, (1995), "Stakeholder Mismatching: A Theoretical Problem in Empirical Research on Corporate Social Performance", The International Journal of Organizational Analysis, 3(3), pp. 229-267.

86. ZHUANG CAI and P. WHEALE, (2004), "Creating Sustainable Corporate Value: A Case Study of Stakeholder Relationship Management in China",

Business and Society Review, 109(4), pp. 507-547.


39

IMPACT OF INDEX DERIVATIVES ON INDIAN STOCK MARKET

VOLATILITY-AN APPLICATION OF ARCH AND GARCH MODEL

S. V. Ramana Rao*, Naliniprava Tripathy**

Abstract The present study examined the impact of introduction of index futures derivative and index option derivative on Indian stock market by using ARCH and GARCH model to capture the time varying nature of volatility presence in the data period from October 1995 to July 2006. The results reported that the introduction of index futures and index options on the Nifty has produced no structural changes in the conditional volatility of Nifty but however the market efficiency has been improved after the introduction of the derivative products. The study concludes that financial derivative products are not responsible for increase or decrease in spot market volatility, but there could be other market factors which influenced the market volatility. Keywords: ARCH, GARCH, option futures, index future and dummy coefficiants

*Assistant Professor (Finance), S.S. Institute of Management, Secunderabad, India Email: [email protected] **Associate Professor (Finance), Indian Institute of Management Shillong, India Email: [email protected]/[email protected]

1. Introduction

During the last decade, many countries across the world have introduced derivative products in their capital markets. The impact of introduction of derivative products on the underlying stock market volatility received major attention all over the world. Investors in the market make their decisions based on their perceptions of the distribution of future stock returns; therefore, any knowledge about future volatility has an enormous influence on their investment behavior. Hence high level of volatility in capital markets often raises issues related to public policies on the stability of financial markets and the economy as a whole. So the present study examined the impact of derivative products viz., index futures and index options on the spot market (Nifty) volatility. The study also investigates whether volatility is time varying and predictable in the market.

The remaining of this paper is structured as follows: the next section presents a brief survey of the literature, in Section: 3 describe the data used in this study, the methodology and the results of the study are presented in Section: 4 and finally Section: 5 conclude the paper.

2. Review of literature Mayhew Stewart (2000) focused on how the introduction of derivative securities affects the underlying market. In most cases, the resulting models predicted that speculative trading and

derivative markets stabilize the underlying market and make spot markets more liquid and more information ally efficient. Huseyin Gulen and Stewart Mayhew (2000) studied the stock market volatility before and after introduction of index futures in 25 countries. They examined whether spot market volatility after the introduction was related to futures market volume and open interest. They found that futures’ trading was related to an increase in conditional volatility in US and Japan, but for the rest of the countries they found no significant effect. It was further found that, except for US and Japan, volatility tended to be lower in periods when open interest was high. Benilde Maria Do Nascimen Oliveira et al (2001) analyzed the impact of futures introduction in the Portuguese stock market. The authors used GARCH model to measure the futures trading impact on the spot market volatility. They found that introduction of the PSI-20 index futures trading increased the Portuguese spot market volatility and that there was no improvement on the market efficiency front. Board John et al (2001) examined the effect of futures market volume on spot market volatility in two parallel markets viz. The equity (spot or cash) market and the market for futures on equity index by using GARCH and stochastic volatility models. They found that there was no evidence that futures trading instantly destabilized the spot market or that an increase in volume in one market, relative to the other, instantly destabilized the other market. The stochastic volatility analysis showed that there is a positive influence on spot market volatility which indicates that spot market volatility has increased. Ali F Darrant, et al (2002)


40

examined the role of futures trading in spot market fluctuations. The Exponential Generalised AutoRegressive Conditional Heteroscedasticity was used to measure volatility and relationship between volatility in the spot and futures markets for the period November 1987 to November 1997. They concluded that index futures trading could not be blamed for the volatility in the spot market. Pilar Corredor and Rafael Santamania examined the effect of the introduction of derivatives (futures and options) in the Spanish market on the volatility and on the trading volume of the underlying index. The period of analysis covers October 1990 to December 1994 and they used GARCH, EGRCH and GJR models to find out the conditional volatility. The findings of the study are that the level of volatility had decreased when derivative markets were introduced but the trading volume of the Ibex 35 had increased. They concluded that the introduction of derivatives did not present a problem for the spot market because their impact was beneficial and possessing characteristics similar to the Spanish market. Kiran Kumar Kotha and Mukhopadhyay Chiranjit (2003) studied the impact of futures introduction on underlying NSE Nifty volatility by using a CUSUM plot and GARCH family of techniques to capture the time varying nature of volatility and volatility-clustering phenomenon’s presence in the data. The results reported that the introduction of index futures trading had no effect on the mean level of returns of Nifty. Nagaraj KS and Kiran Kumar Kotha’s (2004) studied the impact of Index Futures trading on the spot market volatility by using GARCH model. The study found that the market had become more efficient in assimilating the information into the market after the introduction of index futures. Furthermore the study indicated that Monday and Friday are significant in terms of day of week effect of the market and also that volatility of the spot market has declined. YPSingh and S. Bhatia (2006) studied the futures trading impact the spot market volatility by using the GARCH (1, 1) model. The study proved that the daily spot market volatility in India has marginally declined since the introduction of futures trading in India. The market efficient has also improved after introduction of derivative products in the Indian capital markets. The study also found that there is a marginal decline in daily volatility on the expiration day of the derivative contracts. These results are quite contrary to the earlier studies results.

3.Time series data &methodology The present study has been carried out with the three-benchmark indices: S&P CNX Nifty, S&P CNX Nifty Junior and S&P 500 index for studying the volatility behavior. Daily closing prices from June 1999 to July 2006 were obtained from NSE website for Nifty and Nifty Junior; similarly S & P 500 index from Prowess. For the data set daily returns are calculated

for Nifty, Nifty Junior and S & P 500 index by using the following formula. Rt = ln (Pt/Pt-1)*100 Pt and Pt-1 are the prices at time t, and t-1 respectively and Rt is the return for time t.

Financial time series usually exhibits a characteristic called volatility clustering which means that large changes tend to follow large changes and small changes tend to follow small changes. In either case, the changes from one period to the next period are typically of unpredictable nature. GARCH model accounts for certain characteristics like fat tails and volatility clustering that are commonly associated with financial time series. Volatility of the stock markets is measured by using Standard Deviation or GARCH model. The GARCH model provides for heteroscedasticity in the observed returns. GARCH Model is a time series modeling technique that uses past variance and the past variance forecasts to forecast future variances. It is observed that the model that takes into account the changing variance can make more efficient use of the data. The Generalized Auto Regressive Conditional Heteroscedasticity (GARCH) model is a variation of the Auto Regressive Conditional Heteroscedasticity (ARCH) model developed by Engle in 1982. Bollerslev originally proposed the GARCH model in 1986. A distinguishing feature of this model was that the error variance might be correlated over time because of the phenomenon of volatility clustering. In analyzing the behavior of volatility due to derivative products, it is necessary to eliminate the influences of other factors. This is achieved by regress the return series over its lags, over its exogenous variables and over day-of-the-week. The GARCH (1, 1) framework has been found to have the best specification in both the works. Hence this work has been carried out with GARCH (1, 1) model with necessary modifications.

ARCH The ARCH (q) process captures the conditional heteroscedasticity of financial returns by assuming that today’s conditional variance is a weighted average of past squared unexpected returns:

2222

2110 .... qtqttth −−− ++++= εαεαεαα

Where ,,0 10 αα > qαα ........2 > and et1 It-1 ~ N (0,

ht) If a major market movement occurred yesterday, the day before or up to q days ago, the effect will be to increase today conditional variance because all parameters are constrained to be non-negative. In fact, the ARCH model with exponentially declining lag coefficients is equivalent to a GARCH (1, 1) model.


41

GARCH The full GARCH (p, q) model adds p auto regressive turns to the ARCH (q) specification and the conditional variance equation take in the form as follows:

............221122

222

110 ptpttqtqttt hhhh −−−−−− +++++++= βββεαεαεαα

qαααα ,......,,0 210 > > 0, pβββ ........21 > 0

In the ARCH (q) process the conditional variance is specified as a linear function of past squared observations only, whereas the GARCH (p, q) process allows lagged conditional variances to enter as well. The GARCH (p, q) process defined above is stationary when –

)......()........( 2121 pq βββααα ++++++

GARCH (1, 1) process model given by

2110 −+= tth εαα tt Dh 211 ββ ++ −

10 ,0 αα > > 0 1β > 0

Where Dt is a dummy variable that takes a value of zero before introduction of derivative and value of 1 after introduction of derivative

The stationary condition for GARCH (1, 1) is

111 <+ βα

The GARCH (1, 1) model is represented as given below.

Where RNifty, t is the daily return on the S&P CNX Nifty index calculated as the first difference of the log of the index, RNifty, t-1 is the lagged Nifty return, RNiftyJunior,t is the return of Nifty Junior, RSP500,t-1 is the lagged S & P 500 index return, and DAYj are day-of-the-week dummy variables for Monday to Friday except Thursday. It is essential to remove domestic market wide factors influence on Nifty returns to isolate the impact of derivative products introduction into the Indian capital market. In order to do this Nifty Junior, a proxy variable which does not have derivative products and which captures market wide influence of domestic markets in India has been included in the above conditional mean equation. In GARCH, the residuals (ut) from the conditional mean equation assumed to be distributed in N (0,ht) where the conditional volatility ht is given by the following equation:

tttt Dy+h+yy+y=h 3122

110 −−∈

Where Dt is a dummy variable that takes a value of zero before introduction of index options, index futures, stock options and stock futures period and value of one in the post introduction period. y1 reflects the impact of recent news (shocks), y2 reflects the impact of old news (persistence volatility). For a GARCH process to be stable, it is necessary that y0 and y1 and y2 are greater than zero and for y1+y2<1. y3 would indicate the impact of introduction of derivative products on the underlying stock market volatility. A significant negative value for y3 would indicate that derivatives introduction decreases the volatility. Similarly, a positive value suggests increase of volatility.

4. Empirical analysis The study tried to find out whether the nature of the GARCH process was changed as a result of the derivative products introduction. The Indian stock markets have witnessed changes like rolling settlement system, electronic trading system, and creation of futures and options segment etc. These changes could have contributed to improve the efficiency of the market operations. Hence, to explore this, the GARCH (1, 1) model is estimated separately for the pre index futures and post index futures, pre index options and post index options. This facilitates obtaining more information about the effect of the introduction of index derivative products in the Indian stock market. The conditional mean and variance equations are similar except that the dummy variable for derivative products would not be there in conditional variance equation. The results of these two equations are helpful in analyzing the information efficiency in the pre and post derivative products period, examining the persistence of volatility and the changes in the day-of-week effect.

The GARCH analysis of before introduction of derivatives products has exhibited in table 1explains that there is an impact of Nifty Junior, Lag S&P 500 and lag Nifty on Nifty. It is observed that Monday coefficient is significant at 5% level. It indicates that before the introduction of index futures, Monday effect was there in the market as it happens to be opening day of the week and there may be weekend effect also. The output of conditional variance equation y1 and y2 are related to efficiency of news. y1 relates to the impact of yesterday’s market specific changes today. It can be inferred that recent news has an impact on price changes. y2 coefficient reflects the impact of old news on yesterday’s variance and as such it indicates the level of persistence in the information effect on volatility. Persistence of volatility means that today’s volatility due to information arriving today will affect tomorrow’s volatility and volatility of the days to come.


42

Table 1. Estimates of the GARCH (1, 1) model before the Introduction of Index Futures

Particulars Coefficient t-Statistic Prob.

α0 Intercept -0.010343 -0.196482 0.8442 α1 Lagged Nifty return -0.03459** -2.355926 0.0185 α2 Nifty Junior return 0.862942* 66.5254 0.0000 α3 Lagged S & P500 return 0.116363* 5.88259 0.0000 α4 Dummy-Monday -0.140195** -1.950164 0.0512 α5 Dummy-Tuesday 0.018224 0.479469 0.6316 α6 Dummy-Wednesday -0.030494 -1.323961 0.1855 α7 Dummy-Friday 0.015306 0.967904 0.3331 yo ARCH (0) 0.018371* 3.309349 0.0000 y1 ARCH (1) 0.074674* 5.42106 0.0000 y2 GARCH (1) 0.907286* 57.03493 0.0000

* Statistically Significant at 1% level ** Statistically Significant at 5% level

Table 2. Estimates of the GARCH (1, 1) model after introduction of Index Futures


α0 Intercept -0.040276 -0.784469 0.4328 α1 Lagged Nifty return -0.014684 -1.020877 0.3073 α2 Nifty Junior return 0.68205* 66.44498 0.0000 α3 Lagged S & P500 return 0.068489* 3.843142 0.0001 α4 Dummy-Monday 0.019376 0.264724 0.7912 α5 Dummy-Tuesday 0.054594 1.400666 0.1613 α6 Dummy-Wednesday -0.020677 -0.823774 0.4101 α7 Dummy-Friday 0.028116** 1.934983 0.053 y0 ARCH (0) 0.032364** 2.347624 0.0118 y1 ARCH (1) 0.040982** 2.781785 0.0054 y2 GARCH (1) 0.909369* 28.45292 0.0000

* Statistically Significant at 1% level ** Statistically Significant at 5% level

It is observed from the above table 2 that lag

Nifty effect has vanished after index futures. In the post index futures period the Monday effect has disappeared and coefficient sign has also changed from negative to positive. The Friday coefficient is statistically significant at 5% level. The conditional variance equation output y0, y1 and y2 coefficients are statistically significant. Hence it can be inferred that there is an impact of old and latest news effect in the market. A close look at the signals indicating the

market efficiency as measured by y1+y2, signify the persistence of shocks in the market. In the pre index futures period (y1+y2) value was 0.98196 and the same has declined to 0.950351 in the post index futures period. This clearly indicates that there is a substantial reduction in persistence of volatility from pre index futures period to post index futures period. This reveals that market efficiency has increased after the introduction of index futures.

Table 3. Estimates of the GARCH (1, 1) model before Introduction of Index Options


α0 Intercept -0.022173 -0.446837 0.655 α1 Lagged Nifty return -0.029525** -2.12089 0.0339 α2 Nifty Junior return 0.786389* 73.24276 0.0000 α3 Lagged S & P500 return 0.100516* 5.713709 0.0000 α4 Dummy-Monday -0.124197*** -1.759886 0.0784 α5 Dummy-Tuesday 0.019548 0.529053 0.5968 α6 Dummy-Wednesday -0.011454 -0.517595 0.6047 α7 Dummy-Friday 0.022725 1.520238 0.1285 y0 ARCH (0) 0.020789* 3.828287 0.0001 y1 ARCH (1) 0.056676* 5.176842 0.0000 y2 GARCH (1) 0.920994* 65.8113 0.0000

* Statistically Significant at 1% level ** Statistically Significant at 5% level *** Statistically Significant at 10% level


43

Table 3 depicts the GARCH analysis of before introduction of index options explains that there is an impact of lag Nifty returns, Nifty Junior, and Lag S&P 500 on Nifty. It is observed that there is a Monday effect in the market as its coefficient is significant before the introduction of index options. The output of conditional variance equation y1 and y2

are related to efficiency of news. y1 relates to the impact of yesterday’s market specific changes today. It is observed that recent news has an impact on price changes. y2 coefficient reflects the impact of old news on yesterday’s variance and as such it indicates the level of persistence in the information effect on volatility

. Table 4. Estimates of the GARCH (1, 1) Model after Introduction of Index Options


α0 Intercept -0.047709 -0.859812 0.3899 α1 Lagged Nifty return -0.021352 -1.192609 0.2330 α2 Nifty Junior return 0.717164* 61.39636 0.0000 α3 Lagged S & P500 return 0.070346* 3.375954 0.0007 α4 Dummy-Monday 0.028089 0.36258 0.7169 α5 Dummy-Tuesday 0.039948 0.937569 0.3485 α6 Dummy-Wednesday -0.017907 -0.626009 0.5313 α7 Dummy-Friday 0.026069*** 1.658368 0.0972 y0 ARCH(0) 0.038553** 2.096763 0.036 y1 ARCH (1) 0.04142** 2.12461 0.0336 y2 GARCH (1) 0.896665* 20.80725 0.0000

* Statistically Significant at 1% level ** Statistically Significant at 5% level *** Statistically Significant at 10% level Table 4 exhibited that lag Nifty effect has

vanished after index options. In the post index options period, Monday effect has disappeared. The Friday coefficient is statistically significant at 5% level. The conditional variance equation output y0, y1 and y2 coefficients are statistically significant. Hence it can be concluded that there is an impact of old and latest news effect in the market. The market efficiency, which is measured by y1+y2, signifies the persistence of shocks in the market coefficient, shows negative sign. But however, it has become positive in the post index options period but not statistically significant. The y1 and y2 coefficients relate to latest and old news respectively. y1 coefficient is significant hence it can be inferred that recent news has an impact on price changes. Similarly y2 coefficient suggests that the market has been picking up impact of old news. The market efficiency as a measure of (y1+y2) indicates that there is a reduction of persistence of volatility in the post index options period because the value of (y1+y2) has come down from 0.97767 to 0.938085. It indicates that the market efficiency has improved after the introduction of index options.

5. Concluding observation The relationship between financial derivatives such as index futures, index options, stock options and stock futures and corresponding spot market volatility of Nifty has been analyzed by using the GARCH (1, 1) model. The GARCH (1, 1) model included the variables of the day of week effect, domestic market factors, previous day’s Nifty effect and worldwide market factors. The analysis of information efficiency through a separate GARCH model for pre and post

derivatives period showed that there is a shift in market efficiency due to information flow. There is a Friday effect in the market after introduction of derivative products into the Indian capital market. Apart from this, the result shows that the spot market volatility has not been affected by the introduction of financial derivatives since the dummy coefficients for index futures and index options are not statistically significant. But the findings indicate that the market has become more efficient after introduction of derivative products. However it concludes that financial derivative products are not responsible for increase or decrease in spot market volatility, there could be other market factors which influenced the Nifty volatility. The reasons for increase in volatility could be the enhancement of Foreign Institutional Investors participation in the equity market. Another reason could be the turnover of NSE grew by 69% and also the retail investors’ participation in securities markets has increased in the recent years.

References 1. Alif Darrant, Shafiqur Rahman, Maosen Zhong,”

(2002) “On the Role of Futures Trading in Spot Market Fluctuations: Perpetrator of Volatility or Victim of Regret?” The Journal of Financial Research, vol. xxv, no.3, pp 431-444.

2. Apte PG., (2004) “Study of the Impact of Derivatives Trading on Volatility and Liquidity in the Spot Market”, Unpublished Paper,

3. Benilde Mania Do Nascimen Oliveora and Mannel Jose da Rocha Armada, (2001) “The Impact of the Futures Market’s on Portuguese Stock Market”, Finance India, , vol. xv, no.4, pp 1251-1278.


44

4. Board John, Sandman Gleb and Sutcliffe Charles,(2001) “The Effect of Futures Market Volume on Spot Market Volatility”, Journal of Business

Finance and Accounting, , Vol. 28, No. 7&8, pp 799-819.

5. Covrig V and Melvin M (2001), “Asymmetric Information and Price Discovery in the FX Market: Does Tokyo Know More About the Yen?”, Journal of Empirical Finance, Vol.9, pp. 271-285.

6. Cox C. C (1976), “Futures Trading and Market Information”, Journal of Political Economy, Vol.84, pp.1215-1237.

7. Figleewski. S., (1981), “Futures Trading and Volatility in the GNMA Market”, Journal of Finance 36 (May): pp. 445-456.

8. Granger C W J and Huang B (1998), “A Bivariate Causality between stock prices and exchange rates: Evidence from Recent Asia Flu”. Quarterly Review of Economics and Finance, Vol.40, pp. 337-354.

9. Golaka C. Nath,( 2003 ) “Behaviour of Stock Market Volatility after Derivatives”, NSE News Letter, November

10. Haris L. (1989),“The October 1987 S&P 500 Stock-Futures Basis”, Journal of Finance, Vol.44, pp.77-99.

11. Hetam Sania Nupur and Soran Deb Saikat, (2003)“Impact of Index Futures on Indian Stock Market Volatility: An Application of GARCH Model”, Proceedings of the International Conference on

Business & Finance, ICFAI University Press, Hyderabad, pp 291-306.

12. Huseyin Gulen and Stewart Mayhew (2000) “Stock

Index Futures Trading and Volatility in International

Equity Markets” www.terry.ugaedu/finance/ research/workingpaper/infact.pdf

13. Kiran Kumar Kotha and Mukhopadhyay, (2003) “Impact of Futures Introduction on Underlying NSE Nifty Volatility”, Proceedings of the International

Conference on Business & Finance 2003, ICFAI University Press, Hyderabad, , pp 326-42.

14. Kawaller I. G., Koch P. D., and Koch T. W. (1987), “The Temporal Price Relationship between S&P 500 Futures and the S&P 500 Index”, Journal of Finance, Vol. 42, pp. 1309-1329.

15. Nagaraju KS and Kiran Kumar Kotha, (2004) “Index Futures Trading and Spot Market Volatility: Evidence from an Emerging Market”, The ICFAI Journal of

Applied Finance, , vol. 10, No. 8, pp 5 -16. 16. Mayhew Stewart, “The Impact of Derivatives on Cash

Markets: What We Have Learned”, (2000), www.terry.ugaedu/finance/research/workingpaper/infact.pdf

17. Pilar Corridor and Rafael Santamania,“Does

Derivatives Trading Destabilize the Underlying

Assets? Evidence from the Spanish Stock Market”, www.google.com.

18. Premalatha Shenbagaraman,(2003) “Do Futures and options Trading Increase Stock Market Volatility?” NSE Research Initiative, Paper 71.

19. Raju MT and Kiran Karande,(2003) “Price Discovery and Volatility on NSE Futures Market” SEB I Working

Paper Series,, No.7. 20. Stoll H. R. and Whaley R. E. (1990), “The Dynamics

of Stock Index and Stock Index Futures Return”, Journal of Financial and Quantitative Analysis, Vol.25, pp. 441-467.

21. Schwarz,T. V., and Laatsch. F. (1991),“Price Discovery and Risk Transfer in Stock Index Cash and Future Markets”, Journal of Futures Markets, 11, pp.669-683.

22. Stennis E. A., Pinar M. and Allen A. J. (1983), “The Futures Market and Price Discovery in the Textile Industry”, American Journal of Agricultural Economics, Vol.65, No.2, pp. 308-310.

23. Thenmozhi M., (2002) “Futures Trading, Information and Spot Price Volatility of NSE-50 Index Futures Contract”, NSE Research Initiative, , Paper 59

24. Turkington J. And Walsh D. (1999), “Price Discovery and Causality in the Australian Share Price Index Future Market”, Australian Journal of Management, Vol. 24, pp. 97-113.

25. Y.P.Singh and S. Bhatia. (2006),”Does futures trading impact spot market volatility? Evidence from Indian financial markets”, Decision, Vol 33, No 2 July-Dec pp41-62


45

CAVEAT WACC: PITFALLS IN THE USE OF THE WEIGHTED AVERAGE

COST OF CAPITAL

Sebastian Lobe*

Abstract

In Discounted Cash Flow valuations, the WACC approach is very popular. Therefore, knowing which limitations the concept inherits is essential. The objective of this paper is thus twofold: First, it is clarified that a constant WACC rate must fail if the implied leverage ratio is time-varying. This seems to be the rationale for defining a nonlinear WACC (NLWACC). However, the NLWACC appears to be rather artificial when allowing for time-varying WACCs. Second, although the NLWACC approach is further amplified in this paper, it must be emphasized that this approach is, even then, applicable only under specific conditions while a time-varying WACC is still able to provide reliable results. In conclusion, the WACC approach is a valid workhorse whose results can be economically interpreted. Keywords: WACC, cost of capital, capital structure * Center of Finance, Chair of Financial Services, University of Regensburg. Universitätsstraße 31, 93053 Regensburg, Germany. E-mail [email protected], phone +49 941 943 2727, fax +49 941 943 4979.

1. Introduction Methodological issues on valuation have experienced a remarkable renaissance in the financial literature over the last years. Fernandez (2004) initiated a provocative discussion claiming that the value of tax shields is not equal to the present value of tax shields. This claim is indeed provocative as it implies inter alia that the principle of value additivity is not working and that the seminal propositions of Modigliani and Miller are flawed. The subsequent discussion led in several journals revealed that the claim was not well substantiated. For example, Arzac and Glosten (2005) reconsider tax shield valuation looking at a value-based debt policy in the spirit of Miles and Ezzell (1980), and prove the validity of the respective valuation formula. Comments to Fernandez (2004) are given in a comparable vein also by Fieten et al. (2005), Cooper and Nyborg (2006), (2007), and Massari et al. (2008).

The advanced textbook by Kruschwitz and Löffler (2006) offers a rigorous and authoritative perspective on Discounted Cash Flow valuation leading to many new insights. Cooper and Nyborg (2008) analyze the case of tax-adjusted discount rates with investor taxes and risky debt.

Ruback (2002) advocates the Capital Cash Flows methodology as a simple approach to incorporate the value of the debt tax shield in valuation formulae. His methodology is founded on a value-based debt policy. Booth (2007) makes the case that the Capital Cash Flows and the Adjusted Present Value methodology is not easy to handle in specific valuation scenarios while the weighted average cost of capital (WACC) is a more flexible methodology.

For valuing companies or projects, WACC is the dominant Discounted Cash Flow approach in practice. The idea of this long-lived concept is that the total market value (debt and equity) is calculated by discounting the unlevered cash flows with the weighted returns for shareholders re and bondholders rd.

1 The tax rate t adjusts the return to bondholders downward to reflect the interest tax shield.

d d e eWACC w r (1 t) w r= − +

Recent survey evidence from the US, UK, and Germany supports the WACC’s dominance in practice.2 Three reasons could support this success: First, the input parameters are rather easily estimated from market data. Second, applying WACC does not require a commitment to judge how risky debt tax shields are. In other words, the debt policy does not have to be specified. This is not an appealing constellation as it is unknown how much the debt tax shield contributes to the company value. However, given today’s knowledge, only a vague idea exists which debt policies companies actually apply. Therefore, putting a valid value on the debt tax shield given these estimation problems is not such an easy task. Third, the WACC is computationally elegant. An iterative or recursive procedure is generally not necessary. However, the WACC also has its known shortcomings. It implies by definition that a periodic rebalancing of debt takes place to maintain the capital structure set forth in the WACC formula. If this

1 For ease of exposition the notation of Miller (2007) is adopted. Expectation operators are therefore also dropped. 2 For the US, see Bruner, Eades, Harris, & Higgins (1998), and Graham, & Harvey (2001), for the UK, see Arnold, & Hatzopoulos (2000), and for Germany, see Lobe, Niermeier, Essler, & Röder (2008).


46

inherent mechanism is not acknowledged, WACC is prone to errors.

(Miller, 2007) challenges WACC with a nonlinear WACC (NLWACC). My concern is to evaluate the validity of his assertion that WACC is not quite right, and to examine the potential contribution of NLWACC to the literature.

The remainder of the paper is organized as follows. In section 2, the newly introduced NLWACC is motivated and applied. Also, the NLWACC is generalized to allow for annuities with growth rates g ≠ 0%. In section 3, the concept of the NLWACC will be revisited from the perspective of rebalancing the capital structure using WACC before taxes. The merits of WACC and NLWACC are discussed. Taxes being crucial in this context as shown in the seminal work by Modigliani and Miller (1958), section 4 analyzes the after tax-case. Finally, section 5 summarizes shortly the findings and offers an outlook. 2. WACC and modified WACC: an example The notion of the modified WACC shall be covered briefly here: 1) Looking at a levered project with given outlays IC0, cost of capital WACC, and duration N, which break-even unlevered cash flows CF has the project to deliver to be acceptable? Financial acceptance is measured with the net present value NPV. To derive a unique solution for this question, an annuity structure (allowing for geometric growth) is imposed:3

00

N

N

N

N

(1 g)CF 1

(1 WACC) IC (WACC g)NPV 0 IC CF

WACC g (1 g)1

(1 WACC)

+ − + ⋅ − = = − + ⇔ =

− +−

+

(2) Adopting the numerical example of Miller (2007) which assumes g = 0%, IC0 = $200,000, t = 0%,

d d e eWACC w r (1 t) w r= − + =

0.25 0.06 0.75 0.12 0.105⋅ + ⋅ = , and N = 8, leads to

8

$200,000 0.105CF $38,173.86

1 (1.105)−

⋅= ≈

−

.

This threshold operating cash flow CF belongs to the shareholders and bondholders of the company. 2) Having performed this exercise Miller (2007) further asks what the equivalent annuity for shareholders CFe and bondholders CFd is? Equivalence here again is achieved by equating the NPV with zero at time T = 0 respectively. This leads in analogy to Eq. (2) to (assuming g = 0%):

e 0 ee

eN 8

w IC r 0.75 $200,000 0.12CF $30,195.43, and

1 (1 r ) 1 (1 0.12)− −

⋅ ⋅ ⋅ ⋅= = ≈

− + − +

3 The following geometric series is evaluated: CF = CF1, CF (1+g) = CF2, … , CF (1+g)N-1 = CFN.

d 0 dd

dN 8

w IC r 0.25 $200,000 0.06CF $8,051.80

1 (1 r ) 1 (1 0.06)− −

⋅ ⋅ ⋅ ⋅= = ≈

− + − +

Obviously, the sum of both annuities differs from the annuity of the sum of both flows:

e d

CF( $38,247.23)

CF ($30,195.43) + CF ($8,051.80) CF ($38,173.86)≠14444444244444443

3) Discounting the sum of both annuities ($38,247.23) with the textbook WACC of 10.5% will overestimate the project value in this numerical example.4 To overcome this misvaluation, Miller (2007) suggests discounting this cash flow with a modified version of the WACC dubbing it NLWACC (nonlinear WACC). This modified WACC is an internal rate of return r which can be expressed as an implicit function when incorporating also the possibility of geometrically growing annuities.5

e e d d

e d

N N N

N N N

w (r g) w (r g)r g

(1 g) (1 g) (1 g)1 1 1

(1 r) (1 r ) (1 r )

− −−= +

+ + +− − −

+ + +

In the numerical example with g = 0%, interpolating for r yields according to Eq. (3):

8 88

0.191236

r 0.75 0.12 0.25 0.06r 0.105553

1 1.12 1 1.061 (1 r)− −−

=

⋅ ⋅= + ⇔ ≈

− −− + 14444244443

Discounting with r now leads to a zero-NPV according to Eq. (2):

0

N

8N

(1 g)CF 1

$38,247.23 1 (1 0.105553)(1 r)NPV IC $200,000 $0

r g 0.105553

−

+ − − + + = − + = − + ≈

−

The final deduction of Miller (2007) is that WACC is more or less flawed and that NLWACC is needed.

In the following section 3, I put the NLWACC in perspective using the initial example. The analysis reveals why a traditional WACC can not work in this scenario. Also, I show how the challenger, the NLWACC, can be interpreted, and I demonstrate that the NLWACC is a rather narrow concept which is not superior to the WACC. 3. WACC and modified WACC revisited – before taxes First, it is helpful to remember that in this before tax-setting and under the assumptions set forth by Modigliani and Miller (1958) the capital structure irrelevance theorems hold. (To be absolutely clear, the just mentioned Nobel Prize laureate Merton H. Miller is not to be confused with Richard A. Miller who proposed NLWACC.) This implies that no additional value can be created while dividing the financing funds in a different manner. In other words, the

4 The present value then is $200,384.42 which is marginally higher than the true present value of $200,000. 5 See (Miller, 2007), p. 8, Eq. (23) for g = 0%. The derivation incorporating g is straightforward, and thus needs not be shown here.


47

weighted average cost of capital equal the unlevered cost of equity. Secondly, it is necessary to introduce a general return definition. A straightforward and commonly applied definition incurs the total return RT in period T6

T T T 1T

T 1

D V VR

V−

−

+ −= , (4)

where DT are inflows/outflows in period T (dividends, etc.), and VT is the market value at the end of period T. The total return R can be defined as a return for shareholders of levered (unlevered) projects re (ru), bondholders rd and for both claimholders combined as WACC. WACC with time-varying input parameters can be written down more generally than in Eq. (1) as follows:7

T d,T 1 d,T e,T 1 e,TWACC w r (1 t) w r− −= ⋅ ⋅ − + ⋅ (5)

One of the constituent characteristics of the WACC concept is that returns to shareholders and bondholders are weighted with their respective market value weights of the prior period: d,T-1

d,T-1e,T-1 d,T-1

Vw

V V=

+

,

and e,T-1 d,T-1w 1 w= − . It is important to emphasize

that the definition of WACC is based on market, and not on book value weights. Thus, discounting multi-period cash flows with a time-constant WACC implies ceteris paribus a constant relative capital structure over time. This is demonstrated for the initial example in Panel A of Table 1 showing the expected levels of equity and debt over time given the information at T = 0.

Based on the calculations in Panel A of Table 1 it is natural to compute the implied cash flows to shareholders and bondholders over the life of the project in Panel B. The results reveal two remarkable points: First, the cash flow when divided between bondholders and shareholders is sufficient to pay each group its necessary cash flow. Discounting these cash flows with their respective returns leads to their implied market values. Second, both flows do not conform to an annuity. In fact, the equity cash flows decrease while the debt cash flows increase over time. Therefore, forcing the cash flows to shareholders and bondholders to be annuities portrays a rather different scenario (see section 2, part 2)). This scenario II is investigated in Table 2.

Panel A of Table 2 now shows that market values (i.e., Ve,T, Vd,T, and VT) are identical with market values given in Panel A of Table 1 only at T = 0. At other valuation dates T, the values differ, hence exhibiting a different scenario than scenario I considered in Table 1. Scenario I (rebalancing at a constant ratio of debt to market value), and scenario II (rebalancing at an implicit time-varying ratio of debt to market value) are obviously not directly comparable.

6 See e.g. (Campbell et al., 1997), p. 12. 7 A derivation of Eq. (5) can be found in Appendix A.

Panel B of Table 2 analyzes the capital structure ratios over time in scenario II and highlights several implications for the returns. First, the debt ratio is expected to shrink over time. Appendix B shows that under plausible conditions this is generally true. It is only the same as in Panel A of Table 1 at T = 0. Second, this observation has an impact on the WACC returns according to Eq. (5). Under plausible conditions, WACC increases over time. Third, because the return to equity is supposed to be constant in this valuation exercise a very specific behaviour of the operating returns over time is implied when the capital structure is changing. Building on the (Modigliani and Miller, 1958) proposition 2, Eq. (6) postulates that the operating return ru,T is expected to behave over time as follows.

d,T 1e d

d,T 1 e,T 1e u,T u,T d u,T

d,T 1e,T 1

e,T 1

wr r

w wr r (r r ) r

ww1

w

−

− −

−−

−

+

= + − ⇔ =

+

This built-in feature seems not very appealing. Confirming the irrelevance theorem again, WACCT = ru,T proves to be true.

Panel B of Table 2 also explains why discounting with a time-constant WACC has to fail. The WACC is simply time-varying in scenario II. Textbooks usually do not emphasize that the WACC can also be time-varying.8 However, the use of a time-varying WACC has the advantage of an easy economic interpretation. A changing WACC points at a changing capital structure over time. The NLWACC does not provide this information. Thus, discounting debt and equity annuity cash flows ($38,247.23) with time-varying WACCs leads to consistent results shown in Panel A of Table 2.

( )N T

1 NT

jj

CFV

1 WACCτ− =τ

=τ

=+

∑∏

How can the modified WACC which was employed in section 2 be interpreted now? Instead of using time-varying WACCs, the NLWACC allows to calculate the present value at T = 0

• with a single (amalgamated) discount rate • if the cash flows to shareholders and

bondholders exhibit an annuity structure, and • if the operating cost of capital incidentally

follow a specific structure to ensure a constant equity return over time (as in the no tax-case shown by Eq. (6)).

8 See, for example, (Brealey et al., 2008), (Copeland et al., 2005), (Daves et al., 2004), (Lundholm and Sloan, 2004), (Stowe et al., 2007), (Titman and Martin, 2008), (Koller et

al., 2005).


48

Table 1. Scenario I (before taxes)

Panel A: Total market value, equity market and debt market value over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

CFT (in $) 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86

VT (in $) 200,000.00 182,826.17 163,849.06 142,879.35 119,707.82 94,103.28 65,810.26 34,546.48 0.00

Ve,T (in $) 150,000.00 137,119.63 122,886.79 107,159.51 89,780.86 70,577.46 49,357.70 25,909.86 0.00

Vd,T (in $) 50,000.00 45,706.54 40,962.26 35,719.84 29,926.95 23,525.82 16,452.57 8,636.62 0.00

we 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 -

wd 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 -

CFT is the aggregated cash flow to shareholders and bondholders during the period T. VT is the total market value (debt and equity, Vd,T + Ve,T) at date T. Ve,T is the equity market value (we⋅VT) at date T, and Vd,T is the debt market value (wd⋅VT) at date T. No taxes t = 0, WACC = ru = wd⋅rd + we⋅re = 0.25⋅0.06 + 0.75⋅0.12 = 0.105, and N = 8. VT is the present value of CFT given WACC. we is the equity weight (equity/total value), and wd is the debt weight (debt/total value). The weights are constant over time.

Panel B: Implied cash flows to shareholders and bondholders over the life cycle of the project

T 1 2 3 4 5 6 7 8

CFe,T (in $) 30,880.40 30,687.19 30,473.70 30,237.79 29,977.11 29,689.06 29,370.76 29,019.04

CFd,T (in $) 7,293.46 7,486.67 7,700.16 7,936.07 8,196.75 8,484.80 8,803.10 9,154.82

CFT (in $) 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86 38,173.86

CFe,T is the implied cash flow to shareholders during the period T: CFe,T = CFT – CFd,T. CFd,T is the implied cash flow to bondholders during the period T based on Panel A: CFd,T = Vd,T-1⋅rd + (Vd,T-1 – Vd,T). The aggregated cash flow to shareholders and bondholders during the

period T is CFT = CFe,T + CFd,T. Discounting these flows at re = 0.12 for CFe,T, and rd = 0.06 for CFd,T leads to the same values of Ve,T and Vd,T as in Panel A.

Table 2. Scenario II (before taxes)

Panel A: Annuity cash flows, and implied market values over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

CFe,T (in $) 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43

CFd,T (in $) 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80

Ve,T (in $) 150,000.00 137,804.59 124,145.71 108,847.77 91,714.07 72,524.33 51,031.82 26,960.21 0.00

Vd,T (in $) 50,000.00 44,948.22 39,593.31 33,917.11 27,900.34 21,522.56 14,762.11 7,596.04 0.00

VT (in $) 200,000.00 182,752.81 163,739.02 142,764.88 119,614.41 94,046.89 65,793.93 34,556.24 0.00

CFe,T and CFd,T is the annuity cash flow to shareholders and bondholders, respectively. The implied equity market value Ve,T at date T is the present value given CFe,T and re = 0.12. The implied debt market value Vd,T at date T is the present value given CFd,T and rd = 0.06.

VT is the total market value of debt and equity (Vd,T + Ve,T) at date T.

Panel B: Implied capital structure ratios and returns over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

we,T 0.750 0.754 0.758 0.762 0.767 0.771 0.776 0.780 -

wd,T 0.250 0.246 0.242 0.238 0.233 0.229 0.224 0.220 -

WACCT = ru,T 0.10500 0.10524 0.10549 0.10575 0.10600 0.10627 0.10654 0.10681

we,T is the time-varying equity weight (equity/total value), and wd,T is the time-varying debt weight (debt/total value) based on Panel A. The time-varying WACCT is computed via Eq. (5) as WACCT = wd,T-1⋅rd + we,T-1⋅re = wd,T-1⋅0.06 + we,T-1⋅0.12, and is confirmed by Eq. (6) calculating ru,T. Applying Eq. (7), the total market values of Panel A are confirmed by discounting CFT (= CFe,T + CFd,T) with WACCT

The NLWACC certainly is not a return as defined by Eq. (4). It is an amalgamation of time-varying returns into one discount rate, as typically is the case with internal rate of return procedures. This limits its economic interpretation. If one wants to use NLWACC also for prospective valuation at dates T > 0, it has to be updated, that is, N in Eq. (3) has to be updated. The conditions set forth in scenario II are rather limiting (even when allowing for g ≠ 0%) and only then the use of NLWACC is admissible. Clearly, the NLWACC does not help in situations where cash flows are not annuities. A time-varying WACC does

not share these limitations. Its use leads to consistent results.

However, it is not clear why under scenario II the flow to equity approach is not used instead. This would be the natural choice. The NLWACC (and WACC) seem like a detour.

Under scenario I, WACC is definitely the right choice, and even under scenario II a time-varying WACC (keeping the linear structure), solves the problem as does the NLWACC. WACC obviously is a technique better able to handle more general situations than NLWACC.


49

4. Analysis after taxes

Modigliani and Miller (1958) have shown in their seminal work that without taxes the WACC concept does not render any additional insights in comparison to the unlevered cost of capital. The after tax-WACC is covered lengthy in Miller's (2007) paper. Thus, to reconcile the argument taxes have to be considered. Again, in scenario I, an after tax-annuity is discounted with a constant after tax-WACC to arrive at V0 = $200,000, which implies ceteris paribus, a constant relative capital structure over time. The after tax-annuity now is $37,488.80. This is demonstrated for the initial example in Panel A of Table 3 showing the expected levels of equity and debt over time given the information at T = 0.

Based on the calculations in Panel A of Table 3 it is obvious to compute the implied cash flows to shareholders and bondholders over the life of the project in Panel B. The results again reveal that actually both flows do not conform to an annuity. In fact, the equity cash flows decrease while the debt cash flows increase over time. Therefore, forcing the cash flows to shareholders and bondholders to be annuities portrays a rather different scenario which is investigated further in Table 4. To strengthen the argument, an alternative calculation based on the Adjusted Present Value technique (APV) is provided in Panel C. As is well known, the operating value (unlevered value) Vu and the value of the debt tax shield Vt are valued separately. Unlike in the no tax-case, it is now important to specify which financing policy is pursued. Different debt tax policies will require a case-by-case discussion. For the purpose of illustration, I consider a value-based debt policy only.9 Miles and Ezzell (1980) and Kruschwitz and Löffler (2006) show for this policy:

( )

d,T-1de d

d,T-1 d e,T-1de u,T u,T d u,T

d,T-1d e,T-1 d

d e,T-1

wr tr r 1

w 1 r wr tr r r r 1 r

w1 r w r t1 1

1 r w

+ ⋅ − ⋅

+ = + − ⋅ − ⋅ ⇔ = + + − ⋅

+

(8) The unlevered value is computed with the unlevered cost of capital as follows:

( )N T

u, 1 NT

u, jj

CFV

1 rτ− =τ

=τ

=+

∑∏

(9)

In the special case of a time-constant ru, the formula is easier:

( )

N Tu, 1 T 1T

u

CFV

1 rτ− −τ+=τ

=+

∑.

The value of the debt tax shield is risky given this financing policy, and is therefore discounted with the unlevered cost of equity apart from the cash flow of the previous period which is certain.

( )( )

N d,T-1 dt, 1 N 1T

u, j dj

V r t V

1 r 1 rτ− −=τ

=τ

⋅ ⋅=

+ +∑

∏ (10)

9 For other policies see (Kruschwitz and Löffler, 2006).

Simplifying with a time-constant ru leads to

( ) ( )

N d,T-1 dt, 1 TT

u d

V r t V

1 r 1 rτ− −τ=τ

⋅ ⋅=

+ +∑ .

The APV approach confirms the consistency of the calculation in Panel C of Table 3. To arrive at the equity value, the value of debt has to be subtracted from the total market value (Vu + Vt).

Under scenario II (forced annuity structure) Panel A of Table 4 exhibits that market values are identical with market values given in Panel A of Table 3 only at T = 0.10 For other valuation dates T, values are different, hence exhibiting a different scenario than scenario I considered in Table 3. Scenario I (rebalancing at a constant ratio of debt to market value) and scenario II (rebalancing at an implicit time-varying ratio of debt to market value) are, again, not directly comparable.

Panel B of Table 4 analyzes the capital structure ratios over time in scenario II and highlights several implications for the returns. First, the debt ratio is expected to shrink over time again. It is only the same as in Panel A of Table 3 at T = 0. Second, this observation has an impact on the WACC returns according to Eq. (5). WACC increases over time. Third, because in this valuation exercise the return to equity is supposed to be constant a very specific behaviour of the operating returns over time is implied when the capital structure is changing. The unlevered cost of equity increases over time. Again, this built-in feature seems not very appealing.

Panel B of Table 4 also explains why discounting with a time-constant WACC has to fail. WACC is simply time-varying in scenario II. The NLWACC does not provide any information about the changing capital structure. Thus, discounting unlevered cash flows with time-varying WACCs leads to results consistent with Panel A of Table 4. The arguments already established in section 3 can be analogously repeated. Additionally, Miller’s (2007) claim that in the tax-case the integration of the interest tax shield in the WACC formula seems misplaced does not have to be followed. The time-varying WACC approach proves this fact. The APV calculation in Panel C of Table 4 confirms that operating returns are time-varying, indeed, and also confirms the results of Panel A and B. For example, the value of the debt tax shield at the beginning of period 6 is calculated based on Eq. (10) as:

t ,6

295.21 151.91V 407.99

1.06 1.10674 1.06= + =

⋅.

The analysis underlines that in the tax-case the

situation is not getting any better for the NLWACC and the aforementioned results hold in principle.

10 Panel A of Table 4 is identical with Panel A of Table 2. Therefore, also the weights in Panel B of Tables 2 and 4 are the same. The returns are different, of course, as different tax situations are considered in Table 2 and 4.


50

Table 3. Scenario I (after taxes) Panel A: Total market value, equity market and debt market value over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

CFT (in $) 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80

VT (in $) 200,000.00 182,511.18 163,273.50 142,112.05 118,834.4

5 93,229.10 65,063.21 34,080.73 0.00

Ve,T (in $) 150,000.00 136,883.38 122,455.12 106,584.04 89,125.84 69,921.82 48,797.40 25,560.55 0.00

Vd,T (in $) 50,000.00 45,627.79 40,818.37 35,528.01 29,708.61 23,307.27 16,265.80 8,520.18 0.00

we 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 -

wd 0.250 0.250 0.250 0.250 0.250 0.250 0.250 0.250 -

CFT is the unlevered cash flow during period T. VT is the total market value (debt and equity, Vd,T + Ve,T) at date T. Ve,T is the equity market value (we⋅VT) at date T. Vd,T is the debt market value (wd⋅VT) at date T. Taxes are considered with t = 0.3333, WACC = wd⋅rd⋅(1-t) + we⋅re = 0.25⋅0.06⋅0.6667 + 0.75⋅0.12 = 0.10, and N = 8. VT is the present value of CFT given WACC at date T. we is the equity weight (equity/total value), and wd is the debt weight (debt/total value). The weights are constant over time. Panel B: Implied cash flows to shareholders and bondholders over the life cycle of the project

T 1 2 3 4 5 6 7 8

CFd,T (in $) 7,372.20 7,547.09 7,739.47 7,951.08 8,183.86 8,439.91 8,721.57 9,031.39 Vd,T-1⋅rd⋅(1-t) + (Vd,T-1 – Vd,T) (in $) 6,372.20 6,634.53 6,923.10 7,240.52 7,589.68 7,973.76 8,396.25 8,860.99

CFe,T (in $) 31,116.60 30,854.27 30,565.70 30,248.28 29,899.12 29,515.04 29,092.55 28,627.81 CFT (in $) 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80 37,488.80

CFd,T is the implied cash flow to bondholders during the period T based on Panel A: CFd,T = Vd,T-1 rd + (Vd,T-1 – Vd,T). To compute the unlevered cash flow CFT Vd,T-1⋅rd⋅(1-t) + (Vd,T-1 – Vd,T) and CFe,T is needed. CFe,T is the implied cash flow to shareholders during the period T: CFe,T = CFT – CFd,T. Discounting CFe,T with re = 0.12 leads to Ve,T, and discounting CFd,T with rd = 0.06 leads to Vd,T. Panel C: Adjusted Present Value and market values over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

Vu,T (in $) 196,260.03 179,420.39 160,809.00 140,239.44 117,505.69 92,380.04 64,610.85 33,919.97 0.00

Vd,T-1⋅rd⋅t (in $) 1,000.00 912.56 816.37 710.56 594.17 466.15 325.32 170.40

Vt,T (in $) 3,739.95 3,090.79 2,464.50 1,872.61 1,328.77 849.05 452.36 160.76 0.00

VT = Vu,T + Vt,T (in $) 200,000.00 182,511.18 163,273.50 142,112.05 118,834.45 93,229.10 65,063.21 34,080.73 0.00

For illustration purposes a value-based debt policy is assumed. The unlevered firm value Vu,T is computed by discounting CFT based on Eq. (9) with a time-constant ru = 0.10521 according to Eq. (8) . The debt tax shield Vd,T-1⋅rd⋅t is discounted with ru and rd to arrive at the value of the debt tax shield Vt,T according to Eq. (10). VT is the total market value (Vu,T + Vt,T) at date T.

Table 4. Scenario II (after taxes)

Panel A: Annuity cash flows, and implied market values over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

CFe,T (in $) 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43 30,195.43

CFd,T (in $) 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80 8,051.80

Ve,T (in $) 150,000.00 137,804.59 124,145.71 108,847.77 91,714.07 72,524.33 51,031.82 26,960.21 0.00

Vd,T (in $) 50,000.00 44,948.20 39,593.30 33,917.10 27,900.33 21,522.55 14,762.11 7,596.04 0.00

VT (in $) 200,000.00 182,752.79 163,739.01 142,764.87 119,614.4

0 94,046.88 65,793.92 34,556.24 0.00

CFe,T and CFd,T is the annuity cash flow to shareholders and bondholders, respectively. The implied equity market value Ve,T at date T is the present value given CFe,T and re = 0.12. The implied debt market value Vd,T at date T is the present value given CFd,T and rd = 0.06. VT is the total market value of debt and equity (Vd,T + Ve,T) at date T. Panel B: Implied capital structure ratios and returns over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

we,T 0.750 0.754 0.758 0.762 0.767 0.771 0.776 0.780 -

wd,T 0.250 0.246 0.242 0.238 0.233 0.229 0.224 0.220 -

WACCT 0.10000 0.10032 0.10066 0.10099 0.10134 0.10169 0.10205 0.10242

ru,T 0.10521 0.10545 0.10570 0.10595 0.10621 0.10647 0.10674 0.10701

we,T is the time-varying equity weight (equity/total value), and wd,T is the time-varying debt weight (debt/total value) based on the results of Panel A. The time-varying WACCT is computed with Eq. (5) as WACCT = wd,T-1⋅rd⋅(1-t) + we,T-1⋅re = wd,T-1⋅0.06⋅0.6667 + we,T-1⋅0.12. The implied unlevered cost of capital ru,T is calculated for illustration purposes in line with a value-based debt policy according to Eq. (8).


51

Panel C: Adjusted Present Value and market values over the life cycle of the project

T 0 1 2 3 4 5 6 7 8

CFT (in $) 37,247.33 37,348.35 37,455.44 37,568.95 37,689.28 37,816.82 37,952.01 38,095.32

Vu,T (in $) 196,383.29 179,798.09 161,410.14 141,015.75 118,387.6

4 93,272.10 65,385.93 34,412.93 0.00

Vd,T-1⋅rd⋅t (in $) 999.90 898.87 791.79 678.27 557.95 430.41 295.21 151.91

Vt,T (in $) 3,616.73 2,954.71 2,328.87 1,749.11 1,226.76 774.78 407.99 143.31 0.00

VT = Vu,T + Vt,T (in $) 200,000.00 182,752.79 163,739.01 142,764.87 119,614.4

0 94,046.88 65,793.92 34,556.24 0.00

CFT is the unlevered cash flow during period T computed as CFT = CFe,T + Vd,T-1⋅rd⋅(1-t) + (Vd,T-1 – Vd,T); see Panel A, rd = 0.06, t = 0.3333. The unlevered value Vu,T is computed according to Eq. (9) with ru,T. The debt tax shields Vd,T-1⋅rd⋅t are discounted according to Eq. (10) with ru,T and rd to arrive at the value of the debt tax shield Vt,T. VT is the total market value (Vu,T + Vt,T) at date T.

5. Summary and outlook The claim that the NLWACC is conceptually superior to a constant WACC seems for two reasons rather artificial: 1) A slightly modified WACC which is time-varying is conceptually superior to the NLWACC. It does not seclude itself from an economic interpretation. 2) Given the special valuation scenario for which the NLWACC is motivated, the flow to equity approach is the direct solution while the (NL)WACC is a detour. Therefore, the foundations of WACC are sound. However, valuation is still a field which has many promising research questions to offer. Just to name a few: Which debt policies can be empirically supported? Given its autarkic nature, this is a question the WACC does not necessarily have to approach. How do more realistic tax regimes with personal taxes influence tax shield valuation? Which terminal value calculations are plausible? These and others seem to be more pressing questions which deserve further attention in future research. References

1. Arnold, G.C. and Hatzopoulos, P.D. (2000), “The

Theory-Practice Gap in Capital Budgeting: Evidence from the United Kingdom”, Journal of Business

Finance and Accounting, 27, pp. 603-626. 2. Arzac, E.R. and Glosten, L.R. (2005), “A

Reconsideration of Tax Shield Valuation”, European

Financial Management, 11, pp. 453-461. 3. Booth, L. (2007), “Capital Cash Flows, APV and

Valuation”, European Financial Management, 14, pp. 29-48.

4. Brealey, R.A., Myers, S.C. and Allen, F. (2008), Principles of Corporate Finance (9th ed.), Mc Graw-Hill Irwin, New York.

5. Bruner, R.F., Eades, K.M., Harris, R.S. and Higgins, R.C. (1998), “Best Practices in Estimating the Cost of Capital: Survey and Synthesis”, Financial Practice

and Education, Spring/Summer, pp. 13-28. 6. Campbell, J.Y., Lo, A.W. and MacKinlay, A.C.

(1997), The Econometrics of Financial Markets, Princeton University Press, Princeton, NJ.

7. Cooper, I.A. and Nyborg, K.G. (2006), “The value of tax shields is equal to the present value of tax shields”, Journal of Financial Economics, 81, pp. 215-225.

8. Cooper, I.A. and Nyborg, K.G. (2007), “Valuing the debt tax shield“, Journal of Applied Corporate

Finance, 19.2, pp. 50-59. 9. Cooper, I.A. and Nyborg, K.G. (2008), “Tax-Adjusted

Discount Rates with Investor Taxes and Risky Debt”, Financial Management, pp. 365–379.

10. Copeland, T.E., Weston, J.F. and Shastri, K. (2005), Financial Theory and Corporate Policy (4th ed.), Pearson Addison Wesley.

11. Daves, P.R., Ehrhardt, M.C. and Shrieves, R.E. (2004), Corporate Valuation: A Guide for Managers

and Investors, Thomson South-Western. 12. Fernandez, P. (2004), “The value of tax shields is

NOT equal to the present value of tax shields”, Journal of Financial Economics, pp. 145-165.

13. Fieten, P., Kruschwitz, L., Laitenberger, J., Löffler, A., Tham, J., Vélez-Pareja, I. and Wonder, N. (2005), “Comment on „The value of tax shields is NOT equal to the present value of tax shields“, Quarterly Review

of Economics and Finance, pp. 184-187. 14. Graham, J.R. and Harvey, C.R. (2001), “The Theory

and Practice of Corporate Finance: Evidence from the Field”, Journal of Financial Economics, 60, May, pp. 187-243.

15. Koller, T., Goedhart, M. and Wessels, D. (2005), Valuation: Measuring and managing the value of

companies (4th ed.), John Wiley & Sons, Hoboken, NJ.

16. Kruschwitz, L. and Löffler, A. (2006), Discounted

Cash Flow: A Theory of the Valuation of Firms, John Wiley & Sons, Chichester, West Sussex.

17. Lobe, S., Niermeier, T., Essler, W. & Röder, K. (2008). Do Managers follow the Shareholder Value Principle when applying Capital Budgeting Methods? A Comparison of Theory and Practice based on German Survey Results and Return Data. Center of

Finance UR Working Paper.

18. Lundholm, R.J. and Sloan, R.G. (2004), Equity

Valuation and Analysis with eVal, Mc Graw-Hill Irwin, New York.

19. Massari, M., Roncaglio, F. and Zanetti, L. (2008), “On the Equivalence between the APV and the wacc Approach in a Growing Leveraged Firm”, European

Financial Management, 14, pp. 152–162. 20. Miles, J.A. and Ezzell, J.R. (1980), “The Weighted

Average Cost of Capital, Perfect Capital Markets, and Project Life: A clarification”, Journal of Financial

and Quantitative Analysis, 15, pp. 719-730. 21. Miller R.A. (2007), “The weighted average cost of

capital is not quite right”, forthcoming in The

Quarterly Review of Economics and Finance, doi:10.1016/j.qref.2006.11.001


52

22. Modigliani, F. and Miller, M.H. (1958), “The Cost of Capital, Corporation Finance and the Theory of Investment”, The American Economic Review, 48, pp. 261-297.

23. Ruback, R.S. (2002), “Capital Cash Flows: A Simple Approach to Valuing Risky Cash Flows”, Financial

Management, pp. 85-103.

24. Stowe, J.D., Robinson, T.R., Pinto, J.E. and McLeavey, D.W. (2007), Equity Asset Valuation, John Wiley & Sons, Hoboken, NJ.

25. Titman, S. and Martin, J.D. (2008), Valuation. The Art

& Science of Corporate Investment Decisions, Pearson Addison Wesley.

Appendix A: Derivation of time-varying WACC The following derivation draws on the proof of a time-constant WACC by Brealey et al. (2008), p. 533. Starting with the

value in the next to last period: N-1 e,N-1 d,N-1V V V= + . The total cash flow to debt and equity investors is the cash flow CF

plus the interest tax shield: N d,N d,N 1CF t r V −+ ⋅ ⋅ . This total cash flow can also be written based on returns:

( )d,N-1 e,N-1N 1 d,N e,N N 1 d,N d,N-1 e,N e,N-1

e,N-1 d,N-1 e,N-1 d,N-1

V VV 1 r r V 1 r w r w

V V V V− −

+ + = + + + +

Equate both definitions and solve for VN-1: N N

N 1d,N d,N-1 e,N e,N-1 N

CF CFV

1 r (1 t)w r w 1 WACC− = =+ − + +

The WACC-definition of Eq. (5) shows up. This can be repeated for VN-2. As the return-definition is based on Eq. (4) the next period’s payoff includes VN-1:

N 1 d,N 1 d,N 2 N 1CF t r V V− − − −+ ⋅ ⋅ + ( )N 2 d,N 1 d,N-2 e,N 1 e,N-2V 1 r w r w− − −= + +

Solving for VN-2 yields:

( )( )

N 1 N 1 N 1 N 1N 2

d,N 1 d,N-2 e,N 1 e,N-2 N 1

N 1 N

N 1 N 1 N

CF V CF VV

1 r (1 t)w r w 1 WACC

CF CF

1 WACC 1 WACC 1 WACC

− − − −−

− − −

−

− −

+ += =

+ − + +

= ++ + +

By recursion, one obtains for any arbitrary valuation date τ–1:

( )N T

1 NTjj

CFV

1 WACCτ− =τ

=τ

=+

∑∏

(7)

Appendix B: Is the debt to total market value ratio falling over time? To show the conditions for this relationship, I compare growth rates of debt and equity market value, e.g. the equity growth

rate is: e,T e,T 1T e

e,T 1

V Vg (V )

V−

−

−= . If debt is not growing as strong as equity, the debt ratio has to shrink. Inserting Eq. (2) yields

the following growth rate:

N T

N T 1

e e

e e

ee eT e

eee

e

N T N T 1

N T N T 1

N T 1

N T 1

(1 g) (1 g)CF 1 CF 1

(1 r ) (1 r ) 1 g1

1 rr g r gg (V ) 1

(1 g) 1 gCF 1 1

1 r(1 r )

r g

−

− +

− − +

− − +

− +

− +

+ + − − + + + −−

+− − = = − + + − −

+ + −

For the debt growth rate equivalently:

N T

N T 1

dT d

d

1 g1

1 rg (V ) 1

1 g1

1 r

−

− +

+−

+ = − +

− +

Under realistic conditions ( d er r< ) equity growth rates are higher than debt growth rates. Therefore, wd,T decreases over

time.11

11 Except for N → ∞, and T = N. Then, gT(Vd) = gT(Ve).


53

DEBT OF FAMILY FIRMS: A COMPARISON BASED ON ACCOUNTING

INDICATORS

Olivier Colot*, Mélanie Croquet**

Abstract The financial behavior of family firms represents a field of research that has been little explored up to the present time. In this context, we wanted to contribute to understanding the problems linked to financing in family firms, and more specifically to family SMEs in Belgium, because they represent a major part of the Belgian economy. This paper uses paired samples methodology to compare the financial debt of family SMEs with that of non-family SMEs. The results show that family SMEs seem to be more indebted. Among all of the ratios tested, six indicators were statistically significant. The theory of the agency cost of debts seems to be confirmed for Belgian family SMEs. Keywords: family firm, debt, SMEs *UMH, Warocqué Management faculty, Place du Parc 20, 7000 Mons, Belgium tel: 003265/37.32.80 (fax: 003265/37.30.54) [email protected] **Ph.D. in Business Administration, UMH, Warocqué Research Center

1. Introduction

According to Kenyon-Rouviniez and Ward (2004), family firms play a very important role in the economy of most industrialized countries, and they represent between 50 and 90 percent of the gross domestic product of all market economies. In the private sector, family firms represent the most common type of business (Donckels & Frohlich, 1991; Cromie et al., 1995; la Porta et al., 1999; Faccio & Lang, 2002; Ifera, 2003; Morck & Yeung, 2003; Kenyon-Rouviniez & Ward, 2004). Despite this, the literature on family firms has developed in scientific research only since the Eighties (Hirigoyen, 1985; Merigot & Hirigoyen, 1987; Kalika, 1988; etc.), and more during the Nineties (Donckels & Frohlich, 1991; Friedman & Friedman, 1994; Allouche & Amann, 1995; Allouche & Amann, 2000; etc.).

The financing problem of family firms has been covered very little by researchers. Only 14% of family firm studies are focused on its financing problem (Allouche & Amann, 2000, p.4; Romano et al., 2000, p. 285). Chua & alii (2003) made a study of 190 papers on family firms that were published between 1996 and 2003. They noticed that 22% of the papers have to do with research in succession, 15% in performance, 10% in governance, 6% in advantages and 6% in conflicts. Entrepreneurship, culture, and strategy are the topics of only 5% of the papers. Finally, the internationalization of family firms represents 3% of the papers and professionalization represents only 2%.

Along these lines, we wanted to help improve knowledge about family firms while focusing on financing behavior, because family firms represent a main part of the Belgian economy (Donckels, 1989; Donckels & Aerts, 1993; Wtterwulghe |et alii, 1994; Van Caillie & Denis, 1996; Colot, 2005). As some authors (Kalika, 1988; Allouche & Amann, 1995; Schulze & alii, 2003) think that it is necessary to consider the family firm as an entity different from the non-family firm, we wanted to verify if family firms were more or less indebted than non-family firms.

Different kinds of financing are possible in this context: self-financing, debt and the issuing of new shares.

In this paper, we will not study the possibility for a firm to open its capital to find financing because family firms will call upon capital opening only as a last resort. This means of financing can involve a loss of family control if the family does not take part in the operation.

Consequently we chose to speak only about the financial indebtedness of Belgian family firms. This financial indebtedness represents an alternate means of financing that makes it possible to avoid the dispersion of ownership and the loss of family control.

Must the family firm be considered as an entity different from the non-family firm? If the answer is yes, then it would be interesting to consider family firm financing behavior as a function of its characteristics. This would mean that the family character of a firm can have an influence on its


54

indebtedness, and consequently, it would mean that there is a difference between debt levels of family and non-family firms.

For our future argumentation, we will put forward that there is a difference between family and non-family firms and that shareholding and management characteristics of family firms have a significant influence on their indebtedness.

We will use the conclusions of three theories that can be applied to family firms. These include the theory of control, the agency costs of debt theory and the agency costs of stocks theory.

We will test our hypotheses on a sample of family firms and we will confront our results with the theory of control, the agency costs of stocks theory and the agency costs of debt theory.

These three major theoretical thrusts make it possible to explain the financing behavior of family firms. The first two lead to the conclusion that family firms are less indebted than non-family firms and the last confirms the inverse relation.

Thanks to the statistical pairing technique, we will try to determine which of these theories can answer our question.

2. Financing behavior of family firms

The characteristics of family firm management and shareholding may have an influence on the decision of how to finance them. These criteria are used to identify family firms: family ownership, the very significant family implication in the firm’s management and, finally, the will to transfer it to the next generation. Consequently, a firm could be considered as a family firm if and only if the same family owns and manages it in order to safeguard it in the family inheritance. This definition is also defended by Anderson & Reeb (2003); they note that family firm shareholders are strongly affected by the long term survival of their firm. Family control and continuity (Casson, 1999) are at the heart of the family firm issue. In this respect, the theory of control seems to apply to the case of family firms.

In his study, Hirigoyen (1982) highlighted a careful financial strategy of family firm managers and also an objective of continuity of the firms. These results are confirmed by Ward (1988), who shows that family firms adopt defensive strategies to avoid the loss of family control. So it is not unrealistic to think that family shareholders try to reduce the firm’s global risk in order to preserve its long term survival. To do that, shareholders probably count on the bankruptcy risk not rising. An increase in financial indebtedness and thus an increase in the risk of bankruptcy might be seen by family shareholders as a loss of family control to creditors. Nevertheless, the self-financing capacity being limited, family firms must find other means of financing. Consequently, they prefer having banking debts to opening their capital (Calof, 1985; Wtterwulghe, 1998).

Family firms would tend to be more involved in debt. In this regard, Friend & Lang (1988) and Friend & Hasbrouk (1987) notice an inverse relation between ownership concentration and indebtedness level. This risk aversion is also reinforced by the will to transfer the family firm to the next generation (Friedman & Friedman, 1994). Gallo & Vilaseca (1996) also showed that family firms have a weak debt ratio that can be explained by the fear of bankruptcy or loss of control of the family firm. However, because the principal values of family firms are based on their independence and their long term survival, Ben Jemaa (2005) asserts that the level of short term banking debt would have to be lower than that of non-family firms. Notwithstanding, we believe that the short term operating indebtedness should have to be higher within family firms than non-family firms. This short term operating indebtedness represents a means of financing that is less costly or risky than short term banking debts.

According to this literature, we can propose the following hypotheses:

Hypothesis 1: the global financial debt of family firms is lower than that of non-family firms.

Hypothesis 2: the short term financial debt of family firms is lower than that of non-family firms.

Hypothesis 3: family firms have a higher level of short term operating debts than do non-family firms.

The first hypothesis refers to the Agency Theory (Jensen & Meckling, 1976), which is based on the neoclassical principle that different kinds of agents try to maximize their own interests. Nevertheless, family shareholders are strongly implicated in family firm management by reason of the weak separation between ownership and control. Consequently, the risk of agency conflicts between managers and owners is weaker if there is a strong implication of family control in the firm’s management. So indebtedness cannot be retained as a means of reducing managers’ opportunism. On the contrary, the private gains that are obtained by family shareholders thanks to their control of the firm’s management could be reduced because of indebtedness. Indeed, indebtedness increases the risk of bankruptcy but also constrains the firm to external monitoring. This external monitoring can involve a reduction in the private gains resulting from control (Harris & Raviv, 1988). Consequently family shareholders do not have an interest in putting the firm into debt. The first hypothesis is thus confirmed by the Agency Theory because, within family firms, indebtedness does not appear as a means to reduce the potential conflicts between shareholders and managers. On the other hand, another current in literature that also comes from the Agency Theory certifies the contrary. It is based on the conflicts that can exist between shareholders and creditors. Indeed the relations between these two categories can create conflicts if the shareholders misappropriate a part of the firm’s substance at the expense of creditors (Desbrières & Dumontier, 1989). This misappropriation can occur


55

through investments. So, imagine a non-family firm that has a very high level of debts. This level of debts comes from a loan contracted by the firm in order to finance an investment project. Suppose that firm has to pay off a sum of X thousand dollars (for X>0) at a given period t to its creditors.

Managers are supposed to have a choice between two investment projects: P1 and P2. For each of these

projects, there are two possible outcomes: one pessimistic and the other optimistic.

The probability of occurrence of each outcome is 0.5. Then, flows of money from each of the two projects are presented in table 1. After the loan repayment, the balance is distributed to the shareholders.

Table 1. Flows of money

. Pessimistic outcome Optimistic outcome

P1 X (X * z) with z > 1 P2 0 (X * y) with y = (1+z)

Expected flows of project P1 = 0.5 * X + 0.5 * (X * z) = 0.5 * X * [1 + z] Expected flows of project P2 = 0.5 * (X * y)

Because y =(1+z), the expected flow of

project P1 is equal to project P2. However, project P2 is more risky than project P1. The risk of each of the projects can be calculated thanks to the standard deviation. So the standard deviation of project P1 (noted τ(P1)) amounts to 0.5*z while the standard deviation of project P2 (noted τ(P2)) rises to 0.5*X*y. Nevertheless we suppose that y = (1+z), thus τ(P1) < τ(P2).

If the firm chooses project P1 it will be able to fulfill its contractual obligations toward its creditors even if the pessimistic outcome occurs. Indeed, whichever outcome occurs, the firm will be able to create a sufficient expected flow to repay its debt of X thousand dollars to its creditors. On the other hand, if the firm invests in project P2, it will not be able to repay its debt in the case of a pessimistic outcome. The expected value of the debt is not more than X thousand dollars but 0.5 X thousand dollars. Thus for the firm’s creditors, project P1 would have to be selected by the managers.

Nevertheless, for the firm’s shareholders, it is the contrary. Indeed, if project P1 is selected, their earnings expectancy will amount to [0.5 * X * (z-1)] thousand dollars; if project P2 is selected their earnings will amount to [0.5 * X * (y-1)] thousand dollars. By replacing “y” with “(1+z)”, we will show that [0.5 * X * (y-1)] is higher than [0.5 * X * (z-1)]. Consequently shareholders may find it beneficial to convince managers to choose project P2. There will thus be a misappropriation of a part of the firm’s substance by shareholders.

The probability of occurrence is higher within non-family than family firms in which the interests of managers, shareholders and the family are often the same. Consequently we can imagine that creditors would lend more easily in the long term to family firms. Markin (2004) found that long term debts were higher within family than non-family firms. These conclusions complete the list of hypotheses above.

Hypothesis 4: The long term indebtedness of family firms is higher than that of non-family firms. 3. Methodology Our main objective is to determine if Belgian family firms have a higher financial indebtedness than Belgian non-family firms. Note that the case of Belgium is marginal in terms of studies carried out on family firms. 3.1. The sample In order to reach a large number of family firms, our study is based on SMEs, which are very often of a family nature (Donckels & Aerts, 1993; Wtterwulghe et alii., 1994; Van Gils et alii, 2004). Despite this, this kind of firm remains little studied (Van Gils et al., 2004, p. 588). Moreover, wanting to identify firms managed by the owner(s), the choice of SMEs asserted itself. Because the family character is not detectable in the financial report and financial statements of the firms, we made a survey by questionnaire. The population of this survey includes all Belgian SMEs (between 3 and 50 workers). Then, we chose our sample among Belgian SMEs created before December 31, 1990 so that the family character or lack thereof is quite impregnated in the firm. In addition, SMEs employing less than two workers were excluded. The population of this study accounted for 8917 firms for a total of 55284 jobs. Within this population, we made a simple random sample of 2000 SMEs. Our questionnaire was sent to the CEO or CFO of the firms. We sent a second questionnaire to firms from which we received no response. Our useable sample finally included 391 answered questionnaires, presenting a statistical representativeness of the Belgian SME population based on three criteria: geographical dispersion, the branch of industry, and the size (based on the number of workers).


56

We considered that the firm is a family business when it satisfies at least two of the following three criteria:

• a family holds at least 50% of the capital; • a family has a decisive influence on firm

strategy and succession (the majority of managers belong to the family);

• the majority of the board of directors is made up of members of a family.

This definition of family SME presents the advantage of using clear and measurable criteria, in opposition to qualitative definitions that are more subjective and arbitrary. Moreover, this definition of the family firm is in accordance with most recent studies (Floren 2002; Anderson and Reeb, 2003).

We thus observed that among the 391 firms from our sample, 318 can be regarded as family businesses (table 2).

Table 2. Family and non-family firms Total SMEs Family SMEs Non-family SMEs 391 318 81.33% 73 18.67%

The result (81.33% family SMEs) illustrates a

large majority of family firms and is consistent with previous papers related to the importance of family SMEs in Belgium (Wtterwulghe et al., 1994; Jorissen et alii., 2002). This percentage is, in addition, very close to the results obtained by Astrachan and Kolenko (1994) for the United States (90%), by Reidel (1994) for Germany (80%), and by Crouzet (1995), who shows that the percentage of family SMEs in the European Union varies from 75% to 99% according to the countries. 3.2. Observations on the sample The following tables synthesize some descriptive statistics about the sample. Note that only 385 firms were identified in Belfirst14. It seems necessary to make the test over several years in order to confirm or invalidate the stability of the results. 3.3. Test used

To compare the debt between family and non-family firms, we chose to work with paired samples. This method makes it possible to avoid demographic data like size and activity sector, for example. Indeed, according to Jorissen et alii (2002), comparative studies of family and non-family firms generally do not take into account this type of variable, which can, however, skew the results highlighting differences of management practices or performance between these two types of firms. In the same way, for Westhead and Cowling (1998), studies that do not control these

14 Financial data of firms required to publish their financial report and financial statements with the National Bank of Belgium.

variables do not make it possible to identify variations related to the family character of firms but rather related to dissimilarities due to demographic data of the sample.

To build the paired samples, it is necessary to choose criteria considered to be relevant, so as to make sure that the measured effect is due to studied variables and not to differences in the composition of samples (Thietart, 1999, p. 198). Other empirical studies using this method (Caby, 1994; Sapusek, 1998; Heldenbergh, 1999) indicate indeed that various accounting measurements of performance are sensitive to the sectorial membership and the size of the firm considered (Ooghe and Van Wymeersch,, 1990, p. 395). With regard to the choice of the criterion of size, the total asset was preferred (it is also one of the three references to the size of firms according to Belgian accounting law, with manpower and sales turnover). Comble (1994, pp. 371-272) showed that the interaction between the size and the activity makes it possible to obtain more reliable results from a comparative point of view.

The control sample will be determined by the following two criteria:

- the branch of industry; - the size: total assets +/- 20 %

These two criteria are applied to the data of the year under review (this year will be noted year n). When several firms correspond to the profile, we retain the firm that owns the closest total assets. If no firm corresponds to the profile, we simplify the mode of selection. So the NACEBEL code (branch of industry) with three numbers was retained for 12 firms and with two numbers for 23 firms.

The two samples were compared thanks to a statistical test that compares paired observations and identifies significant results. For each ratio, we systematically withdrew the value of the control firm from the corresponding value for the family firm. The test of comparison is in fact practiced on the average of the differences between paired values: the postulation to be tested is that these differences in debt are null while the alternative postulation affirms the existence of differences. This method does not require normality in the distributions (Afnor, 1988, p. 366), which is particularly interesting insofar as many ratios are not normally distributed (Ooghe and Van Wymeersch, 1990, p. 392). The comparison test is used on the average of the differences between paired values. Thus it is recommended to eliminate the illogical differences.

In fact, an important difference could be due to an element that is independent of family firm characteristics. We use the Cochran test to eliminate these differences. For each indicator, we calculated the difference in the firm pairs (the differences named Di). From the Di values, we calculated the Gi values (which are the relation between the Di2 of each pair and the sum of the Di2 of all the pairs) for each indicator. We eliminated the Di values for which the Gi was greater than 0.12.


57

Table 3. Sectors of activity

Activity sector Frequency Percent Industries (Nacebel: 01 � 41) 69 17.92% Building (Nacebel 45) 85 22.08% Trade (Nacebel 50 � 55) 169 43.90% Services (Nacebel : 65 � 93) 62 16.10% Total 385 100%

Table 4. Shareholding structure

Frequency Percent Cumulative Percent 100% by close family 248 64.42% 64.42% 100% by extended family 60 15.58% 80.00% 90% by extended family 11 2.86% 82.86% 50% by extended family 16 4.16% 87.01% + than 50% by other firms 16 4.16% 9117% - than 50% by other firms 32 8.31% 99.48% 50% family - 50% other family 2 0.52% 100.00% Total 385 100.00%

Table 5. Added value

Added value (in €) Number of firms Percent to 100,000 83 21.56% Between 100,001 and 250,000 133 34.55% Between 250,001 and 500,000 97 25.19% Between 500,001 and 750,000 37 9.61% Between 750,001 and 1,000,000 15 3.90% More than 1,000,001 20 5.19% Total 385 100%

The list of ratios in table 6 was established on the basis of the firms’ financial statements.

Table 6. Definition of indicators

Indicators Accounting code

Total debt ratio (|16|+|17/49|) / (<10/15>)

Total financial debt ratio (|1704|+|43|+|42|) / (<10/15>) Financial independence ratio (<10/15>) / (|16|+|17/49|) Long term debt ratio (|16|+|17|) / (<10/15>) Long term financial debt ratio (|1704|) / (<10/15>) Long term financial independence ratio (<10/15>) / (|16|+|17|) Coverage of long term debt by cash flow

(|70/67|-|67/70|+|630|+<631/4>+<635/7>+|6501|+<651>+|6560|-

|6561|+|660|+|661|+<662>+|663|+|680|-|760|-|761|-|762|-|780|-|9125|) / (|16|+|17|)

Short term financial debt ratio (|43|+|42|) / (<10/15>) Short term operating debt ratio (|42/48|-|43|) / (<10/15>) Self-financing quote (|13|+|140|-|141|) / (|10/49|) Current ratio (|29/58|-|29|) / (|42/48|+|492/3|)


58

4. Results and discussion

Table 7. Paired samples test

Differences between paired firms

T df Sig. (2-tailed)

Mean

Standard deviation

Sandard mean error

Confidence interval 95% of difference

Lower Upper Total financial debt ratio n 1.42136 9.85072 1.21254 -1.00025 3.84297 1.172 65 .245 Total financial debt ratio n-1 -.11591 5.68482 .69975 -1.51341 1.8159 -.166 65 .869 Total financial debt ratio n-2 .27031 2.78341 .34524 -.41939 .96000 .783 64 .437 Total financial debt ratio n-3 .32984 3.21319 .40165 -.47279 1.13247 .821 63 .415 Financial independence ratio n -.27115 1.17161 .16247 -.59733 .05502 -1.669 51 .101 Financial independence ratio n-1 -.01321 1.20360 .16533 -.34496 .31855 -.080 52 .937 Financial independence ratio n-2 -.35182 1.12349 .15149 -.65554 -.04810 -2.322 54 .024 Financial independence ratio n-3 -.31000 1.01042 .13502 -.58059 -.03941 -2.296 55 .026 Long term debt ratio n -.88000 4.15698 .63393 -2.15933 .39933 -1.388 42 .172 Long term debt ratio n-1 -.15660 2.43193 .35473 -.87064 .55745 -.441 46 .661 Long term debt ratio n-2 .20409 1.90994 .28793 -.37658 .78477 .709 43 .482 Long term debt ratio n-3 .20383 2.07277 .30234 -.40476 .81242 .674 46 .504 Long term financial debt ratio n .76409 7.78501 .95827 -1.14970 2.67789 .797 65 .428 Long term financial debt ratio n-1 -.37212 5.05911 .62273 -1.61581 .87156 -.598 65 .552 Long term financial debt ratio n-2 .25538 2.04713 .25391 -.25187 .76264 1.006 64 .318 Long term financial debt ratio n-3

.57585 2.64096 .32757 -.07855 1.23024 1.758 64 .084

Long term financial independence ratio n .78556 30.27619 4.51331 -8.31042 9.88153 .174 44 .863 Long term financial independence ratio n-1

1.75000 22.96067 3.34916 -4.99150 8.49150 .523 46 .604

Long term financial independence ratio n-2

-.50047 12.18486 1.85817 -4.25041 3.24948 -.269 42 .789

Long term financial independence ratio n-3

-.07936 6.80282 .99229 -2.07675 1.91802 -.080 46 .937

Coverage of long term debt by cash flow n 15.70543 178.99668 26.39162 -37.45002 68.86089 .595 45 .555 Coverage of long term debt by cash flow n-1

61.75826 197.20054 29.07563 3.19693 120.31959 2.124 45 .039

Coverage of long term debt by cash flow n-2

14.79128 198.58066 28.96597 -43.51419 73.09674 .511 46 .612

Coverage of long term debt by cash flow n-3

26.94320 193.73911 27.39885 -28.11685 82.00325 .983 49 .330

Short term financial debt ratio n .47092 2.22479 .27595 -.08035 1.02220 1.707 64 .093 Short term financial debt ratio n-1 .25697 1.42287 .17514 -.09282 .60676 1.467 65 .147 Short term financial debt ratio n-2 .01545 1.16129 .14294 -.27003 .30093 .108 65 .914 Short term financial debt ratio n-3 -.07246 1.29805 .16100 -.39410 .24918 -.450 64 .654 Short term operating debt ratio n -1.02415 7.55045 .93652 -2.89506 .84676 -1.094 64 .278 Short term operating debt ratio n-1 .70576 7.25200 .89266 -1.07701 2.48852 .791 65 .432 Short term operating debt ratio n-2 -.35500 5.06206 .62310 -1.59941 .88941 -.570 65 .571 Short term operating debt ratio n-3 .14953 4.17434 .52179 -.89319 1.19225 .287 63 .775 Current ratio n .14154 1.92720 .23904 -.33600 .61908 .592 64 .556 Current ratio n-1 .24016 1.69919 .21240 -.18429 .66460 1.131 63 .262 Current ratio n-2 .17508 1.63739 .20309 -.23065 .58080 .862 64 .392 Current ratio n-3 -.13833 1.67023 .20559 -.54893 .27226 -.673 65 .503

Legend: the mean corresponds to the mean difference, for each ratio, between family and control firms. t corresponds to the t of Student ; df corresponds to the degree of freedom ; Sig corresponds to the test’s significance

The results show that family firms seem to be

more involved in debt than their non-family counterparts in the majority of cases. Amongst all the indicators, six of them are statistically significant. Consequently we cannot generalize our conclusions to the population of all Belgian firms. Our conclusions will concern only our sample (which is a representative sample of the population).

The results presented in table 7 show that, on average, Belgian family firms have a total financial debt ratio higher than that of non-family firms. This

observation seems to contradict the first hypothesis. Furthermore, this result can be strengthened by the result obtained for the financial independence ratio. This ratio shows that family firms are less independent than non-family firms. The results obtained are statistically significant to a degree of 10%. The conclusions can therefore be generalized to the Belgian population of firms. Their self financing capacity being reached, those firms have to prefer to be in debt rather than to create new shares. The


59

issuing of shares can create a shareholding dilution and increase the risk of the loss of family control.

The results obtained for the short term financial debt ratio also show that family firms have a proportion of short term financial debt higher than that of non-family firms. This observation also contradicts the second hypothesis.

The results on the long term debt ratio, the long term financial independence ratio and the short term operating debt ratio are not statistically significant. We note that there are two positive differences and two negative differences between family and non-family firms. Because of the ambiguous results obtained for the “short term operating debt ratio”, we cannot confirm or invalidate hypothesis 3 (that family firms have more short term operating debt than non-family firms).

On the other hand, our fourth hypothesis can be confirmed: long term financial debt would seem higher within family firms than within non-family firms. The results obtained for the year n-3 are statistically significant to a degree of 10%. Consequently, our conclusion can be generalized to Belgian SMEs.

The cash flows of family SMEs seem to be higher than those of non-family firms. Thus, it would mean that, all things being equal, family firms are better able to cover their financial charges than non-family firms. This observation could explain why family firms have a higher level of long term financial debt than non-family firms (the coverage of long term debt by cash flow being better than that of non-family firms). This result is also confirmed by the current ratio. Indeed family firms have a better current ratio than non-family firms. This would seem to confirm that family firms are able to carry out their financial engagements. 5. Conclusions and development tracks According to the results in table 7, Belgian family SMEs seem to be more involved in debt than non-family firms. The results also show that the managers of family firms choose careful global financial strategies for fear of losing family control. The managers will prefer self financing to indebtedness and indebtedness to the issuing of new shares.

The issuing of new shares is the last alternative because it can induce shareholding dilution and increase the loss of family control. Consequently an arbitrage exists between the increase in the risk of default (in relation with a higher level of leverage) and the increase in the risk of losing family control.

The statistically significant results of this study seem to confirm this point of view. Managers of family firms will be involved in debt rather than issue new shares when their self financing capacity is reached. The principal motivation within family firms seems to be the maintaining of family control.

Consequently, the two first hypotheses cannot be validated. So the total financial debts are not lower within family than within non-family firms (hypothesis 1 not confirmed) and the short term financial debts are not lower within family than within non-family firms. In regard to the third hypothesis, we cannot draw a conclusion because the results presented in table 7 are not consistent.

On the other hand, we can validate the fourth hypothesis: the long term financial debt of family firms is higher than that of non-family firms. These results confirm the conclusions of the agency costs of debt theory. According to this theory, family firms would obtain long term financial credits more easily than their non-family counterparts thanks to their strategy and their long term vision.

In conclusion, the debt of Belgian family SMEs seems to be higher than that of non-family firms. Among all of the debt ratios tested, six indicators were statistically significant.

References 1. Afnor (1988), Statistique: vocabulaire, estimation et

tests statistique, Volume 1, Paris, p. 366. 2. Allouche J. & Amann B. (1995), “Le retour

triomphant du capitalisme familial”, in de Jacques

Cœur à Renault : Gestionnaires et Organisations, Presses de l’Université des sciences sociales de Toulouse, 23 pages.

3. Allouche J. & Amann B. (2000), “L’entreprise familiale: un état de l’art”, Revue Finance – Contrôle

– Stratégie, Vol. 3 n°1, March 2000, pp. 33-69. 4. Anderson R. & Reeb D. (2003), “Founding family

ownership and firm performance: evidence from the SetP 500”, The Journal of Finance, vol. 58, n°3, pp. 1301-1327.

5. Ben Jemaa S. (2005), “La structure financière des entreprises familiales tunisiennes”, Actes de Colloque. Colloque Medinnov. Marseilles.

6. Caby J. (1994), “Motivations et efficacité des offres publiques d’achat et d’échange en France de 1970 à 1990”, Doctoral dissertation, Université de Nancy II, 676 pages.

7. Calof J. (1985), “Analysis of small business owners financial preferences”, Journal of Small Business and

Entrepreneurship, 3(3), pp. 39-44. 8. Casson M. (1999), “The economics of the family

firm”, Scandinavian Economic History Review, n°47, pp. 10-23.

9. Chua J. H., Sharma P. & Chrisman J. J. (2003), “Current trends and future directions in family business studies: Toward a theory of the family firm”, Written as part of Coleman Foundation White Paper Series, 62 pages.

10. Colot O. (2005), “Guide de la transmission des entreprises familiales”, IFAPME, 247 pages.

11. Comble K. (1994), “Restructurations financières par leveraged management buy-out (LMBO). Approche du marché belge et impact sur les performances des enterprises”, Doctoral dissertation, Université de Mons-Hainaut, 468 pages.

12. Cromie S., Stephenson B. & Monteith D. (1995), “The management of family firms: an empirical


60

investigation”, International Small Business Journal, Vol. 13, n°4, pp. 11-34.

13. Crouzet F. (1995), “Rôle et géographie des dynasties d’affaires”, Entreprise et Histoire, n°9, pp. 5-11.

14. Derijke G. & Ooghe H. (1983), “A comparison of financial structure between public and private firms in Belgium”, Tijdschrift voor Economie en Management, Vol. 28, n°4, pp. 489-507.

15. Desbrières P. & Dumontier P. (1989), “Dettes ou fonds propres: comment choisir?”, Revue Française de Gestion, n°75, September-October, p.5-14.

16. Donckles R. (1989), “A remettre entreprises familiales: La problématique de la succession”, Fondation Roi Baudouin, Roularta Books SA, Brussels, 152 pages.

17. Donckels R. & Frohlich E. (1991), “Are family businesses really different? European experiences from STRATOS”, Family Business Review, Vol. 4, n°2, pp. 149-160.

18. Donckels R. & Aerts R. (1993), “Plein feux sur les PME, de la théorie à la pratique”, Fondation Roi Baudouin, Roularta Books, Brussels, pp. 162-164.

19. Faccio M. & Lang L. (2002), “The ultimate ownership of Western European Corporations”, Journal of

Financial Economics. 20. Floren R. H. (2002), “Cijfers en feiten van het

familiebedrijf, 10 jaar onderzoek onder familiebedrijven”, BDO Accountants et Adviseurs, 59 pages.

21. Friedman M. & Friedman S.(1994), How to run a

family business, Cincinnati, OH: Betterway Books. 22. Friend I.& Lang L. (1988), “An empirical test of the

impact of managerial self interest on corporate capital structure”, Journal of Finance, 43(2), pp.271-281.

23. Friend I. & Hasbrouk J. (1987), “Determinants of capital structure”, in Andy Chen, Research in Finance, 7.

24. Gallo M & Vilaseca A. (1996), “Finance in family business”, Family Business Review, Vol. 9, n°4, pp. 387-405.

25. Harris M. & Raviv A.(1988),”Corporate control contests and capital structure”, Journal of Financial

Economics, 20, pp.55-86. 26. Heldenbergh A. (1999), “Les motivations et

l’efficacité des opérations stratégiques d’OPA sur le marché belge”, Doctoral dissertation, Université de Mons-Hainaut, 475 pages.

27. Hirigoyen G. (1982), “Le comportement financier des moyennes entreprises industrielles familiales”, Banques, n°417, pp. 588-593.

28. Hirigoyen G. (1985), “Les implications de la spécificité des comportements financiers des moyennes entreprises industrielles (M.E.I.) familiales”, Cahier de recherche, IAE Toulouse, n°35, September.

29. Ifera (2003), “Family businesses dominate”, Family

Business Review, Vol. XVI, n°4, pp. 235-239. 30. Jensen M.C. & Meckling W.F. (1976), “Theory of the

firm: managerial behavior, agency costs and ownership structure”, Journal of Financial

Economics, 3, October, pp. 305-360. 31. Jorissen A., Laveren E., Martens R., Reheul A-M.

(2002), “Differences between Family and Nonfamily firms: The impact of different research samples with increasing elimination of demographic sample differences”, Conference Proceedings, RENT XVI,

16th workshop, Nov. 21-22, Barcelona, Spain, Universitat Autònoma de Barcelona, 20 pages.

32. Kalika M. (1988), “Structures d’entreprises : réalités, déterminants, performances”, Economica gestion, 428 pages.

33. Kenyon-Rouviniez D. & Ward J.L. (2004), “Les entreprises familiales”, Que sais-je ? PUF, 127 pages.

34. La Porta R., Lopez-De-Silanes F & Shleifer A. (1999), “Corporate ownership around the world”, The

Journal of Finance, n°2, pp. 471-517. 35. Laveren E. & Bisschops M. (2004), “De

opvolgingsproblematiek in Vlaamse bedrijven: resultaten van een empirisch onderzoek”, Onderzoeksrapport UAMS, July, 30 pages.

36. Markin A. (2004), “Family Ownership and Firm Performance in Canada”, Working Paper, Simon

Fraser University, 27 pages. 37. Merigot J.G. & Hirigoyen G. (1987), “La transmission

des PME familiales: constat et suggestions”, in Futur et Gestion de l’Entreprise, Journées Nationales des IAE. 8. Poitiers, 11, pp.18-20.

38. Morck R. & Yeung B. (2003), “Agency problems in large family business groups”, Entrepreneurship

Theory and Practice, VOL. 27, n°4, pp. 367-382. 39. Ooghe H. & Van Wymeersch C. (1990), “Traité

d’analyse financière”, PUN, 4th edition, volume 1: 423 pages and volume 2: 110 pages.

40. Romano C., Tanewski G. & Smyrnios K. (2000), “Capital structure decision making: a model for family Business”, Journal of Business Venturing, 16, pp. 285-310.

41. Sapusek A.M. (1998), “Fundamental performance of initial public offerings: empirical evidence from Germany”, Congrès international de l’AFFI, Lille, 42 pages.

42. Schulze W.S., Lubatkin M.H., Dino R.N. (2003), “Toward a theory of agency and altruism in family firms”, Journal of Business Venturing, Vol. 18, pp. 473-490.

43. Sonfield & Lussier (2004), “First-, second-, and third-generation family firms: a comparison”, Family

Business Review, Vol. XVII, n°3, pp. 189-202. 44. Thiétart R.-A. (1999), Méthodes de recherche en

management, Dunod, Paris, 535 pages. 45. Van Caillie D. & Denis S. (1996), Etat de la

transmission des petites et moyennes entreprises en province de Liège, Résultats d'une enquête, Faculté d'Economie, de Gestion et de Sciences sociales, Centre de Recherche et de Documentation PME de l'Université de Liège, 8 April, 99 pages.

46. Van Gils A., Voordeckers W., Van Den Heuvel J. (2004), “Environmental uncertainty and strategic behavior in Belgian family firms”, European

Management Journal, Vol. 22, n°5, pp. 588-595. 47. Ward J. (1988), “The special role of strategic planning

for family business”, Family Business Review, 1 (2), 105-117.

48. Westhead P. & Cowling M. (1998), “Family firm research: the need for a methodological rethink”, Entrepreneurship, Theory and Practice, pp. 31-56.

49. Wtterwulghe R., Janssen F., Mertens S. & Olivier F. (1994), “Le financement de croissance des entreprises moyennes belges par le recours au capital à risque”, Louvain-La-Neuve, IAG.


61

AUDITOR CONSERVATISM AND EARNINGS MANAGEMENT:

EVIDENCE FROM TUNISIA

Nadaa Hachicha Elfouzi*, Mohamed Ali Zarai**

Abstract

The aim of this paper is to examine the effect of the type of audit opinions on the earnings management measured by the discretionary accruals in the tunisian capital market context. In particular, we investigate whether abnormal accruals are influenced by modified audit opinions. We find that the probability to manage earnings management to the decrease is related to the issuance modified audit opinion and the presence of No-Big Four auditors. Keywords: auditors, earnings management, Tunisia *ISG-Tunis, 41 Street of the liberty city Bouchoucha, Bardo 2000-Tunis. E-mail/phone : [email protected] (21 34 93 30), **Superior Institute of Management of Tunis. [email protected] (98 22 42 79).

1. Introduction The annual financial statements constitute a synthesis of the activity of firm exploitable by the outside. They serve to the different actors in an optics of assessment, or decision making (shareholders, managers, State and the other public authorities and bankers). Thus, these statements must be established in the respect of principles of regularity and sincerity (Hatfield and al., 2009). However being given that they are considered a product of management, the latitude of which they arrange can let hover a doubt on the sincerity of communicated information, illustrated especially by the notion of the creative accounting (Stolowy, 2000).

The importance to arrange some reliable data on the accounts annuals explains the necessity to resort to an external auditor, who warns the earnings management.

However, since the recent financial scandals, several shapes of earnings management adopted by firms to camouflage their real financial situation of these last. Him in spring that the auditor can be brought to certify the improper accounts, under the effect of pressures of their auditees. These conditions increase the degree of the doubt on the quality of audited financial statements.

Otherwise, These recriminations also dragged a discount in question important of the quality of services provided by the external auditors that are called, in principle, to give of mannered objective, exact and precise the financial information to the profit of investors (Lin and Fraser, 2008).

The Enron business is only a good example, has been accused to have used various methods of earnings management.

In the wake of the collapse of Enron and Andersen, several questions were raised about Enron’s accounting and the behavior of its auditors,

among those the survey of Krishnan and al. (2007) treats if ex-Andersen’s clients received more conservative treatment by their new auditors. They found that auditors were less likely to issue going concern modified audit opinions to small clients who switched from Andersen than to their existing clients. In the same way, Cahan and Zhang (2006) and Lobo and Zhou (2006) report lower discretionary accruals for ex-Andersen clients compared to other clients, suggesting that ex-Andersen clients were treated more conservatively by their new auditors.

Thus, the new laws carrying on the auditor's role increased following the financial scandals and accountants (Sarbanes-Oxley law to the United States, law of the financial safety in France and in Tunisia). They established some more coercive rules therefore on the different aspects of the independence of the external auditor screw to screw of their client. The issue of auditor independence is of serious concern to regulators, investors, creditors and the general public.

In these conditions, the auditor is held to resist pressures of managers, and to express his independent opinion on the financial statements prepared by management which should help to reduce the information asymmetry between the company and its potential investors.

However, in the setting of interest divergences between shareholders and managers (or between creditors and managers), a negative opinion rendered by auditor is often interpreted like a discount in reason of the leader management. To avoid this situation can affect their reputation, managers can exercise pressures on their external auditors, in order to bring them to formulate unqualified opinions. He sacrifices, therefore, the quality of his audit opinion to the profit of his personal interest, in order to preserve his mandate, under light of the theory of the


62

dependence towards resources (Nichols and Price, 1976 and Kato, 1989).

It incites the auditors to prefer of choices conservative accountants. Thus, income-increasing accruals are somewhat more likely to result in reporting conservatism than income-decreasing accruals. Indeed, several recent papers test the hypothesis that earnings management increases the likelihood of receiving a modified audit opinion (Francis and Krishnan, 1999; Bartavo and al., 2000; Johl and al., 2007; Francis and Wang, 2008 and Cameran and al., 2008). These authors use accounting accruals as proxy of the audit quality. In general, these papers posit that modified audit opinions should be a function of accounting accruals

But then, are little the previous studies that use the type of audit opinion as proxy of the audit quality, mainly in anglo-saxon countries (Butler and al., 2004; Chen and al., 2005; Brown and al., 2006; Hunt and Lulseged, 2007 and Charito and al., 2007). The majority of these works corroborates the existence of a negative relation between the earnings management and the modified audit opinions ce at the time of, it seems applicable to fear the possible interaction between the type of audit opinion and the discretionary accruals in an emergent country as Tunisia.

Otherwise, a study led in the tunisian context proves to be crucial, in order to study the real capacity of the external auditor to resist pressures of managers, to express his independent opinion, being given that his nomination decided by the Annual General Meeting, depends managers who solicited him the most often.

Thus, in this study, we seek evidence of the type of audit opinion affect the earnings management measured by the discretionary accruals.

The remainder of the paper is organized as follows. The second section discusses the opinion/accruals relation. In the third section the legal context of auditing in Tunisia is presented. The fourth section details our sampling methodology. The fifth section reports and discusses the results, and the final section contains the conclusions of the study.

2. The effect of the auditor's conservative behavior on the earnings management: theoretical study synthesis The extant literature (Jensen and Meckling, 1976 and Watts and Zimmerman, 1983) suggests that auditing is an important means of mitigating agency conflict between managers and outside shareholders. Auditing is a monitoring device for the shareholders because auditors would report detected material misstatements in audited financial statements. Auditing is also a bonding device of the managers who engage auditors to signal to the shareholders that they will not behave opportunistically. In short, auditing reduces the information asymmetry between

shareholders and managers, and increases the creditability of financial statements.

Following such an agency cost argument, several studies (Francis and Krishnan, 1999; Bartavo and al., 2000 and Johl and al., 2007) argue that modified audit opinions are influenced by more pervasive earnings management. In essence, these studies test the hypothesis that earnings management increases the likelihood of receiving a modified audit opinion. Among those, the study of Francis and Krishnan (1999), it presents evidence of conservative auditor behaviour by examining the relationship between the issuance of modified audit reports and the reported level of accruals for a large sample of United States publicly listed companies. Specifically, Francis and Krishnan (1999) find that auditors of firms reporting high levels of accruals are more likely to issue modified audit reports for asset realisation uncertainties and for going concern problems than auditors of low accrual firms, even after controlling for client-specific and market risk variables. However, they also find that their results only apply for the Big Six group of auditors. They conclude that this latter finding is consistent with the argument that Big Six auditors have greater reputation capital at risk and, therefore, greater incentives for acting conservatively.

As for, Bartov and al. (2000) show that the association between audit opinion and abnormal accruals is negative. Their results appear to be due to severely distressed firms (with going concern opinions), rather than due to firms engaging in extreme earnings management. In the vein of Bartov and al. (2000), Johl and al. (2007) examine auditor reporting behaviour in the presence of aggressive earnings management in the context Malaysia. They find that Big Five auditors in Malaysia appear to issue modified audit reports more frequently than their No-Big Five counterparts when high levels of abnormal accruals are present.

However, Butler and al. (2004) regress discretionary accruals on audit opinion type. They examine whether certain modified audit opinions (scope limitations, departures from Generally Accepted Accounting Principles (GAAP)) are associated with discretionary accruals. They find that the documented relation between modified opinions and abnormal accruals rests with companies that have going concern opinions. These firms have large negative accruals that are likely due to severe financial distress. Overall, they find no evidence to support inferences in previous research that firms receiving modified audit opinions manage earnings more than those receiving clean opinions.

In addition, Butler and al. (2004) show that abnormal accruals are more negative for going concern companies audited by the Big Five than for those audited by the No-Big Five would support the auditor conservatism explanation. If accruals are not different for Big Five versus No-Big Five audited going concern companies, or if accruals are more


63

negative for going concern companies audited by the No-Big Five, then the relation between accruals and troubled firms is less likely due to auditor conservatism.

On the same american market, Lai (2003) shows that after the Act, auditors are more likely to issue modified audit opinion to their clients and their clients are more likely to be associated with lower discretionary accruals. Thus, using the issue of modified audit opinion and the provision of discretionary accruals as measures of auditor independence, their results suggest a positive association of the passing of the Act and improvement of auditor independence. Their results corroborate with those of Ashbaugh and al. (2003), Chung and Kallapur (2003), Cahan and Zhang (2006) and Krishnan and al. (2007).

In a context of financial distress, Hunt and Lulseged (2007) examine the link between client size and two outputs of the audit process, abnormal accruals and the going concern report. Their study supplements Reynolds and Francis (2000) by extending their analysis to the clients of No-Big Five auditors, a relatively small but important and growing segment of the audit market. Otherwise, Reynolds and Francis (2000) find that larger clients of Big Five auditors have lower levels of accruals compared to otherwise similar smaller clients and potentially financial distressed larger clients of Big Five auditors are more likely to receive a going concern opinion. However, their results cannot be generalized to the clients of the No-Big 5 auditors that are excluded from the study because on the one hand, there is a difference in the quality of audits that Big 5 and No-Big 5 auditors provide. So, Big 5 auditors offer higher quality audits [better trained employees and use technology that allows them to better detect errors and irregularities (DeAngelo, 1981 and Craswell and al., 2002), protection of the high reputation (DeAngelo, 1981)]. On the other hand, there are differences in the underlying characteristics of clients of Big 5 and No-Big 5 auditors that might influence the clients’incentives for earnings management and the likelihood of a clients’ bankruptcy (cash flows, size, leverage and financial health) (Craswell and al., 1995 and Krishnan and Schauer, 2000).

Hunt and Lulseged (2007) find that No-Big 5 auditors, like Big 5 auditors do not allow their larger clients greater leeway to manage earnings. In addition, they find that larger clients are at least as likely, if not more likely, to receive a going concern report as are otherwise similar smaller clients.

Still in the same financial distress context, and in the same furrow of ideas, Charito and al. (2007) examine the earnings behaviour of managers during the distressed period by looking at sources of abnormal accruals prior to the bankruptcy-filing year, for 859 United States bankruptcy-filing firms over the period 1986-2004. Their Results show that managers of highly distressed firms shift earnings

downwards prior to the bankruptcy filing. They specified that one of factors that explains this result is qualified audit opinions exert pressure on managers to follow more conservative earnings behaviour during the distressed period.

In sum, the theoretical development whole exposed above permits us to show the role of auditors in the process of improvement of the quality of the financial information distributed to the external users. In particular, auditors are susceptible to attenuate the earnings management. For it, he seems desirable to reinforce their independence and to assure the control of the quality of their work. Arrangements foreseen by the law of the financial safety of October 2005 go in this sense.

Thus, in the setting of the following section we explore this track of research on the basis of a sample of the tunisian firms.

For it, we consider to test the validity of our hypotheses: Hypothesis 1.1. Firms with qualified audit report

will be more susceptible to manage the discretionary

accruals to the decrease that those with unqualified

audit report.

Hypothesis 1.2. Firms with qualified audit report

will be more susceptible to manage the discretionary

accruals to the decrease audited by Big Four audit

firms that those audited by No-Big Four.

3. Framework of auditing in Tunisia In Tunisia, the auditor's mission was organised by the code of commerce published in 1959. In order to improve the quality of the audits several laws were promulgated in 1982, 1988 and 2000. Indeed, the law of 1982 is characterised by a real entrance of the legal audit in the economic environment. This law carried the creation of the Tunisian Institute of Certified Public Accountants. For the 1984-1999 period, this Institute published standards on accounting and auditing. However, these standards presented serious insufficiencies and hiatuses that pushed auditors to use the international standards as referential complementary1. More lately, another law was promulgated in 1988. The objective of this law was to improve the function and work of tunisian’s auditors, who are held to express audit opinion in all independence.

Besides, National Standards on Auditing n°7 and n°15 of the Tunisian Institute of Certified Public Accountants2 and International Standards on Auditing n°700 and n°701 of the International Federation of Accountants3 put in evidence the importance of the opinion paragraph expressed by the statutory auditors, since it sums up findings of the audit mission. They stipulate that the expression used by the auditor to formulate his opinion must be simple and don't vary an exercise to the other. They define the different categories possible of the audit opinions.


64

Otherwise, according to the National Standard on Auditing, the auditor concludes expressly either to the unqualified opinion, either to qualified opinion, and either to disclaimer. An unqualified opinion is one that states the financial statements to which it relates give a true and fair view of the client’s financial affairs.

As for the International Standard on Auditing, the ISA 700 treated the standard audit report where the audit opinion is unqualified. The ISA 701 treated the modified audit report. In this case audit opinions are three categories according to circumstances: qualified opinion, adverse opinion and disclaimer.

In short, the code of commerce, which was issued in 1959, was modified in 2000. It was transformed to the code of commercial companies. This modification entailed several innovations for the accounting profession. These innovations are dealing with: (a) auditor's designation for all commercial companies, (b) improvement of tools used by independent auditors in their mission and (c) date and presentation of auditor’s report.

We develop our methodology of the empiric research adopted, to know sample data, model, variables and their measures, and in short our found results. 4. Research methodology The aim of this paper is to study the relation between modified audit opinions and abnormal accruals, while controlling the effect of other exogenous factors. With this intention, we will detail our sampling methodology.

4.1. Sample data We attempt to collect and classify the 435 audit reports of 137 Tunisian firms from 2002 to 2007 as having 311 unqualified opinions, 123 qualified opinions and only one case of disclaimer.

The accounting variables are determined from the financial statements and data bound to the audit

are determined from the audit reports. These data were obtained from the council of financial market bulletin.

For the 123 qualified opinions, we clear 8 types of qualifications: non conformity with accounting principles, error or irregularity, scope limitations of an audit, absence of provisions, insufficiency of provisions, absence of follow-up rigorous of engagements out balance, failing of the system of the internal control and going concern.

Of the 435 audit opinions, 247 relate to utilities and financial firms that we eliminate because of inherent differences associated with accounting accruals for these firms. We lead to a final sample of 53 no-financial firms and 188 audit reports for the 2002-2007 period. 1 In 2002, the Tunisian Institute of Certified Public Accountants adopted the international standard on auditing, elaborated and published by the international federation of accountants (IFAC) and committed to distribute them close to its members. The IFAC elaborated several recommendations and instructions about ethics, formation of the accountants and audit reports. Among these recommendations, guideline No. 13 “The Auditor’s Report on Financial Statements” was issued in 1983. The motivation for issuing the guideline was to promote the reader’s understanding and help to measure uniformity in the form and content of auditor’s report. After several years of revision works, the publication of the standard ISA 700 “The Auditor’s Report on Financial Statements” (IFAC, 1994) was approved in 1994.

2 National Standard on Auditing n°15 of the Tunisian Institute of Certified Public Accountants “The report of the independent auditor on the financial statements” and National Standard on Auditing n°7 of the Tunisian Institute of Certified Public Accountants “Diligences of the statutory auditors concerning the report on the social accounts”.

3 International Standard on Auditing n° 700 of the International Federation of Accountants “The auditor’s report (independent) on financial statements” and International Standard on Auditing n° 701 of the International Federation of Accountants “Modifications brought to the content of the auditor's report (independent)”.

Table 1. Number of audit opinions

Unqualified opinions Qualified opinions Type of opinions Years Big Four No Big Four No

Totals

2002 5 6 1 12

2003 8 18 2 4 32

2004 6 16 5 9 36

2005 5 18 5 9 37

2006 7 22 5 8 42

2007 5 16 3 5 29

Totals 36

(19.15%)

96

(51.06%)

20

(10.64%)

36

(19.15%)

188

Results of this table show that the majority of audit

reports in Tunisia is unqualified (70. 21%), on the one hand, and are on the other hand audited by No-Big 4 auditors. It can be explained by the fact that the majority of the economic cloth in Tunisia is formed by the small

and averages enterprises, these last generally choose that their financial statements are audited by No-Big 4 auditors.


65

4.2. Empirical model The general model used to determine which factors influence the abnormal accruals is as follows:

titititititiCURRFISIBIGAUOPABAC

,,4,3,2,10,4 εβββββ +++++=

Where ABAC: abnormal accruals. AUOP: one if opinion is qualified and zero otherwise. BIG4: one if a Big 4 audit firm and zero otherwise. FISI: natural logarithm of total assets. CURR: current assets to current liabilities ratio. In order to clear determinants of accounting manipulations, we appraise coefficients of the multivariate regression in data of Panel on the software STATA 10.

4.2.1. Dependent variable: Measuring discretionary

accruals

Abnormal accruals (ABAC) represent that part of total accruals that is more susceptible to manipulation by managers and are frequently used in prior studies as a proxy for earnings management (Jones, 1991 and DeFond and Jiambalvo, 1994).

Abnormal accruals are the difference between total accruals and non-discretionary accruals. Total accruals are the difference between operating income and cash flows from operations. No-discretionary accruals are the expected (or predicted) portion of total accruals in Jones model modified that regresses total accruals on changes in revenue, gross property plant and equipment, and return on assets (Kothari and al., 2005). We define the following model of accruals:

( ) ( ) ( ) ( )itititititititititit

ASETROAASETGPPEASETREVEASETASETACCR εαααα +++∆+=−−−−− 141312111

/1

Where ACCRit: total accruals for sample firm i for year t. ASETit-1: total assets for sample firm i for year t-1. ∆REVEit: change in net revenues for sample firm i for year t. GPPEit: gross property, plant, and equipment for sample firm i for year t. ROAit: return on assets for sample firm i for year t. εit: unexpected portion of total accruals for sample firm i for year t. The residuals εit from this equation are the discretionary accruals.

4.2.2. Independent variables

AUOP: audit opinion is the variable of interest, variable dummy who takes the value one if the firm receives modified audit opinion and zero otherwise (Leventis and al., 2005; Ballesta and Garcia-Meca, 2005; Davidson III and al., 2006 and Al-Thuneibat and al., 2008). However, some authors replace the audit opinion variable with variables corresponding to the audit opinion categories (Soltani, 2002 in France; Butler and al., 2004 to the United States and Pucheta and al., 2004 in Spain).

Besides, a majority of studies interested itself to only one type of audit opinion “going concern” in a context of financial distress (Charito and al., 2007; Hunt and Lulseged, 2007; Ogneva and Subrananyam, 2007; Citron and al., 2008 and Lam and Mensah, 2008).

In our tunisian context, we are not going to proceed to a distinction between the qualified audit opinion categories, being given that the number of modified audit opinions is weak (29.79%) of the final sample.

BIG4: reputation of the auditor (audit firm size) is the variable dummy who takes the value one if a Big 4 audit firm and zero otherwise. Based on prior research, we expect abnormal accruals to be negatively related to the Big Four audit firms. So a negative Big 4 coefficient would suggest that Big 4 auditors don't permit their larger clients to manage earnings (Johl and al., 2007; Hunt and Lulseged, 2007 and Cameran and al., 2008).

Firm size (FISI) and current ratio or liquidity (CURR) are firm characteristics associated with level of accruals. The size of the firm is natural log of auditee’s total assets, which prior research finds is negatively related to accruals. Otherwise, Cameran and al. (2008) suggest larger companies will have more formal and developed internal financial control systems and will therefore be less likely to manage earnings

And in short, as a proxy for liquidity we consider current ratio is measured by current assets scaled by current liabilities (Bartov and al., 2000 and Butler and al., 2004). 5. Empirical results and analysis

5.1. Univariate results

Table 2 presents the descriptive statistics

Table 2. Descriptive statistics for dependent and independent variables

variables Obs Mean Std. dev Min Max

ABAC 167 -1.420 1.900 -2.410 1.690 AUOP 192 .307 .462 BIG4 202 .262 .441 FISI 232 17.310 1.346 12.646 21.269

CURR 214 2.057 2.573 .002 14.718

According to table 2, we can advance that on average the discretionary accruals are negative (-1.420). It means that the average of the earnings

management detected is to the decrease. Besides, the information about abnormal accruals is available for 167 firms-years. Let's note also that a third (26. 20%)


66

observations present Big Four auditors. The standard deviation of the measurement of

audit opinions and audit firms enables us to note that these variables vary slightly inside our sample. Nevertheless, the standard deviation of ratio current makes it possible to notice that this indicator is very volatile compared to the other variables of the study.

A multicollinearity problem is also likely to exist when explanatory variables correlate significantly with each other. Multicollinearity in our data set was investigated by the correlation matrix and Variance Inflator Factor (VIF).

Table 3. The correlation matrix ABAC AUOP BIG4 FISI

CURR

ABAC 1.0000 AUOP -.1306 1.0000 BIG4 .0566 .0810 1.0000 FISI -.2138 .0029 .1853 1.0000 CURR .0327 -.2890 .0611 -.0840 1.0000

The correlation matrix of the variables of the sample is shown in table 3. Most of the coefficients of correlation are small, no correlation exceeds the value of .290. It seems to us that the multicollinearity is not a problem while interpreting results of regression, the big value of correlation is -.289 represents the correlation between AUOP and CURR.

Table 4 shows that the VIF of all the variables is lower than 5, from where there are no problem of multicolinearity. Nevertheless, it proves to be interesting to carry out a multivariate analysis taking of account the simultaneous effect of the various

studied variables. 5.2. Multivariate analysis According to results of our model, the statistical Wald Chi two to 4 degree of liberty is the order of 12.460, either a level of significance is equal to .0143 lower to 5%. We reject the joint hypothesis, we keep so the presence of individual effect in the following of our work.

The evaluation of our model under data of Panel suggests some previous tests, as the test of Hausman (1978). It serves to discriminate the fixed effect/random effect. The statistical of Chi two to 4 degree of liberty is the order of 1.710, either a level of significance of .789 superior to 5%. Thus, we keep the random effect.

In order to arrive to the best results, the question of the heteroskedasticity, the autocorrelation and the normality of residues in the setting of data in Panel is addressed. Indeed, while using the Breusch-Pagan (1980) test, the statistical of Chi two is the order of

.930, either a level of significance of a value of .334 superior to 5%, imply that the model doesn't endure the problem of heteroskedasticity. While using the test of Ramsey, we notice that the model endures the problem of correlation and dependence between terms of mistakes.

And in short the distribution of residues is no normal, according to tests of distribution (Skewness and Kurtosis).

Being given that conditions of OLS are not verified, the least square method generalized (GLS) seems thus most suitable, in the following of our work.

Table 4. VIF

Variables VIF 1/VIF

AUOP 1.190 .843 BIG4 1.060 .946 FISI 1.270 .790

CURR 1.150 .871 Mean VIF =1.170

Table 5. Results of estimation

Variables Coefficients Std. error Z P>│Z│ Expected sign Found sign

AUOP -6.310 3.390 (-1.860)** .063 (-) (-)

BIG4 5.090 3.480 1.460 .144 (-) (+)

FISI -3.360 1.110 (-3.030)* .002 (-) (-)

CURR -28.387 60.912 -.470 .641 (+) (-)

CONST 5.800 1.940 2.990 .003 *Risque de rejet est de 1%, **Risque de rejet est de 10%

Table 5 presents the estimation results for our

model. Consistent with our first hypothesis, the coefficient on modified audit opinion is negative and statistically significant with p-value of .063. Otherwise, the tunisian firms exercise of accounting

manipulations. This phenomenon of the earnings management seems then to us very rife in Tunisia, where the shareholding is concentrated, the level of the investor protection is relatively middle and the little developed stock market, on the one hand. On the


67

other hand, firms with modified audit reports tend to have lower abnormal accruals, that is to say, income decreasing accruals reflect a conservative application of Generally Accepted Accounting Principles (GAAP) by managers. Indeed, the conservative external auditors don’t accept material misstatements of in audited financial. Consistent with Butler and al. (2004), Hunt and Lulseged (2007) and Charito and al. (2007) in a context of financial distress. They find that firms receive going concern opinions have negative accounting accruals.

The coefficient for Audit firm size is positive but statistically insignificant. In of other terms, discretionary accruals are positively related to auditor type, indicating that clients of Big 4 auditors have higher discretionary accruals, contrary to our second hypothesis. This result finds his justification in the tunisian context, being given that nearly the two third of audit firms are No-Big four. These last are less heedful to detect mistakes and irregularities accountants, and therefore, the discretion in earnings management is permitted, due to a concern about losing the significant fees. For the control variables, the coefficient on FISI is negative and statistically significant (p<0.01). Indeed, As in Beasley, Carcello and Hermanson (2000) and Cameran and al. (2008), Larger firms tend to have lower accruals. Current ratio has a negative coefficient and not significantly different from zero, contrary to the predictions of Butler and al. (2004), Uang and al. (2006) and Caramanis and Lennox (2008). They find that accruals are managed to the decrease for firms in financial distress with unqualified opinion and those with going concern opinions. 6. Summary and conclusions The objective of our research was to examine the impact of the audit opinion given out on the earnings management measured by the discretionary accruals, while controlling the effect of certain exogenous factors (auditee size and ratio of liquidity). For tending towards our objective of research, we derived and tested our model in the context of data in Panel, on a sample of 188 audit reports for the 2002-2007 period.

The most significant findings are that the probability to manage earnings management to the decrease is related to the issuance modified audit opinion and the presence of No Big 4 auditors. Otherwise, firms of which audit opinions are qualified manage the abnormal accruals more negative and more meaningful that those with unqualified audit opinions. In addition, we can advanced that the inefficiency of Big Four auditors in the reduction of accounting manipulations done by managers, to our sense can explain himself by the economic business reality, that proves that ties of friendship become knotted between managers and auditors. In the setting of this script, appeared the law of the financial safety in October 2005, the major objectives of this law is to reinforce the independence of auditors screw to screw

of the auditee, and is to protect investors by improving the accuracy and reliability of corporate disclosures and to restore investors’confidence in the integrity of firms’ financial reporting. More explicitly, the auditor must be more conservative when he give out audit opinion on the financial statement prepared by management, in case where the result of seems to them over-valued.

Our study is subject to some limitations. We note that our survey is restricted to the type of audit opinion, we encourage future research to proceed a distinction between the qualified audit opinion categories.

This work may be of interest to market regulators, institutional bodies, investors and large audit firms, because the audit report is often the only mean of communication observable between auditors and all users of financial statements. Additional research could extend the results to other mechanisms that improve the quality of information communicated: audit committees and internal governance mechanisms such as the size of the board of directors.

References 1. Abbott, L. J. Parker, S. and Peters, G. (2006),

“Earnings management, litigation risk, and asymmetric audit fee responses”, Auditing. A Journal of Practice

and Theory, Vol. 25 No. 1, pp. 85-98. 2. Al-Thuneibat, A. and Al-Fayoumi, N. A. (2008), “The

effect of qualified auditors’opinions on share prices: evidence from Jordan”, Managerial Auditing Journal, Vol. 23 No. 1, pp. 84.

3. Ashbaugh, H. Lafond, R. and Mayhew, B. W. (2003), “Do non-audit services compromise auditor independence? Further evidence”, The Accounting

Review, Vol. 78, pp. 611-639. 4. Ballesta, J. P. S. and Garcia-Meca, E. (2005), “Audit

qualifications and corporate governance in Spanish listed firms”, Managerial Auditing Journal, Vol. 20 No. 7, pp. 725-738.

5. Bartov, E. F. Gul, F. A. and Tsui, J. (2000), “Discretionary accruals models and audit qualifications”, Journal of Accounting and Economics, Vol. 30 No. 3, pp. 421-452.

6. Butler, M. Leone, A. J. and Willenborg, M. (2004), “An empirical analysis of auditor reporting and its association with abnormal accruals”, Journal of

Accounting and Economics, Vol. 37, pp. 139-165. 7. Brown, W. D. He, H. and Teitel, K. (2006),

“Conditional and the value relevance of accounting earning: an international study”, European Accounting

Review, Vol. 15 No. 4, pp. 605-626. 8. Cahan, S. F. and Zhang, W. (2006), “After Enron:

auditor conservatism and ex-Andersen clients”, The

Accounting Review, Vol. 81, pp. 49-82. 9. Cameran, M. Prencipe, A. and Trombetta, M. (2008),

“Earnings management, audit tenure and auditor changes: does mandatory auditor rotation improve audit quality?”, Working Paper, University Bocconi, Milan and Instituto of Empresa Business School, Madrid.

10. Craswell, A. Francis, J. and Taylor, S. (1995), “Auditor brand name reputations and industry specialization”, Journal of Accounting and Economics, Vol. 20.


68

11. Craswell, A. Stokes, D. and Laughton, J. (2002), “Auditor independence and fee dependence”, Journal

of Accounting and Economics, Vol. 33, pp. 253-275. 12. Charitou, A. Lambertides, N. and Trigeorgis, L. (2007),

“Earnings behaviour of financially distressed firms: the role of institutional ownership”, A Journal of

Accounting and Business Studies, Vol. 43 No. 3. 13. Chen, S. Su, X. and Wang, Z. (2005), “An analysis of

auditing environment and modified audit opinions in china: underlying reasons and lessons”, International Journal of Auditing, Vol. 9, pp. 165-185.

14. Chung, H. and Kallapur, S. (2003), “Client importance, non-audit services, and abnormal accruals”, The

Accounting Review, Vol. 78, pp. 931-955. 15. Davidson III, W. N. Jiraporn, P. and Dadalt, P. (2006),

“Causes and consequences of audit shopping: an analysis of auditor opinions, earnings management, and auditor changes”, Quarterly Journal of Business and

Economics, Vol. 45, pp. 69-87. 16. DeAngelo, L. E. (1981), “Auditor independence, low

balling, and disclosure regulation”, Journal of

Accounting and Economics, Vol. 3, pp. 113-127. 17. DeFond, M. and Jiambalvo, J. (1994), “Debt covenant

violation and manipulation of accruals”, Journal of

Accounting and Economics, Vol. 17, pp. 145-176. 18. Fama, E. F. and Jensen M. C. (1983), “Agency

problems and residual claims”, Journal of law and

Economics, Vol. 26, pp. 327-350. 19. Francis, J. and Krishnan, J. (1999), “Accounting

accruals and auditor reporting conservatism”, Contemporary Accounting Research, Vol. 16 No. 1.

20. Francis, J. R. and Wang, D. (2008), “The joint effect of investor protection and Big Four on earnings quality arround the world”, Contemporary Accounting

Research, Vol. 25 No. 1, pp. 157-191. 21. Hatfield, R. C. Agoglia, C. P. and Sanchez, M. H.

(2009), “The effect of client characteristics on the negotiation tactics of auditors”, Journal of Accounting

Research, Forthcoming. 22. Heninger, W. G. (2001), “The association between

auditor litigation and abnormal accruals”, The

Accounting Review, Vol. 76 No. 1, pp. 111-126. 23. Hunt, A. K. and Lulseged, A. (2007), “Client

importance and Non-Big 5 auditors’ reporting decisions”, Journal of Accounting and Public Policy,

Vol. 26, pp. 212-248. 24. International Standard on Auditing n° 700 of the

International Federation of Accountants. The auditor’s report (independent) on financial statements.

25. International Standard on Auditing n° 701 of the International Federation of Accountants. Modifications brought to the content of the auditor's report.

26. Jackson, B. Scott. Pitman. and Marshall, K. (2001), “Auditors and earnings management”, CPA Journal,

Vol. 71 No. 7, pp. 39-45. 27. Jensen, M. C. and Meckling, M. H. (1976), “Theory of

the firm: managerial behaviour, agency costs and ownership structure”, Journal of Financial Economics,

Vol. 3 No. 4, pp. 305-360. 28. Johl, S. Jubb, A. and Houghton, K. A. (2007),

“Earnings management and the audit opinion: evidence from Malaysia”, Managerial Auditing Journal, Vol. 22 No. 7, pp. 688-715.

29. Jones, J. (1991), “Earnings management during import relief investigations”, Journal of Accounting Research, Vol. 29, pp. 193-228.

30. Kothari, S. P. Leone, A. and Wasley, C. (2005), “Performance matched discretionary accrual measures”, Journal of Accounting and Economics,

Vol.39, pp. 163-197. 31. Krishnan, J. and Schauer, P. (2000), “The

differentiation of audit quality among auditors: evidence from the not-for-profit sector”, Auditing: A

Journal of Practice and Theory, Vol. 19, pp. 9-23. 32. Krishnan, J. Raghunandan, K. and Yang, J. S. (2007),

“Were former Andersen clients treated more leniently than other clients? Evidence from going-concern modified audit opinions”, Accounting Horizons, Vol. 21 No. 4, pp. 423-436.

33. Kato, T. (1989), “Une interprétation de l’indépendance des auditeurs”, Revue Française de Comptabilité, Octobre, pp. 98-102.

34. Lai, K. W. (2003), “The Sarbanes-Oxley act and auditor independence: preliminary evidence from audit opinion and discretionary accruals”, Working Paper, Hong Kong Polytechnic University- General, pp. 1-29.

35. Law n° 2005-96 of 18 October 2005 of the financial safety (Tunisia).

36. Law n° 2003-706 of 1 August 2005 of the financial safety (France).

37. Leventis, S. Caramanis, C. and Weetman, P. (2005), “Determinants of audit report lag: some evidence from the Athens Stock Exchange”, International Journal of Auditing, Vol. 9 No. 1.

38. Lin, K. Z. and Fraser, I. A. (2008), “Auditors' ability to resist client pressure and culture: perceptions in China and the United Kingdom”, Journal of International

Financial Management and Accounting, Vol. 19 No.2. 39. Lobo, G. J. and Zhou, J. (2006), “Did conservatism in

financial reporting increase after the Sarbanes-Oxley act? Initial evidence”, Accounting Horizons, Vol. 20 No. 1, pp. 57-73.

40. National Standard on Auditing n°15 of the Tunisian Institute of Certified Public Accountants. The report of the independent auditor on the financial statements.

41. National Standard on Auditing n°7 of the Tunisian Institute of Certified Public Accountants. Diligences of the statutory auditors concerning the report on the social accounts.

42. Nichols, D. and Price, K. (1976), “The auditor-firm conflict: an analysis using concepts of exchange theory”, The Accounting Review, pp. 335-346.

43. Pucheta, M. C. Vico, M. A. and Benau, M. A. (2004), “Reactions of the spanish capital market to qualified audit reports”, European Accounting Review, Vol. 13 No. 4, pp. 689-711.

44. Reynolds, J. and Francis, J. (2000), “Does size matter ? The influence of large clients on office-level auditor reporting decisions”, Journal of Accounting and

Economics, Vol. 30, pp. 375-400.

45. Sarbanes-Oxley Act: 2002, Public Law 107-204, 107th

Congress, U.S. Statutes at Large. 46. Soltani, B. (2002), “Timeliness of corporate and audit

reports: some empirical evidence in the french context”, The International Journal of Accounting, Vol. 37, pp. 215-246.

47. Stolowy, H. (2000), “Comptabilité créative”, In Encyclopedie de Comptabilité, Contrôle de Gestion et Audit, Economica, pp. 157-178.

48. Watts, R. and Zimmerman, J. (1983), “Agency problems, auditing, and the theory of the firm: some evidence”, Journal of Law and Economics, Vol. 3, pp. 614-615.


69

EFFECTS OF HUMAN RESOURCE PRACTICES ON FAMILY FIRMS

SOCIAL PERFORMANCE

Olivier Colot*, Claire Dupont**, Mélanie Volral***

Abstract The aim of our research is to analyse social performance (through turnover rate) of large family owned business in relation to their human resource practices. We made multiple regressions on a sample of 60 large firms. Our global model, considering large family owned business and non-family owned business, shows that part-time contracts increase turnover significantly, while training reduces it. We observe the same relation when we analyse family owned business specifically where we also note that to belong to the trade sector influences turnover significantly. When we consider non-family owned business on the other hand, then variables like pay, training, firm’s age and services or building sectors tend to affect turnover significantly. Keywords: family firm, family owned business, human resource management, social performance, turnover *UMH, Warocqué Management faculty, Place du Parc 20, 7000 Mons, Belgium, tel : 003265/37.32.80 (fax : 003265/37.30.54) [email protected] **PhD in Business Administration, UMH, Warocqué Research Center ***Master in Business Administration, UMH, Warocqué Research Center

Introduction

Representing more than half of the gross domestic product of all the market economies (KENYON-ROUVINIEZ and WARD 2004) and assuring in Western Europe between 45 and 65% of gross national product and jobs (ALLOUCHE and AMANN 2000), family firms are a very important actor in most countries economy. Because it is the most common type of firm in the private sector (MORCK and YEUNG 2003), we can understand why academic literature is interested in it.

We have to admit that family firms have specific characteristics which arouse interest. While wanting to perpetuate their activities from generation to generation, family firms also seem to show specific agility and flexibility in comparison with non-family firms. Family character also lets assume a stronger emotional closeness between staff members and their leader. Such companies are said to be more preferred by consumers, to give more opportunities to women, to develop better social policies and to respect much more traditions than non-family firms (ALLOUCHE and AMANN 2000).

Knowing these numerous facets, family firms literature have experienced, according to ARREGLE et al. (2004), a spectacular development with interest in succession, performance and governance issues.

In this paper we want to focus on human resource practices (HR practices) used in family

firms. We want to determine which impact such practices can have on family firms social performance in comparison with social performance of non-family firms.

Although researchers do not seem to take an interest in such human issues (CARLSON et al. 2006, CHUA et al. 2003) we think that we have to give more attention to it. Indeed we believe that family heritage holding and development across generations depends on family firms ability to involve and retain their staff. Maintaining skilled staff is the second priority of family firms according to PRICEWATERHOUSECOOPERS (2007) study.

The specific family mind can undoubtedly influence staff attachment for its firm. But owing to the temporal skyline in which they act, we think that family firms tend perhaps to invest more in their staff in order to develop a real policy of involvement and retention of their human resources. We have to note that while there is some research studying HR practices in family firms (CARLSON et al. 2006, DE KOK et al. 2006, ALLOUCHE and AMANN 1995, …), few studies specifically seem to have analyzed the impact of such practices on family firms performance (HARRIS et al. 2004).

The major goal of our paper is thus to analyze and to explain the social performance of family firms by determining which HR practices could influence it, and by comparing these practices impact with non-family firms.


70

1. Social performance approach

If there are a lot of studies that examine the relations between specific HR practices or a set of HR practices on performance taken all round (BOSELIE et al. 2005, GUEST et al. 2003, HILTROP 1996, HUSELID 1995, …), few of these studies analyze specifically the firm’s social performance. We note that when performance doesn’t principally focus on financial measures (GUEST et al. 2003, ROGERS and WRIGHT 1998,…), it is analyzed more generally. In that case there are various levels in performance analysis that can be combined and that can influence each other: a social level through turnover or absenteeism analysis; an organizational level through productivity or innovation analysis; an economic level through profitability, financial performance or sales growth analysis…

In this paper we want to focus on social performance. What HOUT and IMBS (2004) tend to say about that concept seems to be very interesting knowing that we want to study HR practices in the family firms context. For these authors social dimension that is represented by staff in organization can be valued by the state of mind (people willingness to be involved in their work. This concept also refers to commitment or adhesion notions), by the cohesion (people ability to work together and to appreciate each other), by conflict level,… .All these various concepts appear more precisely in the functioning description of family firms. Such firms have to rely on a committed and motivated staff in which there are few strained relationships if they want to perpetuate their activities. MARTORY and CROZET (1988) also see the firm social effectiveness as its capacity to increase the satisfaction level of its staff and to improve social climate and work conditions.

In line with these authors, we want to link to social performance all the impacts that can have firm social policy on staff behaviors. If social performance represents the firm actions that directly influence its staff in order to stimulate it, to increase its involvement level and/or its adhesion level to organization, it can then be illustrated by the following equation (BOSELIE et al. 2005, p.72): Social performance = f (staff ability, motivation, opportunity to participate).

It seems interesting to define precisely the measures that could help to better analyze the social performance concept in family firm. According to FABI et al. (2004), ROGERS and WRIGHT (1998) or LIOUVILLE and BAYAD (1995), social performance can generally be measured by indicators such as turnover, absenteeism or staff satisfaction that directly refer to HR practices.

Due to the data we can have access to make the empirical part of this paper, we will estimate social performance through the turnover that seems to be a measure reflecting staff state of mind towards their firm. This indicator can also show the firm willingness to stabilize its staff. We think that a staff

that is involved, motivated and satisfied with its job conditions should not intend to leave the firm in which it works. We can also think that the firm that wants to maintain its staff should introduce some practices leading to decrease staff departures. Such characteristics should plead for a low turnover rate.

Now we have defined the way we will analyze social performance in family firms, let’s see what literature tends to say about HR practices that can impact such performance and, more precisely, that can lead to a low turnover rate.

2.HR practices and turnover

Several authors (ARTHUR 1994, HUSELID 1995, BECKER and GERHART 1996, HILTROP 1996,1999, etc.) have tried to analyze the relations between firms performance and the introduction of specific HR practices called « high performance » or « high commitment practices ». We are interested in such practices because they tend to reinforce staff knowledge. Moreover such practices can improve the way staff uses its knowledge thanks to specific incitements. Such practices could then directly affect firms social performance by emphasizing the role that each member staff can play in organization and by strengthening its motivation and adherence to the firm.

In that way through the introduction of « High

Performance Work Practices » (including in particular specific recruitment and selection process, incentive compensation systems, training programs and staff involvement practices), HUSELID (1995) tends to conclude that such practices influence negatively and in a significant way turnover but affect positively productivity and financial performance.

PFEFFER (1998) notes that the most effective firms are marked by flexible work systems, by high compensation systems related to organizational performance, by decentralization of decision-making, by training policies, by promotion and career plans, and the like. Effective firms would also tend to reduce status distinctions and to act on a longer-term perspective.

Analyzing which HR practices can contribute to retain talents in firms, HILTROP (1999) notes that practices such as trainings, teamworking, decentralization of decision-making, etc., are among the more efficacy HR practices. On the other hand job security doesn’t appear in the “best” practices list. However BATT et al. (2002) think that turnover should be lower among employees that see their job as secure. More recently GUEST et al. (2003) as well as BATT (2002) note that firms using more frequently trainings, employment security or higher compensation levels tend to reduce their turnover.

Similar studies have been developed in SMO context. For example LIOUVILLE and BAYAD (1995) note that firms that develop a proactive vision of HRM and that want to favour their staff will have


71

more probabilities to attain higher organizational performances. On the opposite, firms that do not count on their social dimension would experience more social problems that could affect their economic outcomes. FABI et al. (2004) have also observed that practices such as strategic or economic information disclosures, financial participation, job descriptions or participation to decision-making tend to characterize the most performing firms concerning staff retention.

We have to highlight that there is not always unanimity between these various authors to what regards HR practices that could be considered as « high commitment practices ». For example ARTHUR (1994) and HUSELID (1995) do not have the same opinion concerning the idea to integrate variable pay or internal promotion systems in high commitment practices. Despite these divergent opinions all the studies that we have consulted help us distinguish HR practices that can be used in order to stimulate or to mobilize staff in organizations.

Now that we know which HR practices could affect firms social performance, and especially turnover, let’s see if human resource management presents some specificities in the context of family firms.

3.HR practices in family firms

Due to specific values that family firms can develop to create a real family feeling (HARRIS et al. (2004) speak about a strong culture of trust, loyalty and inclusion), we can question about the eventual specificity of HRM policy and HR practices of family firms in comparison with non family firms. According to FREDY-PLANCHOT (2002), the aim of family firms is develop staff loyalty and to maintain it in time. In this way, family firms tend to protect themselves from outside: with time employees that are not in the family circle will be seen as members of the family if they are loyal and devoted (in CABY et HIRIGOYEN 2002, p.196). We could think that HRM policy in family firms should be more paternalistic and more protective. Such policy should also be marked by the willingness to treat staff fairly and to ensure of the trustworthiness and loyalty of staff in the long term. Such a social consciousness could lead family members to make more sacrifices and losses with the intention of saving their company. The commitment and loyalty of people towards the family firm could favor the development of a particular atmosphere that reinforces the membership feeling and highlights the common goals of all the staff (FLAMENT 2006, p.31). According to FLAMENT (2006, p.27), the family firms willingness to maintain jobs can lead to staff happiness. Because they listen more to their employees such firms could favour social advance.

Concerning specific HR practices CASCIO (1995) notes a more flexible approach of HRM in family firms. ASTRACHAN and KOLENKO (1994)

observe some particular HR practices in family firms such as the frequent and regular use of formal procedures of contact with employees, specific pay policies or written job descriptions.

In France ALLOUCHE and AMANN study (1995) demonstrate that, in comparison with non- family firms, family firms have older staff and maintain managerial staff and engineers on a longer period (the authors speak of a « tacit membership contract » between staff and family firms). Family firms CEOs have on average a monthly wage that is lower than the wage of non family firms CEOs. We also have to note that a family firm gives higher rewards outside the wage and uses less part-time jobs because it can deteriorate staff trustworthiness. It seems that family firms use more temporary staff in order to face productivity excess and in order not to disturb permanent staff. We also know that family firms tend to give a lot of attention to staff training and devote a more important part of their payroll to abilities valorization.

With regard to HR practices, CARLSON et al. (2006) analyzed 168 family SMOs and noted that performing family firms give more attention to training, performance appraisals, recruitment, competitive compensation levels, etc., in comparison with non-performing family firms. The study of PRICEWATERHOUSECOOPERS (2007) in 1.454 family SMOs through 28 countries shows moreover that family firms give priority to staff training and that an annual bonus is generally used as a means to reward managers.

Besides these studies that highlight the specificity of HR practices in family firms, an other trend of literature tend to have doubts about such specificity in family firms. According to HAYTON (2006), family firms tend to invest less in training, to train less staff members and to use HR practices that are less complex in comparison with non family firms HR practices. DE KOK et al. (2006) tend to evoke REID and ADAMS (2001) by highlighting the absence of professional HR practices such as appraisal systems or merit-based pay in family firms. That could be explained by the social interactions between family members that could make use of informal procedures easier. The risk to lose flexibility in the relations with the staff by using more professional HR practices could be another argument to account for the lack of such professional practices in family firms (DE KOK 2006). HARRIS et al. (2004) also note that family firms resort to less practices leading to staff involvement. For example, these authors observe a lack of communication and consultation techniques (absence of unions, staff don’t receive information about their company financial situation, no regular meetings with the staff,…) These authors note for example a lack of communication and consultation techniques (absence of unions, staff has no information related to their company financial situation, no regular meetings with staff,…). For HARRIS et al. (2004), family firms culture can exert


72

a strong influence because such culture could be based on the fact that such firms don’t need such specific involvement practices. We have to highlight that these trends arguing for non- specific HRM in family firms derive from studies realized in SMOs (20 and 100 people). The size of such organizations is perhaps too restricted to lead to the development of professional HR practices.

4.Problem statement

The aim of our research is to establish if specific HR practices influence social performance of large family firms (more than 100 workers). We will also compare social performance of Belgian family firms to the social performance of their non-family counterparts. In order to do that, we consider that turnover can constitute a good measure of social performance because it can give an idea of social climate and staff movements within the firm. We would therefore like to analyze if specific HR practices of family firms can influence turnover or if the family character can influence the social performance of the firm.

If some authors explained the lack of professional HR practices in family firms by a specific culture and a paternalistic kind of management, we think that in large firms, the paternalistic aspect or feeling to belong to the same family is perhaps less present because of the firm size and its more complex functioning. In comparison with family SME’s, we then consider that large family firms tend perhaps to use more particular HR practices or to develop more formalized practices in order to make sure of their staff fidelity and of their commitment on length. Let’s note that these large firms are more likely, in comparison with SME’S, to have a HR department15 that can develop more specific HR practices. We also have to remind that these goals related to HRM tend to be particularly important in family firms because of their willingness to transmit their heritage from generation to generation. A high turnover in these firms could reflect a lack of fidelity of workers in relation to their organization or the opposite (the firm does not show its willingness to support its staff). The working climate among the staying collaborators could then degrade.

5.Methodology

5.1. The sample

As justified above, we have worked on a sample of large Belgian family firms, presenting a statistical representativeness on the large Belgian firms population. Each year these firms have to draw up a social balance sheet which includes information

15 According to MAHE DE BOISLANDELLE (1998), many firms begin to take formally charge of HR function when they have between 150 and 250 staff members

relating to social performance, while it is facultative for SME’S. In order to do that, we used all the firms in the obligation to publish their Financial Report and Financial Statements in complete schema to the National Bank of Belgium. Financial and public firms were taken away, which give 477 large firms. To avoid reciprocal holdings, all large firms which are detained in more than 20% by other firms present in the sample were taken away (24 firms). Then, we kept only firms with concentrated shareholding (more than 50% of shareholding), what represents 402 firms. Finally, firms with concentrated foreign shareholding (282) were taken away, in order to keep only large Belgian firms (with a concentrated Belgian shareholding), which represents 120 firms.

So as to keep a sample presenting a statistical representativeness of the large Belgian firms population, we used a random sampling stratified according to family character and to activity sector. Indeed, according to VICINDO DATA MARKETING16 (2002), 52% of the 100.000 larger Belgian firms are family firms. We have then taken the same proportion for the sample. We kept an equal distribution of activity sector between family firms and non-family firms. To this end, we regrouped the different activity sectors in four meta-sectors: services, trade, industries and construction. We checked family character on the basis of criteria defined and mentioned below. We gathered information through various interviews with CEOs. Information was then checked on Internet websites and in the annual financial reports of these firms.

On the basis of this criteria (Belgian firms, activity sector and information availability), our sample includes 74 large Belgians firms, among which half (37) are family owned business.

5.2. Criteria for familial character

We considered that the firm is a family firm when the family holds at least 50% of capital.

This family firm definition is rather limited in comparison with the most recent definitions of family firm (COLOT 2007, ANDERSON and REEB 2003, FLOREN 2002), who use other criteria such as a decisive influence on firm strategy and succession, or the majority of the board of directors made up of family members. But this definition of family firm presents the advantage to use a clear and a measurable criterion, in opposition to qualitative definitions which are more subjective and arbitrary. Consequently, we will use the term « family owned business (FOB) » rather than « family firm » (MILLER and LE-BRETON MILLER, 2008).

Our sample includes finally 74 big Belgians firms, among which half (37) is family owned business.

16 In Institut des Experts Comptables, www.iec-iab.be, « Zoom sur les entreprises familiales », 19/05/05.


73

5.3. Model

The model to be tested is the following: Turnovert = β0 + β1 Compensationst-1 + β2

Flexibilityt-1 + β3 Wellnesst-1 + β4 Job securityt-1 + β5 Trainingt-1 + β6 FAM + xj β7 + νj

where Tut is the turnover of the firm at time t,

Compensationst-1 are either the salaries and fringe benefits (k€) per worker17 or the Bonus (k€) per worker, at time t-1, Flexibility is the ratio between the average number of part-time staff and the sum of the average number of part-time staff and the average number of full-time staff, Wellness is the number of worked hours by workers, Job security is the ratio between the permanent contracts and the sum of the fixed-term contracts and the permanent contracts, Training is either the training rate or the training costs per worker, FAM is a dummy variable equal to one if the firm is a family owned business and zero otherwise, xj includes firm characteristics (the age, the total asset at time t-1, the ROA (Return On Assets) at time t-1, the value added per worker at time t-1 and the sectoral affiliation (3 dummies)) and νj is the error term.

In this model, the year t is 2006, that is to say the last available year of Financial Statements of the firms. So we estimate the impact of HR practices of 2005 on the turnover of 2006, because we think that these practices have an impact but with a certain delay, given the necessary time for their implementation and for their potential effects on staff and its feeling of involvement in the firm.

We calculated the explained variable, that is to say turnover, in the following way: Turnover= (exit permanent contract in full-time equivalent + fixed-term contract in full-time equivalent)/ full-time equivalent 18.

The following table shows different ratios used to illustrate HR practices used in our study. To select these practices, we used studies analysing links between HR practices and firms performance (HUSELID 1995, PFEFFER 1998, HILTROP 1999, GUEST et al. 2003, etc.). The choice of these practices also depended on data that were communicated by firms. The control variables that we used are also introduced there.

5.4. Descriptive statistics Data concerning the social balance sheet, as well as financial data, were extracted from database Belfirst19. Data missing or wrong sums in database lead us to

17 Workers are always expressed in full-time equivalent. 18 PC = permanent contract; FTE = full-time equivalent; FTC = fixed-term contract 19 Financial data of firms that have to publish their Financial Report and Financial Statements to the National Bank of Belgium.

withdraw 14 firms, which brings back the sample to a size of 60 firms. The table 2 presents average and distance type for our variables (the number of workers is in full-time equivalent).

Our sample is finally constituted of 29 non-family owned business and 31 family owned business. The observation of the statistics of the sample shows that:

- staff turnover is lower in family owned businesses. It would confirm the results of FREDY-PLANCHOT (2002) that shows a lower staff turnover in these firms ;

- family owned businesses are smaller than their non-family counterparts: there are on average 658 workers in non-family owned businesses while only 371 workers in family owned businesses ;

- salaries and fringe benefits are on average higher, as well as the percentage of part-time, the number of worked hours, the part of trained workers and the training costs by trained workers. On the contrary, the percent of permanent contract in family owned businesses is higher than in non-family owned businesses ;

- the average age is almost identical in both types of firm. So we do not have to take into account the influence of firms life cycle ;

- The ROA of family owned businesses is higher than the ROA of non-family owned businesses.

Let’s note that our findings relating to the human resource management in family owned are opposed to the results of ALLOUCHE and AMANN (1995), specially for training and for bonuses distributed to the staff. They tend rather to confirm HAYTON (2006) results according to which family owned businesses invest much less in training.

6.Results

6.1. Global model: HR practices and turnover Table 3 reports our estimates, obtained from OLS regression, of the effect of human resource practices on social performance of family and non-family owned businesses taken together. Some variables have been dropped due to a problem of multicollinearity.


74

Table 1. Indicators reflecting HR practices and control variables

HR Practices Indicators used Information Compensations Salaries and fringe benefits per worker

= salaries and fringe benefits in full-time equivalent (k€) / full-time equivalent Bonus per worker = extra compensations (k€) / full-time equivalent

Salaries and fringe benefits are compensations, welfare costs and pensions. Extra compensations represent the non-taxable social benefits package (wedding presents, sports centres, advantages resulting from the access to a medical service, …)

Flexibility Part-time rate = average number of part-time staff / (average number of part-time staff + average number of full-time staff)

The goal is to determine the intensity of part-time use by firms.

Wellness Number of worked hours by worker = Number of worked hours by workers in full-time equivalent / full-time equivalent

Number of really worked hours and paid during a year (we do not take into account overtime that is not paid, and sick leaves)

Job security Part of permanent contracts (PC) = Full-time equivalent with permanent contracts / (full-time equivalent with fixed-term contracts + full-time equivalent with permanent contracts)

The goal is to determine the intensity of permanent contract used by firms.

Training Training rate = Number of trained workers / FTE Training cost per trained worker = Training cost (in thousand euros)/Number of trained workers

Quantitative approach of training that can give an idea of training intensity. Qualitative approach of training that can give an idea of training quality.

FAM = a dummy variable equal to one if the firm is a family owned business and zero otherwise Control variables 20 : - Age = Age of the firm in 2006 - Services = a dummy variable equal to one if the firm operates in the service sector and zero otherwise, with the industrial sector as reference - Trade = a dummy variable equal to one if the firm is a trade firm and zero otherwise, with the industrial sector as reference ; - Construction = a dummy variable equal to one if the firm operates in the construction sector and zero otherwise, with the industrial sector as reference ; - Asset = total of the assets of the firm (in thousands €) - ROA = measure of the rentability of the total assets (financial results of the firm) - VA/L = added value by worker in thousands € (measure of staff productivity)

Table 2. Means and Standard Deviations (SD) of the Main Variables

Variables Mean SD Mean SD

Non-family owned business Family owned business

Staff turnover 0.27 0.24 0.22 0.20

Number of workers 658.24 1703.51 370.94 298.63

salaries and fringe benefits (k€)/ Number of workers 49.39 17.53 42.87 10.72

Bonus (k€)/ Number of workers 0.23 0.33 0.22 0.39

Percent of part-time contracts 0.17 0.25 0.14 0.16

Number of worked hours / Number of workers 1607.28 202.46 1515.13 112.53

Percent of permanent contracts 0.91 0.23 0.92 0.22

Part of trained workers 0.37 0.35 0.30 0.33

Training costs / trained workers (k€) 1 1.16 0.72 0.68

Firm age 26.07 13.91 25.65 18.13

Total assets (k€) 94872.28 154912.6 45261 35565.52

ROA 5.76 7.50 7.34 7.52

Value added by worker (k€) 80 71.84 61.68 25.45

Sectoral affiliation

Services 5 5

Trade 7 4

Construction 3 4

Industries 14 18

Total number of firms 29 31

20 It would have been interesting to include other control variables such as the presence of trade unions in firms, the presence of a HR department, etc., but such information is not available in DVD-Rom Belfist from which we extracted data.


75

Table 3. HR practices and turnover

Dependent variable Turnover

Whole sample Non-family

owned business Family owned business

Intercept 0.02

(0.27) 0.55°° (0.27)

-0.65 (0.57)

Flexibility 0.44** (0.13)

0.16 (0.14)

0.93** (0.29)

Salaries and fringe benefits per worker -0.002 (0.002)

-0.005* (0.002)

0.003 (0.005)

Bonus per worker -0.08 (0.07)

-0.16°° (0.08)

0.04 (0.14)

Wellness 0.0002

(0.0002) 0.00004 (0.0002)

0.0005 (0.0004)

Training rate -0.12° (0.08)

-0.20* (0.08)

-0.34°° (0.19)

Control variables

FAM -0.06 (0.05)

Age 0.0003 (0.001)

-0.005* (0.002)

-0.001 (0.002)

Services 0.08

(0.08) 0.36* (0.10)

-0.02 (0.12)

Trade Sector -0.07 0.06

0.02 (0.06)

-0.20° (0.13)

Construction 0.05

(0.08) 0.24* (0.09)

-0.06 (0.11)

ROA 0.004

(0.003) 0.005

(0.004) -0.0007 (0.006)

Total asset 0

(0) 0

(0) 0

(0)

Adjusted R² 0.39 0.71 0.21

F-stat 4.08** 7.09** 1.73°

Notes : **/*/°°/° indicate significance at the 1%, 5%, 10% and 15% level, respectively. Standard errors are reported between brackets.

Findings emphasize the existence of a positive and significant relationship between the use of part-time jobs, which can refer to a specific kind of flexibility, and turnover and a negative and significant relationship between the training rate and turnover.

With regard to flexibility, we observe a positive coefficient that links part-time jobs use and turnover. That means that a 1% increase of part-time contracts tends to increase turnover of the following year of 0.44%. We can link such results with ALLOUCHE and AMANN (1995) observations. These authors think that part-time use can be a risk for staff trustworthiness. Even if their study analyzes only family firms and even if our findings concern family and non family firms, we can think that such part-time contracts don’t match staff waitings. Indeed if a firm proposes to its staff member such contract, there is a risk the worker leaves its company to go to another firm that would propose a permanent contract. Such reasoning is all the more likely if it concerns a young staff member who seeks a paid full-time job that helps him to take charge of his personal expenditures (house, car, etc.). By imposing part-time contracts, the firm risks of altering its ability to retain staff. In this case staff loyalty is then proportional to work duration proposed by the company. Let’s however note that some staff members can want such contracts and that could lead to the opposite relation.

We also observe a significantly negative relation between training rate and turnover. An increase of training rate of 1% leads to a decrease of 0.12% of turnover the year following the training. Such findings agree with the results of ARTHUR (1994), HUSELID (1995), HILTROP (1999) or GUEST et al. (2003). By investing more and more in training, we think that the company considers its staff with a longer term vision in which staff members have a crucial role in firm functioning. Training contributes to develop staff intellect because it can renew or extend its knowledge. Such practice can be seen as a kind of reward for staff because they can develop their competences, knowledge or attitudes in order to become more performing in their work and to evolve in the company with better paid jobs or jobs with more responsibilities. In this way training leads staff to rely more on its company and can contribute to reinforce staff involvement in the firm.

We also note a negative – but non significant - relation between the family character and turnover. Such observation tends to confirm the idea of “tacit membership contract” (ALLOUCHE and AMANN (1995)) between staff and its family firm that wants to maintain its workforce for a long time (FREDY-PLANCHOT 2002).

Now we have a general idea of HR practices that affect turnover of family and non family businesses, we will subsequently compare the HR practices that


76

can influence turnover of family / non-family businesses.

6.2. RH practices and turnover in family owned businesses and non-family owned businesses Results obtained from OLS regression on non-family owned businesses are presented in the third column of Table 3.

Among family owned businesses it seems that several HR practices are significantly related to turnover. When we look at compensation we note that salaries and fringe benefits and bonuses per worker tend to impact significantly and negatively turnover. Such findings agree with HUSELID (1995), BATT (2002) or GUEST et al. (2003) studies. We can easily understand that a company will be more attractive if it offers its staff higher compensation levels or more social advantages (benefit packages). Monetary incitements could then be a means to retain people in companies. According to RANDSTAD (2007) research high compensation levels seem to be the first factor that impacts the decision to work in a company. When we know that half of the Belgian people need to work for financial reasons we can understand that staff members will need to stay in companies which offer them attractive pays.

Our findings also show a negative and significant relation between training rate and turnover. As stated earlier training can be seen as a means to improve staff skills and to manage them in a longer perspective because company needs their competences and knowledge to flourish. Through training firms can then highlight the role of each staff members in its functioning and show that all employees are crucial in its development.

Firms age is another significant variable. We note that the older a company is, the lower its turnover tends to be. LACOURSIERE et al. (2002) find the same results and such a relation seems to be logic: staff is perhaps less frightened of working in a company that is older and has more experience and fame. On the contrary we can think that people will be less inclined to work in a younger firm that has still to assert itself (new markets or products development, search for new customers, etc.). Moreover, an older company could have introduced specific involvement practices that could have been reconsidered according to staff feelings about them. On the opposite a younger company only begins to formalize its HR practices (especially if its HR department is recent) and risks having not enough resources (financial or human) to introduce such involvement systems.

When we look at activity sectors, it seems that services and construction sectors are positively and significantly related to turnover. Working conditions are generally harder in such sectors (people have to work outside, their tasks are sometimes very arduous, non-standard work schedules in services sector and so on) but lead to lower wages in comparison with other

sectors. According to a study that RANDSTAD (2007) administered to 10.000 people, such sectors are not listed among the most attractive sectors. 6.3. HR practices and turnover in family owned businesses

While several HR practices are significantly related to non-family owned business turnover, only 2 HR practices are significantly linked to family owned business turnover: the use of part-time contracts that refer to flexibility and the training rate, as the fourth column of Table 3 emphasizes.

In flexibility terms we note a positive relation

between part-time jobs and turnover. Indeed while part-time contracts use is 1% up, family owned businesses turnover tends to be 0.93% up the following year. If family owned businesses tend to consider each of their member staff as a family member (FREDY-PLANCHOT 2002), we can then believe that firms offering part-time jobs don’t contribute to get their staff more implied in their functioning. It is all the more probable for staff member that does not need to work with such contract. As stated by ALLOUCHE and AMANN (1995), such contracts risk to alter staff commitment feeling towards their company because it partially integrates people to company life.

Training rate is also significantly and negatively related to family owned business turnover. While there is a 1 percent training rate increase, there is a 0.34 percent turnover decrease. We can explain such a relation in the same way as for global but also for non-family owned business findings. We have to emphasize that while family owned businesses tend to decrease their turnover with more intensive trainings, such companies tend to train less than non-family owned businesses as our descriptive statistics suggest.

When we look at the activity sector impact, we note that companies in trade sector tend to have a lower turnover than in other sectors. This result is somewhat bit surprising. We can possibly assume that trade sector offers more incitement policies (regular trainings, performance based pay, etc.) and faster promotion opportunities in comparison with other sectors. Such practices could then contribute to retain staff in company. 7.Conclusions and development tracks

In this paper we have wanted to analyze if large Belgian family owned businesses developed more than large Belgian non-family owned business their human resource management in order to retain staff. We have attempted to see if specific HR practices could decrease turnover by assuming that a low turnover proves the firm ability to retain its staff. According to the hypothesis of staff retention we can think that family owned business invest more in their human resource management. However, our sample


77

sometimes proved the opposite. Indeed if turnover is lower in a family context, we noted that family owned business have lower labour costs, give less bonuses to their staff and train less than their non-family counterparts. We could possibly link these trends to the smaller size of family owned businesses in comparison with non-family owned businesses of our sample. However, family owned businesses use more permanent contracts and less part-time workers.

We then changed the specification of the model to analyse the impact of HR practices on turnover of the year following their introduction because we thought that these practices did not influence immediately turnover. Firstly, the regression on a sample of family and non-family owned businesses reveal that only 2 HR practices significantly affect turnover. The use of part-time jobs increases turnover while training rate decreases it. We have to note that the binary variable that represents the family character is negatively but not significantly related to turnover. Such character could then contribute to lower turnover in firms. Secondly, we have tested our model on specific samples (family owned businesses / non-family owned businesses). In a non-family context several HR practices and control variables are significant. In this, we note a negative and significant relation between wages (including labor costs and fringe benefits), trainings, firm age and turnover. On the contrary there is a positive relation between turnover and services or construction companies. However, in a family context, the regression only reveals a significant and positive relation between turnover and part-time jobs and a negative relation between training rate and turnover. Let’s note that trade sector affects significantly and negatively turnover.

We note that more HR practices are significantly related to turnover when they are analyzed in a non family context. Such results do not mean that family owned businesses do not care for their staff. Indeed, because of the methodology we used, we cannot measure the incentive character of HR practices in these firms. We can possibly be connected with authors like HAYTON (2006), DE KOK et al. (2006) and HARRIS et al. (2004) that dealt with the possible lack of HR professional practices in family owned businesses. Specific culture or the use of informal involvement practices in family context could explain why such firms have a lower turnover. But the smaller size of the family owned businesses of our sample, compared with non family businesses, could also account for the reasons why lower wages and less intense trainings are observed in a family context. In an analysis more focused on SME’s, we should pay more attention to specific factors related to the internal context of these companies (CEO role, staff social awareness, social interactions, etc.). Such factors could perhaps offset the possible lack of specific HR practices (monetary advantages, trainings,…) that such small firms cannot introduce in comparison to larger companies.

Some improvements have to be brought to this study. Future research should rely on panel data in order to control for the non-observed characteristics of firms. In order to strengthen our findings we could also extend our sample to more companies. We also think that it would be interesting to collect more qualitative data with interviews or questionnaires. For example information about the intenseness of HR practices used in firms, about staff feelings with regard to these practices, about the characteristics of family firms functioning (culture, conflicts management, etc.) should help us to better define the model to use in order to estimate turnover rate. Moreover information related to activity sector of our firms sample would be useful because we have to be cautious when we analyze the links between lower turnover and staff satisfaction. Indeed, by being distinguished through specific characteristics (kind of contracts, sector in decline with limited appointments, low wages, etc.), some activity sectors can dissuade people with specific skills from leaving their jobs even if these people are dissatisfied with their job. In the same way, we think we could improve the indicator used to measure turnover. By focusing more particularly on voluntary exits we could have a clearer idea of social climate in a company. It could also be interesting to analyze performance in firms more globally by combining social, economic and financial approaches. References 1. ALLOUCHE J., AMANN B. (2000), « L’entreprise

familiale : un état de l’art », Finance Contrôle

Stratégie, 3(1), pp.33-79 2. ALLOUCHE J., AMANN B. (1995), « Le retour

triomphant du capitalisme familial », in « de Jacques Cœur à Renault : Gestionnaires et Organisations, », Presses de l’Université des sciences sociales de Toulouse, pp.1-23

3. ANDERSON R. and REEB D. (2003), « Founding family ownership and firm performance: evidence from the S&P 500 », The Journal of Finance, 58(3), pp.1301-1327

4. ARREGLE J.L., DURAND R., VERY P. (2004), « Origines du capital social et avantages concurrentiels des firmes familiales », Management, 7(1), pp.13-36

5. ARTHUR J.B. (1994), « Effects of Human Resource Systems on Manufacturing Performance and Turnover », Academy of Management Journal, 37(3), pp.670-687

6. ASTRACHAN J.H., KOLENKO T.A. (1994), « A Neglected Factor Explaining Family Business Success: Human Resource Practices », Family

Business Review, 7(3), pp.251-262 7. BATT R. (2002), « Managing Customer Services:

Human Resource Practices, Quit Rates and Sales Growth », Academy of Management Journal, 45(3), pp.587-597

8. BATT R., COLVIN A.J.S., KEEFE J. (2002), « Employee Voice, Human Resource Practices and Quit Rates: Evidence from the Telecommunications


78

Industry », Industrial and Labor Relations Review, 55(4), pp.573-594

9. BECKER B., GERHART B. (1996), « The Impact of Human Resource Management on Organizational Performance: Progress and Prospects », Academy of

Management Journal, 39(4), pp.779-801 10. BOSELIE P., DIETZ G., BOON C. (2005),

« Commonalities and contradictions in HRM and performance research », Human Resource

Management Journal, 15(3), pp.67-94 11. BRABET J. (1993), « Repenser la Gestion des

Ressources Humaines ? », Paris, Economica 12. CABY J., HIRIGOYEN G. (2002), « La gestion des

entreprises familiales », Paris, Economica 13. CARLSON D.S., UPTON N., SEAMAN S. (2006),

« The Impact of Human Resource Practices and Compensation Design on Performance: An Analysis of Family-Owned SMEs », Journal of Small Business

Management, 44(4), pp.531-543 14. CASCIO W.F. (1995), « Managing Human

Resources: Productivity, Quality of Work Life, Profits », 4th Edition, Mc Graw-Hill: London, in REID R., MORROW T., KELLY B., ADAMS J., McCARTAN P. (2000), « Human Resource Management Practices in SME’s: A Comparative Analysis of Family and Non-Family Businesses », IBAR, 21(2), pp.157-181

15. CHUA J.H., SHARMA P., CHRISMAN J.J. (2003), « Current Trends and Future Directions in Family Business Studies: Toward a Theory of the Family Firm », Written as part of Coleman Foundation White Paper Series, 62 pages

16. COLOT O. (2007), « La transmission des PME

familiales non cotées : approche de la transmission en

Wallonie et impact sur la performance des

entreprises », Thèse de doctorat, Université de Mons-Hainaut, 468 pages

17. DE KOK J.M.P., UHLANER L.M., THURIK A.R. (2006), « Professional HRM Practices in Family Owned-Managed Enterprises », Journal of Small

Business Management, 44(3), pp.441-460 18. FABI B., RAYMOND L., LACOURSIERE R.,

ARCAND M. (2004), « Les PME les plus performantes se distinguent-elles par leurs pratiques de GRH ? », 7e congrès international francophone en entrepreneuriat et PME, Montpellier (du 27 au 29 octobre), pp.1-16

19. FLAMENT F. (2006), « La stratégie et la gestion des ressources humaines en PME familiale : recension des écrits », note de recherche n°2006-05, Chaire de recherche du Canada sur les enjeux socio-organisationnels de l’économie du savoir, mai.

20. FLOREN R. H. (2002), « Cijfers en feiten van het familiebedrijf, 10 jaar onderzoek onder familiebedrijven », BDO Accountants et Adviseurs.

21. FREDY-PLANCHOT A. (2002), « La gestion des ressources humaines dans les PME à caractère familial », pp.193-204, in CABY J., HIRIGOYEN G. (2002), « La gestion des entreprises familiales », Paris, Economica

22. GUEST D.E., MICHIE J., CONWAY N., SHEEHAN M. (2003), « Human Resource Management and Corporate Performance in the UK », British Journal

of Industrial Relations, 41(2), june, pp.291-314 23. HARRIS R., REID R., McADAM R. (2004),

« Employee Involvement in Family and Non Family-owned Businesses in Great Britain », International

Journal of Entrepreneurial Behaviour and Research, 10(1-2), pp.49-58

24. HAYTON J.C. (2006), « Explaining Competitive Advantage in Family Firms: the effectuation paradox », papier présenté à « l’US Association of Small Business and Entrepreneurship (USASBE) Annual Conference », Tucson AZ, January, pp.1-8

25. HILTROP J.-M. (1999), « The Quest for the Best: Human Resource Practices to Attract and Retain Talent », European Management Journal, 17(4), pp.422-430

26. HILTROP J.-M. (1996), « The Impact of Human Resource Management on Organizational Performance: Theory and Research », European

Management Journal, 14(6), pp.628-637 27. HOUOT T., IMBS P. (2004), « GRH : une source de

performance organisationnelle », Personnel, n°455, décembre, pp.14-17

28. HUSELID M.A. (1995), « The impact of Human Resource Management practices on turnover, productivity and corporate financial performance », Academy of Management Journal, 36(3), pp.635-672

29. KENYON-ROUVINIEZ D., WARD J.L. (2004), « Les entreprises familiales », Que sais-je ?, PUF, 127 pages

30. LACOURSIERE R., FABI B., ST-PIERRE J., ARCAND M. (2002), « Impacts de certaines pratiques de GRH sur la performance organisationnelle et financière de PME », communication présentée au 12e congrès de l’Association Internationale de Psychologie du Travail et des Organisations, juillet, pp.1-7

31. LIOUVILLE J., BAYAD M. (1995), « Stratégies de gestion des ressources humaines et performances dans les PME : résultats d’une recherche exploratoire », Gestion 2000, 1, pp.159-179

32. MAHE DE BOISLANDELLE H. (1998), « Gestion

des Ressources Humaines dans les PME », Economica

33. MARTORY B., CROZET D. (1988), « Gestion des Ressources Humaines », Paris, Nathan

34. MILLER D. and LE-BRETON MILLER I. (2008), « Stewardship vs Stagnation: an empirical comparison of small family and non-family businesses, The journal of Management Studies, Vol. 45, Iss. 1, p. 51.

35. MORCK R., YEUNG B. (2003), « Agency Problems in Large Family Business Groups », Entrepreneurship

Theory and Practice, 27(4), pp.367-382 36. PRICEWATERHOUSECOOPERS (2007), « Making

a Difference – The PricewaterhouseCoopers Family Business Survey 2007/08 », pp.1-60

37. PFEFFER J. (1998), « Seven Practices of successful organisations », California Management Review, 40(2), pp.96-124

38. RANDSTAD (2007), « Randstad Award 2007 – Most attractive employer », pp.1-33

39. REID R.S., ADAMS J.S. (2001), « Human Resource Management: A Survey of Practices within Family and Non-Family Firms », Journal of European

Industrial Training, 35(6), pp.310-320 40. ROGERS E.W., WRIGHT P.M. (1998), « Measuring

Organizational Performance in Strategic Human Resource Management: Problems, Prospects, and Performance Information Markets », Human Resource

Management Review, 8(3), pp.311-331 41. SECUREX (2007), « Saint-Valentin de l’Entreprise

2007 – Bonheur et Motivations au travail », White Paper, 14 février, pp.1-30


79

РАЗДЕЛ 2 СОВЕТ

ДИРЕКТОРОВ

SECTION 2 BOARD OF DIRECTORS

BOARD MONITORING AND FIRM PERFORMANCE: CONTROLLING

FOR ENDOGENEITY AND MULTICOLLINEARITY

Mohammad I Azim*, Dennis W Taylor**

Abstract

Prior corporate governance studies have resulted in inconsistent findings on the significance of relationships between combinations of board monitoring characteristics and firm performance, due to a failure to properly control for endogeneity and multicollinearity problems inherent in the multivariate analysis of their data. In this study, panel data of the top 500 listed companies from the Australian Stock Exchange is used over three years. Results reveal that all but one of the five board characteristics and seven board committee characteristics considered in this study are significantly related to both return on assets and earnings per share in each of the three years. It is concluded that results in this study are much stronger and more consistent than prior governance-performance studies because the structural equation modelling and lagged measures of performance used are able to control for endogeneity and multicollinearity. Key words: Boards monitoring, board committees, firm performance, endogeneity, multicollinearity. * School of Commerce, University of South Australia, City West Campus, GPO Box 2471, Adelaide SA 5001 Phone: + 61 8 8302 7046 Fax. +61 8 8302 0992 Email. [email protected] ** School of Accounting & Law, RMIT University, Australia We acknowledge the suggestions received from the participants of the Accounting and Finance Association of Australia and New Zealand Annual Conference, 2006 and the Global Accounting and Organisational Change Conference 2008.

1. Introduction This study examines whether board monitoring characteristics affect firm performance after controlling the endogeneity and multicollinearity problems in the data analysis. Multicollinearity occurs when independent variables are highly correlated with each other, which makes it difficult to come up with reliable estimates of their individual regression coefficients (Hermalin & Weisbach, 2003). The board monitoring mechanisms should work together in practice, and so are likely to be inherently interrelated to each other. Good corporate governance seeks to

align managers’ to shareholders’ interests through the efficient use of a bundle of board monitoring variables, rather than any single mechanism (Boo & Sharma, 2008; Fernández & Arrondo, 2005; Agrawal & Knoeber, 1996; Rediker & Seth, 1995). If board monitoring variables are adopted with the intention of complementing each other, then they are unlikely to be independent, thus giving rise to the multicollinearity problem (Fernández & Arrondo, 2005). In order to avoid the problem of multicollinearity, multi-variate analysis needs to control for the interrelationships among the monitoring devices of the board and its committees.


80

To achieve this, structural equation modelling is used in this study to model the interrelationships.

Similarly, the issue of endogeneity needs to be dealt with in the analysis. Kole (1996) and Loderer & Martin (1997) present evidence of a reverse-causality or a two-way causality between managers’ equity ownership and firm performance. They determine that at high levels of firm financial performance, management tends to liquidate part of their shareholding, confirming the existence of a relationship whereby managerial ownership derives from firm performance. However, evidence has not been provided in prior studies about the possibility that various board governance variables may not have a relationship with firm performance that is mono-directional running from governance to performance. MacAvoy & Millstein (1999) contend that the failure of previous studies to find significant relationships between board monitoring and performance is because current year performance, rather than lagged performance, has been used in prior studies. Intuitively, the progressive emergence in firms of their financial results during a current year may trigger steps to re-structure the board and its committees or to have them meet more frequently during that same year. That is, board monitoring characteristics may be changed during a current year in response to anticipated end-year financial results, or end-year financial results may be stronger due to improved board monitoring during the year. Therefore, to address this endogeneity issue, this study considers lagged years’ performance.

There are three specific motivations for undertaking this study. First, recent studies (e.g., Boo & Sharma, 2008) have highlighted the interrelationships among corporate governance variables in terms of the way they complement and substitute for each other as a bundle, as well as the way they both drive and are being driven by corporate performance. Extending on these interrelationship issues, this study is motivated to provide a more robust test of the inter-relationships between the board (and board committees) monitoring variables and corporate financial performance. Second, this study is motivated by the regulatory importance of board monitoring. As has occurred in many countries, the Australian Stock Exchange (ASX) set up a code of

good corporate governance practice to improve the standards of board monitoring. By providing findings about how board monitoring characteristics and devices work together in firms, this study can give better insights to securities market regulators about the relative effectiveness of such characteristics and devices which it includes in its code of good governance. Third, the study provides evidence from an under-researched country. The majority of studies which have examined board structures and directors’ characteristics as mechanisms that can better align management-shareholder interests and improve performance have been conducted in the US, UK, Japan and Germany. Limited research has been undertaken in Australian (Bonn, 2004). The corporate regulatory regime in Australia draws on, but differs in its detail, from other countries. Australian corporate governance regulators responded to the specific circumstances that arose in the late-1990s and early-2000s from corporate collapses and scandals of HIH, Ansett, One Tel and Harris Scarfe.

The data analysis in this study is distinguished from prior research in two ways. First, structural equation modelling is used to model the inter-relationships that exist among the corporate governance variables in a way that can control the effects of multicollinearity. Second, a lagged-year model for the dependent variable is used to limit the problem of endogeneity that can plague multi-variable corporate governance research. 2. Framework of the Study The analysis of governance-performance relationships in this study is limited to data on those governance variables concerned with aspects of board monitoring only. These board-monitoring variables are depicted in the framework in Figure. In this framework, key board-monitoring devices and characteristics are deemed to fall into the following categories: � Structuring of the board (Audit Committee,

Remuneration Committee, Nomination Committee)

� Operating of the board (duality of Chair-CEO, size of board, frequency of meetings)

� Characteristics of directors (financial literacy of directors, independence of directors)


81

Figure 1. Framework of board monitoring

Fama and Jenson (1983) recognise the board as the most important control mechanism available because it forms the apex of a firm’s internal governance structure. Boards are responsible for ensuring that management acts for the interests of the owners. Boards and their committees have the power to recruit, take action, ratify and monitor important decisions where executive managers are concerned (Jensen 1993). In the absence of any formal theory for constructing an effective board, different practices are followed for the construction of the board and the enhancing of its monitoring ability (Hermalin and Weisbach, 2001).

Previous research has emphasised different characteristics of the board. For example, Jensen (1993) considers board composition, board leadership and board size as the prerequisites to capture the board’s monitoring ability. Agerwal and Knoeber (1996) examine a range of governance variables and find that board independence is the only governance mechanism which consistently affects corporate value. With respect to board leadership, the question of conflict of interest due to the dual role of CEO and Chairperson by the same person has been studied. With respect to size, the general finding is that smaller boards are more effective (e.g., Jensen, 1993 and Lipton and Lorsch, 1992). In support of these findings, Hermalin and Weisbach (2001) provide findings of the effects of the size of boards and their proportion of outside directors on several aspects of the effectiveness of monitoring of managers. They find smaller boards and higher proportions of outside directors to lead management teams to take actions that are more in line with shareholders interests, to be more inclined to remove poor performing managers and make better firm acquisition-related decisions.

The effectiveness of monitoring by the boards of directors varies with factors such as board size, board composition, number of meetings, background of directors, CEO/Chair duality and committees. Corporate boards better represent stockholder

interests when they are smaller, contain more outside directors and having a separate person holding the CEO and chairperson position (Jensen, 1993). Further, the background and the experience of the directors have an influence on the monitoring ability of the directors (DeZoort, 1997) 3. Literature on Board Monitoring The extent to which the board behaves as an active monitor in the corporate governance system depends on the way it is structured, its operating features and the characteristics of its directors. This section reviews the literature relating to each of the key board-monitoring variables shown in above Figure.

3.1 Board Size Board size is an important factor for monitoring management. If the size of the board becomes too big, it increases problems of directors’ free-riding and becomes more difficult for directors to express their ideas and opinions in the limited time available to them. It is also argued that large boards are relatively ineffective and are not easy for the CEO to control (Habib and Azim, 2008; Jensen, 1993 and Lipton and Lorsch, 1992). Empirical results in Eisenberg et al. (1998) support the notion that smaller boards enhance firm performance. In contradiction of this argument, Kiel and Nicholson (2003) find evidence in the Australian context that large size boards are not necessarily impediments to good performance.

There is also a potential monitoring problem if the board size is too small. Kiel and Nicholson (2003) suggest that there is an “inverted U” relationship between board size and performance in which adding directors can bring the board to an optimal skills/experience mix level. Beyond that point the difficult dynamics of a large board prevail over the skills/expertise advantage that additional directors might bring. Eight directors is cited as the upper limit

Firm Performance

Boards of Directors

Audit Committee

Remuneration Committee

Nomination Committee

Structuring of the Board

Operating of the Board

Characteristics of Directors

CEO/Chair Duality

Board Meetings

Financial Literacy of Directors

Independence of directors

Board Size


82

and 6.6 as the mean board size in a study by Kiel and Nicholson (2003: 194). Another study by Larcker et al. describes eight as “typical” (2004: 7), while Leblanc and Gillies note that eight to eleven is viewed as optimal (2004: 5). 3.2 Number of Meeting Major decisions of the firm are made at the board meeting. Therefore, it is important that directors spend a considerable amount of time for board meetings. Board activity, measured by board meeting frequency, is an important dimension of board operations. Vafeas (1999) finds that the annual number of board meetings is inversely related to firm value. However, their results were driven by increases in board activity following share price declines.

Sometime, it is argued that the quality of time directors spend in board meetings is important rather than the quantity of time. However, quantity of director’s time is emphasised by shareholder activist groups and labour unions, where their measure of a director’s performance includes such factors as attendance and number of directorships. Therefore the number of board meetings is an important consideration in judging the effectiveness of monitoring mechanisms.

3.3 CEO/Chair Duality The two most important positions of firms are the Chair of the board and Chief executive officer (CEO). The position of chairperson significantly influences the outcome of board decisions because he/she controls the board meetings, sets its agenda, makes committee assignments and also influences the selection of new directors. The position of CEO is also influential as he/she is responsible for any operating and financial decision making of the firm.

If the same person holds the position of both CEO and Chair there will be a problem of proper monitoring of the performance as the CEO will be able to control the board and will reduce the board’s independence from management and make decisions in their self-interest and at the expense of shareholders. Therefore, to maintain independence, it is necessary that the board is independence from the CEO (Hermalin and Weisbach. 2001). Cadbury (1995) also recommends that the role of the chairman of the board of directors should be separate from that of the CEO.

A number of empirical studies have provided important insights into the relationship of leadership structure to performance (Heracleous, 2001; Leblanc and Gillies, 2004; Rechner and Dalton, 1989 and 1991; and Baliga, Moyer and Rao, 1996). However, the evidence is far from conclusive. For example, Heracleous (2001) provides a literature review of evidence that shows that whether the CEO and Chair are separate or the same person does not, on its own, appear to make much difference to performance. Leblanc and Gillies (2004) argue that empirical research has failed to find a clear link between the

separation of CEO and Chair positions and enhanced firm performance. Rechner and Dalton (1989) examine shareholder returns over a five-year period (1978 – 1983) and find no significant distinction between the performances of separated and combined structure firms.

3.4 Independence of Directors Independent directors21 are directors who do not hold any executive position in the company or have any direct or indirect interest in the company. It is generally argued that independent directors, because of their lack of interest in any financial benefit from the firm, are more likely to protect shareholders interests and reduce the agency problem. Empirical results also support the argument that outside directors are more effective monitors and a critical disciplining device for managers (Ahmed, Hossain and Adams, 2006; Davidson, Goodwin-Stewart, and Kent, 2005; Hermalin and Weisbach, 1988). Fama (1980) and Fama and Jensen (1983) argue that board outsiders, by providing expert knowledge and monitoring services, add value to firms. Being financially independent of management, independent directors have the ability to withstand pressure upon management.

There is much professional and research interest in the role of non-executive directors’ monitoring role in corporate governance. However, there are mixed findings. Pfeffer (1972) and Zald (1967) reveal a positive association between improved efficiency and corporate performance when boards of directors are dominated by non-executive directors. These results have been disputed by Kesner (1987), Pearce (1983) and Vance (1964 and 1978) who found a superior financial performance in firms that had boards dominated by executive directors. This finding was supported by Dechow, et al. (1996) and Beasley (1996) who conclude that there is a negative relationship between the number of independent directors and the incidence of financial statement fraud. However, Hermalin and Weisbach (1991) did not find any significant relationship in regression of board composition on firm performance.

Although there are conflicting findings in the previous research, in general this paper views that outside directors improve board quality by increasing its independence from management and working for the best interest of the shareholders (Cadbury 1995). Independent non-executive directors are regarded as being in a better position than non-independent directors to effectively monitor executive management. Independent non-executive directors in

21 Suchard et al, (2001) mentioned that Australian board members can be classified into two broad categories, executives and non-executives. While the executives are employed by the firm, the non-executives can be further classified into two categories, independent and non-independent. For monitoring purpose independent directors are more effective.


83

turn have incentives to develop reputations as experts in decision control and monitoring (Fama and Jensen 1983).

One of the common limitations of the above studies is that most of them focused on the executive and non-executive classification as an indicator of the independence of the board. However, not all non-executive directors are independent (Psaros and Seamer, 2002). To capture the monitoring ability of independent directors they need to classify into three categories: insider, grey and outsider. This three way classification of the directors was first done by Baysinger and Butler (1985)22. Commonly, ‘insider’ directors refer to the directors with the executive position in the company; ‘grey’ directors are not full time employees of the company but are associated with the company in other capacities (such as acting as professional adviser or consultant, supplier or customer, or previous employee of the company); and ‘outsiders’ are those who have no affiliation with the firm except for their directorship. Previously, no monitoring effects were identified when ‘grey’ directors were excluded from non-executive directors. 3.5 Educational/Working Background of the Directors Directors educational background and work experience is an important consideration in the monitoring process. The working experience and/or financial background is expected to lead to better monitoring of management. DeZoort (1997) and Bull and Sharp (1989) emphasise on the board members to have accounting and auditing expertise. Ramsay mentioned that financial literacy is an important component of the general standards of care, skill and diligence required of company directors (2001: 155). It is expected that the directors who are financially literate can monitor management efficiently. To capture the educational/experience background it is important to consider whether the directors had worked in any firm for more than five years as directors or whether they had any business or economics background. Higher levels of educational background and stronger work experience help better understand the business and properly monitor management. 3.6 Committees of the Board Another factor in considering the monitoring ability of the board is the ability of different board committees, especially audit, compensation and nomination committees. The legislation requires that these committees be independent for the purpose of proper monitoring (Austin, 2002).

Audit Committee: The primary function of the audit committee is to review management information, financial statements and internal control

22 They use Australian Accounting Standard AASB 1017: Related Party Disclosure to classify directors into three categories for better reflection of the board composition.

system (Bosch, 1995; Klein, 1998). The importance of audit committees as a corporate monitoring mechanism has been emphasised by many researchers in recent years (e.g., Chen et al., 2005, Abbot and Parker, 2000).

An important recommendation by the Ramsay Report (2001) is the mandatory rule for all Australian Stock Exchange (ASX) listed companies to have an audit committee23. Australian companies have adopted the recommendation by Corporate Law Economic Reform Program Act 2004 (CLERP 9, Commonwealth of Australia 2004), where the top 500 companies listed on the Australian stock exchange are required to have an audit committee. A report issued by the Joint Committee of Public Accountants and Audit (JCPAA, 2002) also highlighted the need for all listed companies to have an audit committee24. According to the Australian Stock Exchange Corporate Governance Council (2003), the audit committee should consist of: (i) only non-executive directors; (ii) a majority of independent directors; (iii) the independent chairperson, who is not chairperson of the board; and (iv) at least three members (see recommendation 4.3: ASX CGC, 2007).

Similar to the board of directors, too many or too few meetings both are the threat to effective decision making of the audit committee. Again if the members of the audit committee are financially literate, it is expected that they will work as an effective monitor. It is expected that effective audit committee monitoring will have an impact on firm performance. Number of audit committee meetings with impendent and financially literate directors will work as an effective monitor for the audit committee. In general, monitoring ability of the audit committee is measured by: the number of audit committee meetings, the proportion of independent directors in the committee and the proportion of financial literate directors in the committee.

Compensation Committee: Compensation committee has become more common in the wake of the Cadbury Committee’s 1992 report. The existence of Compensation committees is consistent with agency theory, which advocates the separation of

23 The Ramsay Report summarises and recommends limited adoption of best practices in the USA, UK and Canada. In addition to the proposal to mandate the formation of audit committees for all listed companies, the Ramsay Report proposes a threshold test of market capitalization to determine the proportion of independent audit committee member required. This initiative takes into consideration the disproportionate cost requirement for smaller listed companies to have an independent audit committee. 24 The JCPAA recommendations on the composition and responsibilities of audit committees are the same as those prescribed in the Ramsay Report. Furthermore, the JCPAA argued that the cost of setting up an audit committee should be an obligation for companies seeking to access the Australian capital market via a listing on the Australian Stock Exchange.


84

management from control (Barkema and Gomez-Mejia, 1998). The main function of compensation committees is to determine and review remuneration packages for senior management of the company (Klein, 1998). When determining compensation it is necessary to consider the company’s needs together with the interests of its shareholders and other stakeholders (Bosch, 1995). There has been an increasing demand for greater accountability for remuneration, substantially contributing to the growth in adoption of the compensation committee (Bosch, 1995). This report recommended that the appointment of Compensation committees consisting wholly or mainly of non-executive directors and chaired by non-executive director.

Compensation committees have an important role in the monitoring of boardroom control. The Compensation committee supports and advises the board in fulfilling its responsibilities to shareholders by appropriately design the Compensation policy for directors, chief executive officer and other senior executives. Monitoring by the Compensation committee is captured by the number of meetings held by the committee and the proportion of independent directors in the committee.

Nomination Committee: The ability of non-executive directors to perform the monitoring function is related to their independence, which in turn is related to director selection by the nomination committee (Conyon and Peck, 1998). Theoretically, directors are selected by the shareholders but, in practice, shareholders simply ratify candidates selected by the board itself (see Vaefas, 1999). Therefore, the appointment of the directors is a critical issue to determine monitoring ability; especially if outside board directors are selected by an incumbent CEO heightening the prospect that they may not execute their duties in a manner congruent with shareholder interests (Hart, 1995). Jensen (1993) also argues that the nomination process is often dominated by a powerful CEO who selects directors under his influence in order to contain the intensity of board monitoring. Board control is more effective in companies that have introduced a nomination committee to select and recruit directors.

The presence of the nomination committee ensures that the board is comprised of individuals who are best able to discharge the responsibilities of a director, having regard to the law and the highest standards of governance, by assessing the skills, knowledge, experience and diversity required on the board and the extent to which each are represented; nomination committee also establish processes for the review of the performance of individual directors and the board as a whole (Conyon and Peck, 1998).

In this selection process presence of nomination committee ensure effective and efficient monitoring through the non-executive directors and frequent meeting. Therefore, this study considers the monitoring of the nomination committee by the number of meetings and the proportion of

independent directors in the committee. The number of nomination committee meetings captures the willingness to select the right person for the firm. A successful nomination decision requires a good discussion of the companies’ needs and proper selection of the committee members which in tern requires directors to meet several times. And the proportion of non-executive directors demonstrates the independence of the nomination committees.

The following framework is developed from the above discussion on different board monitoring characteristics. In this framework key characteristics of board monitoring are: compositions (board size, number of meeting, and proportion of independent directors); characteristics (background of directors, separation of CEO and Chair) and committees (audit, Compensation and nomination committees). 4. Hypothesis Development The above literature review and discussion leads to the following hypotheses related to boards of directors and firm performance:

The first hypothesis is based on the accounting performance of the firm, as reflected in return on assets. It is expected that the boards of directors’ monitoring will influence the management to work for the best interest of the company, including the reporting to shareholders of relevant and reliable accounting information. This quality of accounting information will be reflected in the accounting based performance of the firm.

H 1 If other things remain the same, monitoring

by boards of directors is positively related to

the return on assets of a firm.

The second hypothesis is based on the market performance of the firm, as reflected in earning per share. In the presence of boards monitoring, management will disclose better quality accounting information which will have an impact on the market. These disclosures of information by management are expected to have a positive impact on market performance measures of the firm.

H 2 If other things remain the same, monitoring

by boards of directors is positively related to

the earning per share of a firm.

5. Method of Analysis This paper uses structural equation modelling (SEM) to find out the relationship between the board monitoring and firm performance. Multiple regressions, multivariate analysis of variance and discriminate analysis provide researchers with powerful statistical tools to address a wide range of corporate governance issues. They have been widely used in prior empirical studies on governance-performance relationships. However, the major limitation of these analytical techniques is that it is only possible to examining a single relationship at a time. Although MANOVA allows for multiple


85

independent and dependent variables, it only represents a single relationship between the dependent and independent variables (Hair et al., 2006). Likewise, multiple regression represents single path relationships between each independent variable and the dependent variable. The main difference between SEM and other multivariate techniques is the use of independent relationships between sets of variables. That is, SEM estimates a series of separate, but interdependent, multiple regression equations simultaneously by specifying the structural model. A structural model can express relationships among independent and dependent variables, even when dependent variables become independent variables in other relationships.

Moreover, in SEM it is possible to include both observed and latent variables in the model. Observed variables have data and are also usually continuous. Latent variables are not directly observed. To observe latent variables, the model should be built which expresses latent variables in terms of observed variables. The latent variables in SEM are continuous variables and can, in theory, have an infinite number of values. Due to the interrelations among the variables, this study selects structural equation modelling as an appropriate statistical tool to test the research questions. No previous corporate governance research has used structural equation modelling to consider the complex interrelationship among the monitoring variables.

There are three advantages of using SEM in this study. Firstly, it has the ability to incorporate latent variables into the analysis. A latent variable is an unobserved concept that can only be approximated by observable or measurable variables. Secondly, in all multivariate analysis it is assume that there is no error in the variables. However, it is well know both in practical and theoretical perspectives, that it is not possible to perfectly measure the concept as there is always some degree of measurement error. By considering this type of error SEM improve the statistical estimation. Thirdly, SEM is a powerful method to deal with multicollinearity in sets of predictor variables. Multicollinearity occurs when two or more variables are not independent. In genera there is a strong interdependence among the corporate monitoring variables. If this interdependence is not considered there is a possibility that the result will be poor and misleading. This study handles the problem of multicollinearity by using the structural equation modelling.

Like all other statistical methods, SEM requires a careful consideration of some basic factors which might affect the research design and analysis of the models. Following are some assumptions on which SEM is based:

(i) The first assumption of the SEM is that the sample should be of a reasonable size. Sample size in SEM generally varies depending on analysis procedures used and with the model characteristics.

(ii) Dependent variables are continuously and normally distributed. According to the normal distribution, it is expected that the sample mean and sample variance tend to be normally distributed as the sample size becomes large. SEM is designed for variables that are assumed to have an underlying continuous distribution. Theoretical basis for model specification is particularly important for SEM because it is considered a confirmatory analysis; that is, it is useful for testing and potentially confirming theory.

(iii) SEM models can never be accepted; they can only fail to be rejected. This assumption leads to provisionally accepting a given model. In SEM in most instances it is recognized that there are equivalent models that fit equally as well as their own provisionally accepted model. Any of these models may be correct because they fit the data as well as the preferred model. This is an attempt to eliminate alternative models, and by extension alternative explanations, but this is not always possible. The use of SEM thus entails some uncertainty, particularly with cross-sectional data that are not collected under controlled conditions.

5.1 Fitness of the Structural Models Evaluating the model is one of the most difficult issues connected with research based on structural equation modelling. The overall goodness-of-fit for structural equation models depends on many factors. There is no single statistical test that describes the goodness-of-fit. As mentioned by Hair et al. (2006), ‘No single magic value for the fit indices separates

good from poor models, and it is not practical to

apply a single set of cutoff rules to all measurement

models and for that matter to all SEM models of any

type’ (p.705) . The application of multiple fit measures will enable a consensus across types of measures regarding acceptability of the proposed model to be attained. Various goodness-of-fit measures assess the results from three perspectives:

Normed Chi-Squire (X2/df): Joreskog and

Sorbom (1994) proposed that the chi-square be adjusted by the degrees of freedom to assess model fitness. This measure can be termed the normed chi-square (X2/df), and is the ratio of the chi-square divided by the degrees of freedom. A normed chi-squire statistics, parallel to the F-change statistics consulted in hierarchical regression analysis, is used to determine which model is better for the observed data (Hoyle, 1995). This measure provides two ways to assess inappropriate models: firstly, a model that may be over-fitted, thereby capitalising on chance, typified by values less than 1.0; and secondly, models that are not yet truly representative of the observed data and thus need improvement, having values greater than an upper threshold. However, because the chi-square value is the major component of this measure, it is subject to sample size effects. The normed chi-square, however, has been shown to be somewhat unreliable. Marsh and Hocevar (1985) find


86

that different researchers have recommended using ratios as low as 2 or as high as 5 to indicate a reasonable fit. Carmines and McIver (1981) find that degrees of freedom with a range of 2 to 1 or 3 to 1, indicate an acceptable fit between the hypothetical model and the sample data.

Goodness–of-Fit Index: The goodness-of-fit index (GFI) is based on the squared residuals from prediction compared to the actual data but is not adjusted for the degrees of freedom. It ranges from 0 (poor fit) to 1.0 (perfect fit) and higher values indicate better fit, but no absolute threshold levels for acceptability have been established (Hair et al., 2006). In this research, GFI is used to compare the fit of different models.

Adjusted Goodness-of-Fit Index (AGFI): The adjusted goodness-of-fit is an extension of the goodness-of-fit index, adjusted by the ratio of degrees of freedom for the proposed model to the degrees of freedom for the null model. It is quite similar to the parsimonious normal fit index (PNFI) described below, and a recommended acceptance level is a value greater than or equal to .90 (Hair et al., 2006).

Comparative Fit Index (CFI): Comparative fit index25 (CFI) represents comparisons between the estimated model and a null model. Values lie between 0 and 1.0 and larger values indicate superior goodness-of-fit. If the CFI is less than one, then the CFI is always greater than the TLI. Because CFI has many desirable properties including its relative, but not complete, insensitivity to model complexity, it is among the most widely used indices. The CFI has been found to be more appropriate in a model development strategy (Hair et al., 2006).

Root Mean Square Error of Approximation (RMSEA): The root mean square error of approximation (RMSEA) calculates the discrepancy from the population per degree of freedom.26 The value is representative of the goodness-of-fit that could be expected if the model was estimated in the population, not just the sample drawn for the estimation. Browne and Cudeck (1993) suggest that a value of the RMSEA of about .05 or less indicates a close fit of the model in regard to the degrees of freedom, while a value of about 0.08 or less provides a reasonable error of approximation. A model with a RMSEA greater than 0.1 should be rejected. Browne and Cudeck (1993) report that, based on an empirical examination of several measures, the RMSEA is best suited in a confirmatory or competing models strategy with large samples. These accords with the models developed in this research.

25 d(Null Model) - d(Proposed Model) CFI = -------------------------------------------------------------d(Null Model)

26 RMSEA = v[(c2/df - 1) /(N - 1)]

The ultimate goal of any of these fit is to assist the researcher in discriminating between acceptability and unacceptably specified models. However, the standard indicates good fit is still a question. Academic journals are replete with SEM results citing a .90 value on key indices, such as the CFI or GFI, as indicating an acceptable model. Some may cite precedent from a previously published paper. Others times the .90 rule is simply cited as a reasonable ad hoc rule with no support from previous research. However, the main aim of using SEM in this research is not to get the good fit but to test the existing theory and practice in corporate governance. SEM is not used to get a good fit; it is used to test theory (Hair, et. al., 2006).

5.2 Definition of the Variables

Latent Variable: Boards of Directors

Monitoring According to agency theory, the main task of the board is to monitor and control management on behalf of the shareholders. Boards of directors are responsible in adopting monitoring characteristics to ensure that management behaviour and actions are consistent with the interest of the owners. To ensure this there are different observed variables that reflect the monitoring ability of the board of directors. This study captures the monitoring ability of the board through the various board and board committee characteristics as detailed in the literature review section. This paper manually collected the information of number of directors, non-executive and independent directors, background expertise of directors and number of meetings of boards and the various committees from the Aspect financial database.

Observed Variable: Firm Performance Previous empirical studies use different types of performance measures to observe the relationship between monitoring variables and performance. As an indicator of performance, this study uses return on assets (ROA) and earning per share (EPS).

Measuring Accounting Performance: Return on Assets (ROA) To measure the accounting performance this study uses the ROA measures. ROA is calculated on the basis of accounting information that is disclosed in the financial report of the firm. The following discussion describes the particular calculation procedure followed to calculate these ratios.

ROA tells an investor how much profit a company generated for each dollar in assets. ROA measures a company’s earnings in relation to all of the resources it had at its disposal (the shareholders’ capital plus short and long-term borrowed funds). Thus, it is considered the most stringent test of return to shareholders. If a company has no debt, the return


87

on assets and return on equity figures will be the same.

There are two acceptable ways to calculate ROA. The lower the profit per dollar of assets, the more asset-intensive a business is. The higher the profit per dollar of assets, the less asset-intensive a business is. All things being equal, the more asset-intensive a business, the more money must be reinvested into it to continue generating earnings. ROA is a key measure of a company's profitability; it is calculated by earnings-before-interest divided by its total assets. Return on assets essentially shows how much profit a company is making on the assets used in its business:

Earning before interestRe turn on Asset (ROA) =

Total assets less outside equity interest

Judge and Zeithaml (1992) in using the LISREL

statistical method, found a positive relationship between board involvement in the strategic decision process and the average return on assets of companies. Measuring Market Performance: Earning per Share (EPS) EPS is a key ratio used in share valuations. It shows how much of the company's profits, after tax, each shareholder owns. This is the single most popular variable in dictating a share's price. EPS is an important measure as it indicates the profitability of a company. The portion of a company's profit allocated to each outstanding share of common stock is calculated as:

Net Income - Dividends on preferred stockEarning Per Share (EPS) =

Average outstanding shares

In a study involving 139 companies from Fortune

500 firms, Pearce II and Zahra (1991) found that there is a positive relationship between participative boards and earnings per share of firms.

Control Variables: Industry and Size

Industry may have several effects on monitoring. For example, banking and finance, and insurance companies have higher monitoring costs involved which results in greater scrutiny of the firms and increased incentives for high quality financial statements. This study controls for the industry effect on firm performance by using the industry adjusted performance measurement. To obtain industry adjusted performance measures firms are classified in different industry categories according to the four digits GICS industry classification. Industry averages are calculated for each performance measures and then find the difference with each performance measures for the firm.

Another variable that this study controls for is the size of the firm. As mentioned before moral-hazard problems are more prominent in large firms (Jensen, 1993). Large firms are under more internal and

external monitoring which eventually increases the difficulty of monitoring and also increase the cost of monitoring. This study uses total assets as a proxy of firm size as a control valuable to make the result easily comparable. 6. Sample Selection and Data Sources This study uses archival sources, such as the Aspect

Financial Analysis database (hereafter Aspect) and the Connect 4 and the Aspect Huntley Financial

Analysis (hereafter Aspect Huntley), for collecting data. The Aspect provides comprehensive data for all ASX listed companies. Similarly the Connect 4 provides annual reports of the top listed companies. The information provided in these two websites is used to track and collect information on corporate monitoring variables, i.e., boards of directors, committees, external auditors and shareholders information. The Aspect Huntley, one of Australia's most comprehensive sources of data for listed companies, was used to collect performance measures information. This information was cross-checked with the annual reports obtained from Aspect and Connect 4. The ASX website was used to obtain industry classifications for each company.

This study uses data of three years observations from top 500 companies listed in the Australian Stock Exchange (ASX). This observation period of 2001 – 2003 was chosen to include the most recent data available at the time of commencing this study. And the choice of publicly listed companies was based on the most efficient data available and the presence of audited financial statements. Initially all the listed companies are downloaded from the ASX website for the year 2001. Next step is to sort them according to their market capitalisation. Form the total list the top 500 companies are selected for 2001. The same procedures are followed for year 2002 and 2003. All corporate governance and financial information are based on year end financial information, which helps to keep consistency in the collected information. In this study a repeated measures design is used, where the same data are collected on each variables across three consecutive periods. In relation to industry classification, most of the companies in the sample operate in the financial sector (22%), followed by material sector (18%). Remaining 60 percent are involved in energy, industrials, consumer discretionary, consumer staples, health care, information technology, telecommunication and utilities.

In examining the relationship between monitoring and performance, this study addressed the impact of monitoring characteristics on the lagged year performance. It is reasonable to believe that the affect of monitoring characteristics will be reflected in the next year’s performance. For example, when assessing the effects of board monitoring in 2001 this study relates it to performance in 2002. Because of


88

this lag, performance measurements for the years 2001 – 2004 are collected.

The details of the sample selection procedures are shown in Table 1. Since performance measures are the independent variables any companies without having the required information on firm performance in the lagged year are excluded from the sample. Therefore, companies de-listed or suspended in the following year are excluded. Consequently, 25 companies in year 2001, 28 companies in 2002 and 37 companies in year 2003 were excluded. In total this study excludes 90 firms from the sample of 1500. It includes the companies that changed their name during the study period. Of these 1500 companies, 1410 have all the required information for this analysis. Characteristics of the Board Monitoring Variables Results in Table 1 reveal that average board size is 6 directors (maximum=17, minimum=3). Average number of board meetings is 10 per annum (maximum=37, minimum=2). More than 82% of the firms (1,158) have a board with a majority of independent members. In the sample, there are 168 firms (12%) where the roles of chair and CEO are occupied by one person. One hundred fifty three firms (7%) do not have any audit committee meetings, and 66% of firms (929) have 2 to 4 audit committee meetings per annum. Eighty percent of firms (1,143) have 1 to 4 independent directors on the board. In 383 (27%) companies there are no financially literate members on the board and in 229 (16%) all directors are financially literate. In the sample, 79% (1,117 firms) have between 1 to 3 meetings per annum. Only 288 (20%) firms have a nomination committee. Amongst these firms, 239 have 1 to 4 nomination committee meetings per annum. Only in three committees are there no non-executive directors (Table 1).

[TABLE 1 ABOUT HERE] 7. Results The results were generally consistent across different year’s models when examining the impacts on board monitoring on firm performance. The discussion of the results will begin with observing the overall fitness of the model followed by discussion on research findings. This study found that hypothesised models fit the data well in terms of absolute, relative and parsimonious fitness for the year 2001, 2002 and 2003.

Boards of directors are the most active monitors of management. The efficiency in monitoring improves when independent, financially literate directors make up the board, and the CEO and Chair are separate persons. Yet, whether monitoring by boards affects firm performance remains unresolved

in the literature. The following results show that such monitoring has a consistent and statistically significant relationship with firm performance after controlling for endogeneity and multicollinearity problem.

[TABLE 2 - 4 ABOUT HERE]

The data for monitoring and accounting performance for 2001 lagged year (Table 2, 3 and 4) show that the impacts of board of directors on ROA is statistically significant (P<0.01), suggesting different board and its’ committees’ monitoring work together to have an effect on the return on assets. This result is consistent with other lagged year models of 2002 and 2003.

[TABLE 5 - 7 ABOUT HERE]

Same result was found when examining the relationship between board monitoring and EPS. Lagged year model of 2001, 2002 and 2003 showing a statistically significant relation, suggesting that different board and committees’ monitoring mechanisms work together to have an effect on the earning per share (Table 5, 6 and 7) show a consistent finding of a significant result in all of the above year.

This study contradict previous research by Hermalin and Weisbach (1991), Mehran (1995), Klein (1998) and Bhagat and Black (2000), who examined the influence of board monitoring on firm performance and failed to find any relationships. However, those studies (mentioned above) did not consider the multicollinearity and endogeneity issues of board monitoring. MacAvoy and Millstein (1999) argue that one reason for not finding any relationship is because they have used “old” data – that is, data that preceded the board monitoring role in the current-year and performance rather than using lagged year performance. Therefore, after controlling the endogeneity and multicollinearity issues, this study finds difference result when examining board monitoring with performance.

Sensitivity Tests Although not reported, this study examined the robustness of the results by taking only the firms which existed in the three year sample period. In this respect there are 285 companies among the top 500 which are listed throughout the study periods of 2001 – 2003. However, the results are consistent with the full sample.

8. Conclusion This study examines the effect of board monitoring on firm performance in Australian context. SEM analysis suggests that there is a significant relation of monitoring by boards of directors and firm performance. In a broader context the finding of this study will add value in the discussion of the board


89

monitoring features and their influence on firm performance.

There are already some steps to enhance the corporate governance code of conduct, which have initiated some changes in the corporate governance and reporting practices. For example the Australian Stock Exchange (ASX) has recently released their corporate governance guidelines will sets out the principles of best practice for companies listed on the ASX. Among others Government’s Corporate Law Economic Reform Program (CLERP 9) also suggests reforming and adopting principles that provide good governance practice.

There are some limitations of this study: firstly, this study does not includes all the companies listed on the ASX. The result might be different if all the companies are included in the sample. Secondly, this study was done for the years 2001 – 2003, which do not include the change that took place in the year 2004.

Future studies should take into consideration of these limitations. The findings suggest that there are significant positive direct relationships between the board monitoring characteristics and firm performance. There remains a need for additional studies to address how the monitoring variables specifically work as complementary and substitute mechanisms to each other as suggested by Rediker and Seth (1995); Agrawal and Knoeber (1996); Fernández and Arrondo (2005).

References 1. Abbott, L.J. and Parker, S. (2000), “Auditor selection

and audit committee characteristics”, Auditing: A

Journal of Practice and Theory, Vol. 19, pp. 47 – 66. 2. Agrawal, A., and Knoeber, C.R. (1996), “Firm

performance and mechanisms to control agency problems between managers and shareholders”, Journal of Financial and Quantitative Analysis, Vol. 31, pp. 377 – 397.

3. Ahmed, K., Hossain, M. Adams, M. B. (2006), “The Effects of Board Composition and Board Size on the Informativeness of Annual Accounting Earnings”, Corporate Governance: An International Review, Vol. 14, pp. 418 – 431.

4. Austin D.V. (2002), “Independence of the board committees”, Hoosier Banker, Vol. 86, p. 10.

5. Australian Stock Exchange Limited Corporate Governance Council (ASX CGC) (2007) Principles of

good corporate governance and best practice

recommendation, Australian Stock Exchange Limited, Sydney.

6. Baliga, B. R., Moyer, R.C. and Rao, R. S. (1996), “CEO duality and firm performance: what’s the fuss?”, Strategic Management Journal, Vol. 17, pp. 41 – 53.

7. Barkema, H.G., and Gomez-Mejia, L.R. (1998), “Managerial compensation and firm performance: a general research framework”, The Academy of

Management Journal, Vol. 41, No. 2, pp. 135 – 145. 8. Baysinger, B. and Butler, H. (1985), “Corporate

governance and the board of governance: performance effects of changes in board composition”, Journal of

Law and Organization, Vol. 1, pp. 101– 124.

9. Beasley, M. (1996), “An empirical analysis of the relation between board of director composition and financial statement fraud”, The Accounting Review, Vol. 71, No.4, pp. 443 – 466.

10. Bhagat, S., and Black, B. (2000), “Board independence and long term firm performance”, Working Paper, University of Colorado.

11. Bonn, I. (2004), “Board structure and firm performance: evidence from Australia”, Journal of the

Australian and New Zealand Academy of Management, Vol. 10, No. 1, pp.4 – 24.

12. Boo, E and Sharma, D. (2008), “Effect of regulatory oversight on the association between internal governance characteristics and audit fees”, Accounting

and Finance, Vol. 48, pp. 51-71. 13. Bosch, H. (1995), Corporate Practice and Conduct,

3rd ed. Business Council of Australia, Pittman Publishing.

14. Browne, M.W. and Cudeck, R. (1993), “Alternative ways of assessing model fit” cited in Hoyle, R.H.(Eds.) (1995), Structural Equation Modeling: Concepts,

Issues, and Applications, Sage publications, USA. 15. Bull, I., and Sharp, F.C. (1989), “Advising clients on

tread way audit committee recommendations”, Journal

of Accountancy, Issue. February, pp. 45-52. 16. Cadbury Committee, (1992), Report of the committee

on the financial aspects of corporate governance, Professional Publishing Ltd, London.

17. Cadbury, A. (1995), Committee on the financial

aspects of corporate governance: Compliance with the

code of best practice. Gee Publishing, London. 18. Carmines, E. and Mclver, J.M. (1981), “Analysing

models with unobserved variables: analysis of covariance structures” cited in Hair, J. F., Anderson, R.E., Tatham, R.L. and Black, W. C. (1998), Multivariate Data Analysis, fifth edition, Prentice-Hall, New Jersey, USA.

19. Chen, Y. M., Moroney, R. and Houghton, K. (2005), “Audit committee composition and the use of an industry specialist audit firm”, Accounting and

Finance, Vol. 45, pp. 217 – 239. 20. Conyon, M.J. and Peck, S.I. (1998), “Board control,

compensation committees, and top management compensation”, Academy of Management Journal, Vol. 41, No. 2, pp. 146 – 157.

21. Davidson, R., Goodwin-Stewart, J., and Kent, P. (2005), “Internal governance structures and earnings management”, Accounting and Finance, Vol. 45, pp. 241 – 267.

22. Dechow, P.M., Sloan, R.G. and Sweeney, A.P. (1996), “Causes and consequences of earnings manipulation: an analysis of firms subject to enforcement actions by the SEC”, Contemporary Accounting Research, Vol. 13, pp. 1-36.

23. DeZoort, F.T. (1997), “An investigation of audit committees’ oversight responsibilities”, Abacus, Vol. 33, No.2, pp. 208-227.

24. Eisenberg, T., Sundgren S. and Wells, M.T. (1998), “Larger board size and decreasing firm value in small firms”, Journal of Financial Economics, Vol. 48, pp 35 – 54.

25. Fama, E. F. (1980), “Agency problems and the theory of the firm”, Journal of Political Economy, Vol. 88, pp. 288–307.

26. Fama, E. F. and Jensen, M.C. (1983), “Separation of control and ownership”, Journal of Law and

Economics, Vol. 26, pp. 301-325.


90

27. Fernández, C. and Arrondo, R. (2005), “Alternative internal controls as substitutes of the board of directors”, Corporate Governance: An International

Review, Vol.13, No. 6, pp.856 – 866. 28. Habib, A., and Azim, I. (2008), “Corporate governance

and value-relevance of accounting information: evidence from Australia”, Australian Research

Journal, Vol. 21, No. 2, pp. 167-194. 29. Hair, J. F., Black, W. C., Babin, B.J., Anderson, R.E.,

and Tatham, R.L. (2006), Multivariate Data Analysis, sixth edition, Prarson Education International, New Jersey, USA.

30. Hart, O. (1995), “Corporate governance: some theory and implications”, Economic Journal, Vol. 105, pp. 678 – 689.

31. Heracleous, L. (2001), “What is the impact of corporate governance on organisational performance?” Faculty of Business Administration, National University Singapore, Singapore.

32. Hermalin, B. E.,Weisbach, M. (2003), “Boards of directors as an endogenously determined institution: a survey of the economic literature”, Economic Policy

Review, Vol. 9, pp. 7–26. 33. Hermalin, B.E. and Weisbach M.S. (1988), “The

determinants of board composition”, RAND Journal of

Economics, Vol. 19, pp. 589 – 606. 34. Hermalin, B.E. and Weisbach M.S. (1991), “The

effects of board composition and direct incentives on firm performance”, Financial Management, Issue. Winter, pp. 101 – 112.

35. Hermalin, B.E. and Weisbach, M.S. (2001), “Board of directors as an endogenously determined institution: a survey of the economic literature”, FRBNY Economic

Policy Review, pp. 1 – 20. 36. Hoyle, R.H. (1995), Structural Equation Modeling:

Concepts, Issues, and Applications, SAGE Publication, Thousand Oaks, California. USA.

37. JCPAA (Joint Committee of Public Accounts and Audit), (2002), Review of Independent Auditing by

Registered Company Auditor, Parliament of Australia, Canberra, Australia.

38. Jensen, M. C. (1993), “The modern industrial revolution, exit and the failure of internal control systems”, Journal of Finance, Vol. 48, pp. 831-880.

39. Joreskog, K.G. and Sorbom, D. (1994), LISREL VI

user’s guide (3rd ed.). Scientific software, Mooresville, IN, USA.

40. Judge, W.Q. and Zeithaml, C.P. (1992), “Institutional and strategic choice perspectives on board involvement in the strategic decision process”, Academy of

Management Journal, Vol. 35, pp. 766 – 794. 41. Kesner, I.F. (1987), “Directors’ stockownership and

organizational performance: an investigation of fortune 500 companies”, Journal of Management, Vol. 13, pp. 499 – 508.

42. Kiel, G.C. and Nicholson, G.J. (2003), “Board composition and corporate performance: how the Australian experience informs contrasting theories of corporate governance”, Corporate Governance: An

International Review, Vol. 11, No. 3, pp. 189 – 205 43. Klein, A. (1998), “Firm performance and board

committee structure”, Journal of Law and Economics, Vol. 41, No. 1, pp. 275 – 299.

44. Kole, S.R. (1996), “Managerial ownership and firm performance: incentives or rewards?” Advances in

Financial Economics, Vol.2, pp. 119-149. 45. Larcker, D. F., Richardson, S. A. and Tuna, I. (2004),

“Does corporate governance really matter?”, Working

Paper, Wharton school, University of Pennsylvanil, USA.

46. Leblanc, R. W. and Gillies, J. (2004), “Improving board decision-marking: an inside view. alternatives beyond imagination”, Company Directors Conference, Port Douglas, Queensland, AICD.

47. Lipton, M., Lorsch, J., (1992), “A modest proposal for improved corporate governance”, Business Lawyer, Vol. 48, pp. 59–77.

48. Loderer, C. and Martin, K. (1997), “Managerial ownership and firm performance: incentives or rewards?” Advances in Financial Economics, Vol.2, pp. 119-149.

49. MacAvoy, P. and Millstein, I.M. (1999), “The active board of directors and its effect on performance of the large publicly trades corporation”, Journal of Applied

Corporate Finance, Vol. 11, No. 4, pp. 8 – 20. 50. Mehran, H. (1995), “Executive compensation structure,

ownership and firm performance”, Journal of

Financial Economics, Vol. 38, pp. 163 – 184. 51. Pearce II, J.A. and Zahra, S.A. (1991), “The relative

power of CEOs and boards of directors: Associations with corporate performance”, Strategic Management

Journal, Vol. 12, pp. 135-153. 52. Pearce, J.A. (1983), “The relationship of internal

versus external orientations to financial measures of strategic performance”, Strategic Management

Journal, Vol. 4, pp. 297 – 306. 53. Pfeffer, J. (1972), “Size and composition of corporate

boards of directors: the organization and its environment”, Administrative Science Quarterly, Vol. 16, pp. 218 -228.

54. Psaros, J. and Seamer, M. (2002), Corporate

Governance Report, Horwath, Newcastle 55. Ramsay, I., (2001), Independent of Australian

Company Auditors: Review of Current Australian

Requirements and Proposals for Reform; Report of University of Melbourne, Melbourne, Victoria.

56. Rechner, P.L. and Dalton, D.R. (1989), “The impact of CEO as board chairperson on corporate performance: evidence vs rhetoric”, Academy of Management

Executive, Vol. 3, pp. 141 – 143. 57. Rechner, P.L. and Dalton, D.R. (1991), “CEO duality

and organisational performance: A longitudinal analysis”, Strategic Management Journal, Vol. 12, pp. 155 – 160.

58. Rediker, K.J. and Seth, A. (1995), “Boards of directors and substitution effects of alternative governance mechanisms”, Strategic Management Journal, Vol. 16, pp. 85-99.

59. Suchard, J., Manohar, S. and Barr, R. (2001) “The market effects of CEO turnover in Australian firms”, Pacific-Basin Finance Journal, Vol. 9, pp. 1–27.

60. Vafeas, N. (1999), “The nature of board nominating committees and their role in corporate governance”, Journal of Business, Finance and Accounting, Vol. 26, No. 1 & 2, pp. 199 – 225.

61. Vance, S.C. (1964), Board of directors: structure and

performance, Eugene, OR: University of Oregon Press. 62. Vance, S.C. (1978), “Assessing corporate performance

by boardroom attributes”, Journal of Business

Research, Vol.6, pp. 203-220. 63. Zald, M.N. (1967), “Urban differentiation,

characteristics of board of directors and organizational effectiveness”, The American Journal of Sociology, Vol. 73, pp. 261 – 272.


91

Table 1. Descriptive statistics regarding monitoring measures

(Sample Size: 1410 Company –years)

Combined Sample 2001 2002 2003

Monitoring Variables

Min

imum

Max

imum

Mea

n

S.D

Min

imum

Max

imum

Mea

n

S.D

Min

imum

Max

imum

Mea

n

S.D

.

Min

imum

Max

imum

Mea

n

S.D Size of the Boards (number) 3 17 6.31 2.11 3 17 6.24 2.153 3 17 6.31 2.132 3 15 6.38 2.07 Boards of Directors’ meetings (per year) 2 3727 10.78 4.33 2 33 10.7 4.284 2 32 10.8 4.217 2 37 10.80 4.52 Proportion of Independent directors on the Boards 0 1 0.71 0.195 0 1 0.71 0.203 0 1 0.7 0.202 0 1 0.71 0.18 Proportion of Financial Literate directors on the Boards 0 1 .4139 .2529 0 1 .4365 .2508 0 1 .4350 .2486 0 1 0.36 .25 Dual role of Chair and CEO (0,1) 0 1 0.12 0.320 0 1 0.14 0.344 0 1 0.14 0.343 0 1 0.08 0.27 Number of Audit Committee Meetings (per year, N = 1265) 0 15 3.03 2.02 0 12 2.83 1.85 0 14 3.04 2.024 0 15 3.24 2.16 Proportion of Independent members on Audit Committee (N=1265) 0 1 0.69 0.35 0 1 0.67 0.351 0 1 0.7 0.357 0 1 0.71 0.36 Proportion of Financially Literate directors on the AC (N=1265) 0 1 0.44 0.34 0 1 0.46 0.34 0 1 0.46 0.34 0 1 0.39 0.34 Number of Compensation Committee Meetings (per year, N = 815) 0 15 1.49 2.12 0 14 1.32 1.98 0 15 1.51 2.245 0 15 1.64 2.11 Proportion of Non-Executive Directors on RC (N = 815) 0 1 0.87 0.223 0 1 0.87 0.218 0 1 0.88 0.196 0 1 0.85 0.249 Number of Nomination Committee Meetings (per year, N =288) 0 17 0.55 1.57 0 17 0.45 1.626 0 13 0.56 1.674 0 13 0.66 1.41 Proportion of Non-Executive Directors on NC (N = 288) 0 1 0.89 0.208 0 1 0.91 0.208 0 1 0.90 0.203 0 1 0.88 0.213


92

Table 2. ROA 2001 (T+1)

Latent Variable Measurement Variable Standardized Regression Weights

Sig. Sq Multiple Correlation

Board BSIZ BDM

PBIND PBFL CECH

.749

.232

.277

.094

.269

.000

.000

.000

.080

.000

.561

.054

.077

.009

.073 Audit Committees ACM

PAI PAFL

.603

.388

.306

.000

.000

.000

.363

.150

.094 Compensation Committees

RCM PRNE

.428

.399 .000 .000

.183

.159 Nomination Committees

NCM PNNE

.252

.499 .000 .000

.064

.249 Control Variable SIZE .701 .000 .491 Performance ROA 2001 (T+1) .209 .000 .044

(CMIN/DF = 3.165, GFI = .942, AGFI = .902, CFI - .925, RMSEA = .039)* *Here, CMIN/DF = Normed Chi-Squire (Acceptable limit 1 – 5; 1 = best fit, 5 = reasonable fit) GFI = Goodness of fit index (acceptable limit => .90) AGFI = Adjusted goodness of fit index (acceptable limit => .90) CFI = Comparative fir index (0 = no fit at all, 1 = perfect fit) RMSEA = Root mean squire (.05 or less indicate a close fit) (Source: Hair, et al. 2006)

Table 3. ROA 2002 (T+1)



Board BSIZ BDM

PBIND PBFL CECH

.704

.139

.217

.046

.238

.000

.018

.000

.395

.000

.496

.019

.047

.002


PAI PAFL

.617

.396

.211

.000

.000

.000

.380

.157


RCM PRNE

.483

.414 .000 .000

.233


NCM PNNE

.395

.481 .000 .000

.156


(CMIN/DF =3.165, GFI = .942, AGFI = .902, CFI = .925, RMSEA = .039)

Table 4. ROA 2003 (T+1)



Board BSIZ BDM

PBIND PBFL CECH

.671

.254

.239

.201

.178

.000

.000

.000

.000

.000

.450

.064

.057

.040


PAI PAFL

.644

.448

.285

.000

.000

.000

.414

.201


RCM PRNE

.566

.500 .000 .000

.320


NCM PNNE

.476

.529 .000 .000

.226




93

Table 5. EPS 2001 (T+1)



Board BSIZ BDM

PBIND PBFL CECH

.743

.224

.278

.096

.267

.000

.000

.000

.074

.000

.552

.050

.077

.009


PAI PAFL

.606

.392

.300

.000

.000

.000

.368

.154


RCM PRNE

.433

.397 .000 .000

.187


NCM PNNE

.264

.509 .000 .000

.069

.259 Control Variable SIZE .702 .000 .492 Performance EPS 2001(T+1) .267 .000 .071


Table 6. EPS 2002 (T+1)



Board BSIZ BDM

PBIND PBFL CECH

.705

.128

.217

.053

.233

.000

.028

.000

.320

.000

.498

.016

.047

.003


PAI PAFL

.613

.394

.206

.000

.000

.000

.375

.155


RCM PRNE

.476

.410 .000 .000

.227


NCM PNNE

.391

.481 .000 .000

.153

.231 Control Variable SIZE .704 .000 .495 Performance EPS 2002 (T+1) .267 .000 .071


Table 7. EPS 2003 (T+1)



Board BSIZ BDM

PBIND PBFL CECH

.681

.237

.242

.201

.170

.000

.000

.000

.000

.001

.464

.056

.058

.040


PAI PAFL

.637

.436

.274

.000

.000

.000

.405

.190


RCM PRNE

.560

.487 .000 .000

.314


NCM PNNE

.471

.526 .000 .000

.222

.276 Control Variable SIZE .680 .000 .463 Performance EPS 2003 (T+1) .200 .000 .040



94

EXECUTIVE BOARD MEMBERS’ REMUNERATION: A LONGITUDINAL

STUDY

Themistokles Lazarides*, Evaggelos Drimpetas**, Koufopoulos Dimitrios***

Abstract Remuneration is considered to be closely connected with financial performance (positively), firm size (positively), the organizational structure (negatively) and corporate governance mechanisms (negatively). Furthermore, a connection of ownership structure and executives’ remuneration has been well established (theoretically and empirically) in the literature (agency theory). The paper examines if these relationships are valid in Greece. Greece hasn’t the characteristics of an Anglo-Saxon country. Overall the study has shown that remuneration levels in Greece are defined by a different set of factors than the ones that are prominent in an Aglo-Saxon country. Notably, fundamental financial measures of performance are more widely used. The age of firms and corporate governance quality have a catalytic impact on remuneration levels. Keywords: board members, remuneration, corporate governance *Department of Applied Informatics in Administration and Economy,, Technological Institute of West Macedonia, Grevena, Greece, [email protected], [email protected] Grevena, 51100, Greece, Tel.: +302462087691, Fax: +302462087692, [email protected], [email protected] **International Economic Relations and Development, Democritus University of Thrace, Komotini, Greece, [email protected] *** Brunel University, [email protected]

Introduction

Remuneration levels of executive directors and managers is the corner stone for the alignment of interests between executives and shareholders (Grossman, and Hart, 1982), the verification of the value of managers (Petra, 2005), managers’ and directors’ contribution to financial performance (Letza and Kirkbride et al., 2008; Conyon et al., 1995; Gregg et al., 1993; Hassan, Christopher, Evans, 2003) and the overall value of the firm (Habib and Ljungqvist, 2005; Stulz, 1990). Remuneration – compensation is the incentive for better managers’ performance and better financial performance (Jensen, 1986). Some researchers (i.e. Petra, 2005) argue that it is necessary to enforce managers in order to enforce productivity. Remuneration control is exerted by the Annual Shareholders Meeting, the Board of Directors or by any committee that has been introduced to control and evaluate executive managers and their performance. The efficiency of these mechanisms has been the focal point of many studies (Petra, 2005; Conyon and Peck, 1998).

Remuneration has been seen by the agency theorists as a (partial) remedy of the agency problem (Bebchuk and Fried, 2003). The Board of Directors (BoD) is responsible for providing a remuneration scheme that will direct executive managers to align their interests with the shareholders’ interests (Minow and Bingham, 1995; Muth and Donaldson, 1998). A characteristic of the BoD in Anglo-Saxon countries is

that directors typically have only nominal equity interests in the firm (Baker, Jensen, and Murphy, 1988; Core, Holthausen, and Larcker, 1999).

Some research findings (Bebchuk and Fried, 2004, p. 2) showed that managerial power has dominated the process of negotiation for remuneration levels. BoD is responsible for determining these levels and the schemes of remuneration. “In light of the historically weak link between non-equity compensation and managerial performance, shareholders and regulators wishing to make pay more sensitive to performance have increasingly looked to, and encouraged, equity-based compensation—that is, compensation based on the value of the company’s stock” (Bebchuk and Fried, 2004, p. 7).

The theory that executives and directors should be motivated to align their interests with the shareholders’ interests, has led to a quintuple (Cassidy, 2002) of executive remunerations in a decade (1991-2001) and the disclosure of frauds. On the other hand executives are willing to invest free cash flows ineffectively, to retain the capital assets within the firm, rather than to distribute them to shareholders (Hellwig, 1998). The basic motive for the executives is the dominance in corporate power game. Dominance guarantees high remuneration and entrenchment.

Agency theory addresses the issues that arise from organizational structure of firms that follow the Anglo-Saxon firm characteristics. There are major


95

differences, relevant to remuneration, between the Anglo-Saxon system and the one in the Continental Europe (Weimer and Pape, 1999) are: a) markets for corporate control, capital and labor market for directors are more active and effective (although there is a growing discussion about how efficient they are). Furthermore, executive managers may entrench themselves in their positions, making it difficult to oust them when they perform poorly (Shleifer and Vishny, 1989), b) As Shleifer and Vishny (1997) argue in Anglo-Saxon countries, capital providers need specialized human capital and executives need capital providers, because they do not have enough capital themselves. On the contrary, in Continental Europe countries executive directors are capital providers and in many cases, members of the dominant group of stakeholders and c) The presence of a large shareholder is likely to result in closer monitoring and reduce of executive directors’ power to impose the pursuit of their interests (Shleifer and Vishny, 1986).

In Continental Europe countries the fact that major shareholders are members of the BoD, CEOs and Presidents of the BoD, reduces the possibility of monitoring and transparency. These members have triple attributes or roles (major shareholder - part of the dominant group, member of the BoD and CEO – President of the BoD). Greece is a typical Continental Europe system’s country.

Corporate Governance status in Greece Greek firms are mainly family or controlled by a group of stockholders (Mavridis, 2002). Free float is relatively small in percentage (20-50%) and the ability to achieve control through the capital market is limited. The members of the family or the controlling group are actively involved in management and normally, there is no distinction between management and ownership. The Board of Directors can be characterized as one tier. Managers that are not members of the family or the controlling group are closely connected with these groups and their decisions are subject to their control and monitoring. Institutional investors, although the catalyst for the adoption of CG mechanisms, have not actively been involved in management or in controlling and monitoring the decisions and actions of the controlling group.

Greece’s legal framework constitutes a mixture of German and French law. According to La Porta, et al. (1998), countries with English Law (common law) tradition have the strongest legal protection for minority investors while French law provides the weakest protection. Countries with German law fall in the middle in terms of protection for shareholders. Anti-director rights measure how strongly the legal system favors minority shareholders against managers and dominant shareholders. La Porta, et al. (1998) believe that a strong legal enforcement system could substitute for weak rules since an effective judiciary

can step in and save minority shareholders from exploitation by the management. So, eventually, Greece has these characteristics as well.

Mertzanis (2001) (before the new law for the CG in Greece was enacted) noted: “the prevailing framework of corporate governance in Greece is not simply considerably outdated, but may cause potential problems, due to inadequate transparency and accountability, regarding the provision of cost-efficient finance that is required to increase investment and raise national competitiveness”. So the Hellenic Capital Market Committee (2000) and the Committee on Corporate Governance28 have made 44 basic recommendations (compiled in seven main categories: rights and obligations of shareholders; the equitable treatment of shareholders; the role of stakeholders in corporate governance; transparency, disclosure of information and auditing; the board of directors; the non-executive members of the board of directors; Executive management. They have also proposed the adoption of IAS (now IFRS)). Only a small number of these recommendations have been adopted and introduced.

Spanos (2005) notes that “the majority of medium and small capitalization (family-owned) companies have adopted the minimum mandatory requirements and lack further efficient CG mechanisms. As long as the competition for capital is increasing, listed companies have to realize that proper CG is a prerequisite in order to attract international capital. Moreover, corporate governance may meet one of the most significant challenges that family-run businesses face: management succession”. The need for CG mechanisms is identified by all market participants as a substitute for trust (as a bonding and problem solving element) among the major stockholders or family members, but they cannot agree on what the mechanisms/processes will be. Also, there are strong resistive forces mainly by the major stockholders/family members who are not willing to pass power and information to “non-trust worthy” stockholders or professional executive managers. As a result the governing/administrative bodies do not function according to statutes or laws and the process that they provide, but according to the common will of the family members. Furthermore, an effective market for corporate control does not exist.

Τhe board is mostly acting as a passive body in the company where it follows the decisions of the management. Non-executive board members, rather than act as shareholders’ agents, do not efficiently supervise the management (Schulze et al., 2003). This is the case in the majority of (family) public companies in Greece, where significant costs result from bias in favouring family interests over the firm’s interests (such as non-family shareholders), because of loyalty toward the family (Schulze et al., 2003). Even though the rules mandate specific requirements regarding board independence, it’s difficult in practice to identify whether the board meets these rules (Spanos (2005). In countries with concentrated


96

ownership structure (continental Europe, Japan and other OECD countries), large dominant shareholders usually control managers and expropriate minority shareholders, in order to extract private control benefits. The question is therefore posed as how to align the interests of strong block-holders and weak minority shareholders (Spanos, 2005, p. 16; Becht, 1997).

On the other hand, investors usually use their exit options if they disagree with the management or if they are disappointed by the company’s performance, signaling – through share price reduction – the necessity for managers to improve firm performance (Spanos, 2005, p. 16; Hirschman, 1970). The lack of market liquidity creates problems in the effectiveness of the shareholders exit option and governing problems (since the main governing body is the general shareholders meeting, but participation is not an easy task). The cost of involvement with management and control for the minor stockholder is greater than the cost of exit and so they may easily choose to sell their stock (“they vote with their feet”) if they are not content with the managements’ choices. The shareholders encirclement does not necessarily mean participation in the company administration. Moreover, family firms disclose less narrative information than non-family corporates, where family-firms may disclose more information than non-family corporates in some selected areas of interest, such as data information about share price policy and number of diagrams used in the interim report (Mavridis, 2002). In countries where business has traditionally been based on relationship and trust, corporate information is thought of as secret; and it is accepted practice to keep different sets of books, e.g. one for taxes, one for outside investors, and one for the majority shareholder (Fremond and Capaul, 2002, p. 18). There is a vicious circle whereby managers consider secrecy as imperative so that shareholders do not vote with their feet and through it they can cover up their lack of efficiency or impotence; minority shareholders (major shareholders already have the information because they are members of the BoD, management or the relevant cost for them is not too high) do not actively demand information because the cost of acquiring and processing it is to high for them.

The proposition of the study is that the agency theory is not valid in a Continental Europe’s system country. Remuneration is considered to be closely connected with financial performance (positively), firm size (positively), the organizational structure (negatively) and corporate governance mechanisms (negatively). Furthermore, a connection of ownership structure and executives’ remuneration has been well established (theoretically and empirically) in the literature (agency theory). The paper sets out the methodological approach adopted for this study by discussing the sample frame and the measurement issues of the various constructs of the study. Then the analysis of the data is presented through descriptive

and regression analysis. Finally conclusions are drawn.

Methods

Sample The study’s time horizon is from 2001 to 2006. Sixty firms, that are ranked in the two major stock indexes (FTSE-20 and FTSE-40) of the Greek Capital market and they are consider to be the biggest firms in terms of capitalization and with the highest free float, are used. Their annual reports are the basic source for the data collection. The data was supplemented by information collected by the corporate web sites. Total sample size is 303 observations. Although remuneration disclosure is mandatory, from the 303 available annual reports only 109 contain information about the executive board members. This is a strong indication of the trend to conceal “sensitive” information. As Bebchuk and Fried (2003) argue that executives have an incentive to “camouflage” their remunerations, in order to minimize the “outrage” of outsiders. In this case it’s the major shareholders that conceal information.

Measurments A panel data study is the most appropriate method to determine the factors that formulate the executive’s director remuneration level. There are a number of methodologies that can be used in a panel study. If the Fixed Effects (FE) models are proved to be statistical better, then firms have an identifiable and steady trend and do not differ one from the other, but only at the intercept level. On the contrary, if Random Effect (RE) models are proved to statistically better, then there is a more dynamic situation, where groups (stratum) and time affect firm’s behavior, and this behavior is statistically different among the firms of the sample. In RE models the estimators and the intercept are considered to be equal within the stratum and time. N advantage of the RE model is that if they are statistical better, then the hypothesis that the sample is representative to a greater population, has merit.

Finally, for the study of time effect the Two Way (TW) model is used. Hence the paper used three types of models. Four (4) variables are used to stratify the sample:

• Binary variable of law (LAW). The variable takes the value of 1 if the year is greater or equal to 2003 and the value 2 if it is smaller. It is used to detect the effect of the law on remuneration levels and financial performance

• Binary variable of Index (INDEX). The variable takes the value of 1 if the firm is ranked at the FTSE-20 index and the value 2 if it is ranked at the FTSE-40. It is used to detect the effect of firm’s size and ownership diffusion on remuneration levels.

• Binary variable of activity sector (FIN). The variable takes the value of 1 if the firm’s activity is financial and the value 2 if it is not. The third variable


97

is used to detected the effect of activity sector on the remuneration levels.

• Binary variables OWN_B (the sum of the percentages of equity capital of the five biggest shareholders) and HERF_B (the square of the sum of the percentages of equity capital of the five biggest shareholders). These variables have been calculated. If the value of the observation is smaller than the

median the variable takes the value 1 and if it is greater the value 2. This variable is used to detect the effect of ownership diffusion to remuneration.

Every stratifying variable slits the sample into two stratums. Finally, the variable YEAR is used to detect the effect of time. The independent variables that are used in the study are presented below.

Table 1. Variables

Variable Type Description

Panel (stratifying) variables Index Binary Participation in FTSE-20 (1) or in FTSE-40 (2) Law Binary Prior to 2003 (1), after 2003 (2) Financial Binary Financial firm (1), Non financial firm (2) Own_G Binary < median of Own (1), > median of Own (2) Herf_G Binary < median of Herf (1), > median of Herf (2) Year Number Year

Independent Variables Ownership variables

Own Percentage Sum of ownership percentages of the biggest five shareholders Herf Percentage Square of the sum of ownership percentages of the biggest five shareholders

Performance variables ROA Continuous Return on Assets TQ Continuous Tobin’s Q

Other independent variables CG Ordinal Quality of CG MERGER Binary M-A (1), no M-A (0) INVP Continuous Investments as a percentage of assets DE Continuous Debt Ratio (Debt / Equity)

Main independent variables OWNCEO Binary Main shareholder is the CEO (1), No (0) CEOCHAIR Binary CEO is the President of the Board of Directors – duality of roles (1), No (2) AUDITC Binary An Audit Committee exists (1), No (2) BOD Ordinal Number of members in the Board of Directors BEXEC Ordinal Number of executive Board members BPS Ordinal Number of firms that the Board members participate as Members of their

Board of Directors BDIS_P Ποσοστό Secessions – Resigns of board members to the total number of board

members BDISI_P Ποσοστό Secessions – Resigns of board independent members to the total number of

board members Control variables

TA Continuous Total assets SMCAP Continuous Stock market capitalization EMPL Continuous Number of employers OC_S Continuous Own Capital to Sales OC_S2 Continuous Square of Own Capital to Sales YEARF Continuous Foundation year

Model construction What has been recorded in the study is the cash – salary payments made to the executive directors. No other way of remuneration (e.g. stock options) could be tracked through annual reports. This may result to the omission of some of the remuneration

mechanisms. The omission of these mechanisms, although is important in the Anglo-Saxon countries, in countries like Greece these mechanisms are not widely used. In Greece the majority of executive directors are major shareholders. The model is:

E_REMit = α + β1ROAit + β2TQit + β3CGit + β4MERGERit + β5DEit + β6HERFit + β7OWNCEOit + β8ΒODit + β9BEXECit + β10BPSit + β11BDIS_Pit + β12BDISI_Pit + β13PROSPAGit + β14TAit + β15EMPLit + β16SMCAPit + β17OC_Sit + β18OC_S2it + β19YEARFit + uit (1) Where: i = 1 … Ν, t = 1 … T

Remuneration should be positively related with the variables of financial performance (ROA, TQ). Agency theory argues that, in order for the interests of executives and shareholders to align, executives

should be paid according to their performance. Hence, the sign should positive.

On the contrary the variables of CG quality index (for the construction of the index see Lazarides and


98

Drimpetas, 2008) and the binary value of mergers and acquisitions are negatively related with remuneration levels. Monitoring, control and market for corporate control reduce the possibility of managers to impose the remuneration level that they wish. BoD’s size BOD, number of members) is a measure for the determination of the monitoring efficiency. On the contrary, high numbers of executive members (BEXEC) of the BoD and secessions – resigns of board members (BDIS_P), to the total number of board members may lead to higher remuneration levels.

Firms depended heavily on debt to finance their operations, present high uncertainty and risk for future returns. This, in turn should lead to reduced remunerations. Firm size (EMPL, TA) is positively related with the remuneration levels. Bigger firms have the capacity to pay more

Concentrated ownership (variables OWN and HERF) is a remedy for the reliance on the skills and knowledge that the professional executives possess. The lack of reliance minimizes the bargaining power of managers and therefore the relation is negative. The same conclusions can me drawn for the variance OWNCEO (the biggest shareholder is the CEO).

Investments (INVP) are positively related with remuneration levels, if the assumption that investment can lead to better short-term financial results. Otherwise, the relation is negative. Younger firms tend to present higher risks and hence firms are reluctant to have relatively high remuneration levels.

Firm age (YEARF) is a critical component in remuneration. As the firm grows old it loses the initial family characteristics due to diffusion of shares

(through IPO’s, capital increases and succession) and so as firm’s grow older there is a higher reliance on professional manages, rather than dominant stakeholders. Firms that were founded in the last twenty five years have 61% ownership concentration percentage (OWN), when in older firms ownership concentration is 53%. Younger firms present higher risk factor and the firm is reluctant to have large remuneration levels. Younger firms (<25 years) pay on average half (0,66 million) the amount the older firms pay (1,24 million). Executive members in Greek firms are also major shareholders and their tenure is long. Long tenures counterbalance the expectation for higher remuneration.

Findings Statistical results Remuneration descriptive statistics are shown in Table 1. The sample was divided in two main categories. The first one is the observations of the firms that are ranked in the FTSE-20 index (the biggest 20 firms in terms of capitalization) of the Athens Stock Exchange and the second is the observations of the firms that are ranked in the FTSE-40 index (the next 40 firms in terms of capitalization).

Two ratios were calculated to relate remuneration with fundamental firm size variables: E_REM_EQ is the ratio of remuneration and equity and E_REM_SM is the ratio of remuneration and stock market value. Table 1 shows that remuneration levels are different from one index to the other (FTSE-20 mean is 2,04, instead the mean for FTSE-40 is 0,68). FTSE-40 remuneration level presents smaller standard deviation from the FTSE-20 observations.

Table 2. Executives Directors Remuneration by Index (2001-2006)

Mean Standard Deviation Min Max

FTSE – 20: Observations 32 E_REM 2.04078 1.92586 .257000 8.15700 E_REM_EQ .198524E-01 .206993E-01 .588672E-03 .724860E-01 E_REM_SM .133671E-02 .130430E-02 .158162E-03 .696350E-02

FTSE – 40: Observations 77 E_REM .680747 .478294 .830000E-01 2.33200 E_REM_EQ .261985E-01 .218827E-01 .146686E-02 .946160E-01 E_REM_SM .890263E-02 .269559E-01 .000000 .177612

Total Sample: Observations 109 E_REM 1.08002 1.26994 .830000E-01 8.15700 E_REM_EQ 243354E-01 .216422E-01 .588672E-03 .946160E-01 E_REM_SM .668144E-02 .228866E-01 .000000 .177612

Remuneration variance through time presents a

peculiar behavior (see Table 2). While in 2002 remuneration levels are reduced in absolute terms, this is not the case for the remuneration ratios. This may be caused by the lack of relation between remuneration and share price premiums. In 2003-2004

stock market prices fell, while the remuneration levels increased. When stock market prices increased (2004-2006), remunerations decreased. Many firms have adopted complex incentive schemes with the use of stock options.


99

Table 3. Executives Directors Remuneration by Year (2001-2006) Mean Standard Deviation Min Max

2001: Observations 20 E_REM .910525 1.10713 .830000E-01 5.02100 E_REM_EQ .238433E-01 .160083E-01 .128053E-02 .613824E-01 E_REM_SM .394517E-02 .810327E-02 .158162E-03 .368509E-01

2002: Observations 21 E_REM .896310 1.11897 .127000 5.40200 E_REM_EQ .211610E-01 .154454E-01 .209451E-02 .514464E-01 E_REM_SM .707724E-02 .163597E-01 .312346E-03 .769415E-01

2003: Observations 20 E_REM 1.27491 1.66174 .145000 7.71000 E_REM_EQ .271230E-01 .220311E-01 .159222E-02 .885706E-01 E_REM_SM .980991E-02 .317564E-01 .411541E-03 .144361

2004: Observations 24 E_REM 1.31988 1.61353 .177000 8.15700 E_REM_EQ .259800E-01 .262013E-01 .972797E-03 .946160E-01 E_REM_SM .109078E-01 .357221E-01 .290133E-03 .177612

2005: Observations 13 E_REM 1.06934 .675756 .190000 2.47600 E_REM_EQ .235146E-01 .223470E-01 .104863E-02 .577701E-01 E_REM_SM .250801E-02 .266340E-02 .000000 .781936E-02

2006: Observations 11 E_REM .873900 .642492 .190000 2.40900 E_REM_EQ .236038E-01 .308050E-01 .588672E-03 .729348E-01 E_REM_SM .923911E-03 .744573E-03 .000000 .221930E-02

Graph 1. Executives Directors Remuneration by Year (2001-2006)

Disclosure levels are higher in the Non Financial sector (see Table 3). Non Financial firms seem to disclose more information than the financial firms.

Table 4. Disclose frequency of remuneration in relation with the activity sector

Non Financial Financial Total Disclosed remuneration 96 (37,1%) 13 (29,5%) 109 (36%) Non Disclosed remuneration 163 (62,9%) 31 (70,5%) 194 (64%) Total 259 (100%) 44 (100%) 303 (100%)

As Table 4 depicts firms with higher ownership

concentration, better corporate governance level (Lazarides and Drimpetas, 2008) and better Tobin’s Q, seem to disclose more information.

Table 5. Disclose frequency of remuneration in relation with other variables (2001-2006)

Disclose 2001 2002 2003 2004 2005 2006 Total

Ownership concentration (OWN) – Mean

No 0,53 0,53 0,53 0,48 0,48 0,47 0,50

Yes 0,59 0,60 0,54 0,55 0,61 0,52 0,57

Total 0,55 0,56 0,53 0,51 0,51 0,48 0,52


100

Table 5 continued Corporate Governance Index (CGC) – Mean

No 2,16 2,11 2,28 2,93 3,00 3,03 2,64

Yes 3,35 3,76 4,40 4,13 3,92 4,42 3,97

Total 2,69 2,83 3,10 3,47 3,22 3,35 3,13

Return on Assets (ROA) – Mean

No 0,08 0,06 2,87 2,32 0,07 0,09 0,88

Yes 0,10 0,07 0,08 0,08 0,07 0,10 0,08

Total 0,08 0,06 1,80 1,31 0,07 0,10 0,59

Tobin’s Q (TQ) – Mean

No 2,39 1,23 1,26 1,22 1,69 1,82 1,60

Yes 1,87 1,52 3,62 1,54 1,52 1,98 2,02

Total 2,16 1,36 2,17 1,37 1,65 1,86 1,75

The LM statistics, in all cases, is small (see Table

6). This indicates that the classical OLS model may be better suited for the analysis of the constructed models (Greene, 2000). The models with stratifying variables seem to have little significant statistical importance. The H statistic indicates that Random Effects models are preferable to Fixed Effects. Due to the fact that the stratifying variables LAW, OWN_G present a higher LM statistic, a decision to further analyze these models using a smaller set of independent variables (the ones with statistical significance). The coefficient of determination (R2)

and the Adjusted R2 (see Table 7) is in all case satisfactory (> 0,34).

The use of White’s method to correct the heteroscedasticity problem (statistic LM X2 for the B-P-G has the value 510,91 (14 degrees of freedom) (possibility of homoscedasticity: 0,00). Heteroscedasticity is common in panel studies. There is no autocorrelation problem (d=1,9795 and r=0,0102).

Table 8 suggests that from seventeen independent variables, seven are statistical significant. Three variables have the opposite than expected sign.

Table 5. Regression model selection statistics

Stratifying Variable - Time LM H INDEX 0,73 0,00 INDEX - YEAR 0,73 0,00 FIN 0,10 0,00 FIN - YEAR 0,11 0,00 LAW 0,99 0,04 LAW - YEAR 0,99 0,00 OWN_G 0,91 0,00 OWN_G - YEAR 0,91 0,00 HERF_G 0,29 0,00 HERF_G - YEAR 0,29 0,00

Table 6. Coefficient of determination

Stratifying Variable - Time R2 R2 Adj. None 0,445 0,3424 INDEX - YEAR 0,7722 0,7106 FIN - YEAR 0,842 0,7992 LAW - YEAR 0,771 0,709 OWN_G-YEAR 0,7724 0,7108 HERF_G-YEAR 0,7826 0,7237

Table 7. Statistical significance test (Stratification: None)

Variable β Standard Error β/ Standard Error Statistical significance Theoretical Confirmation ROA -311.91 184.93 -1.687 .0951 *** No TQ 12.740 5.0218 2.537 .0129 ** Yes CG 7.9785 5.8004 1.376 .1723 No MERGER -12.117 14.870 -.815 .4172 Yes HERF -101.82 143.96 -.707 .4812 Yes DE -2.3035 1.3594 -1.695 .0935 *** Yes OWNCEO -13.684 28.722 -.476 .6349 Yes BOD -6.6545 5.7545 -1.156 .2505 Yes BEXEC 7.9342 6.3132 1.257 .2120 Yes BPS 1.6664 2.2070 .755 .4521 Yes


101

Table 8 continued BDISI_P -745.35 367.98 -2.026 .0457 ** Yes BDIS_P 43.473 47.775 .910 .3652 Yes INVP 99.892 58.070 1.720 .0888 *** Yes YEARF .3737 .4382 .853 .3960 No OC_S2 -7.6742 3.5992 -2.132 .0357 ** Yes TA .0038 .0020 1.823 .0716 *** Yes EMPL -.0003 .0038 -.099 .9215 Yes

* p< 0,01 ** p< 0,05 *** p< 0,10

Final model Model (1) has been further analyzed to clear all non statistical significant variables. OLS did not produce any significant results. Using panel data methods and the stratifying variables of LAW-

ΥΕΑR, OWN_G-YEAR the models were regressed. The model with LAW as a stratifying variable has given good statistical results.

E_REMit = α + β1ROAit + β2CGit + β3TAit + β4YEARFit + uit (2) Where: i = 1 … Ν, t = 1 … T

Even though the LM statistic for Model (2) was bigger than 3,8 the possibility to opt FE / RE models

is more than 90%. The H statistic indicates that RE models are better suited for model (2).

Table 8. Statistical significance test (Stratification: None) Stratifying Variable - Time LM H

All independent variables LAW 0,99 0,04 LAW - YEAR 0,99 0,00

Smaller set of variables LAW 3,82 0,00 LAW - YEAR 4,77 0,00

The test for the determination of the statistical

difference of the two stratums has been conducted with the X2 and F statistic (see Table 10). The F statistic test shows that the model with the combined effect of the stratifying variable and time does not differ from the OLS model

( =3,98<6,861). The F statistic test for the

model with the use of LAW as stratifying variable illustrates that in this model the two stratums statistically differ on from the other

( =2,15>1,172). The methodology that is selected is RE model with LAW as stratifying variable.

Table 9. Determination Test for between the two stratums

Stratifying Variable - Time X2 Prob. F P - value LAW 7,029 0,00802 6,861 0,01014 LAW - YEAR 8,765 0,26996 1.172 0,3257

Coefficient of determination has been marginally

reduced from 0,77 to the 0,7293 (Adj. R2 = 0,70168). Four variables have been proven to be statistical

significant. All of them have the sign that is theoretically correct.

Table 10. Statistical significance test (Stratification: LAW)

Variable β Standard Error β/ Standard Error Statistical significance Theoretical Confirmation ROA 2.03865 .60838 3.351 .0008* Yes CG -.10251 .04268 -2.401 .0163** Yes YEARF -.00822 .00409 -2.008 .0446** Yes TA .000088 .000013 6.706 .0000* Yes

* p< 0,01 ** p< 0,05 *** p< 0,10

Conclusions One major finding of this study is that only 36% (109/303), breaking the law because disclosure is

mandatory, of the firms have disclosed in their annual reports the remuneration levels of their executive members. The selection of non disclosure is conscious. Major shareholders, groups of shareholders


102

and families are the dominant stakeholders in the firm. They are unwilling to release information that may shake the status quo or question their power to make decisions.

The use of RE methodology denotes that every firm has its own policies about remuneration of executive members and that the sample is representative of a larger population. Stratifying variables, except the variable of LAW, have no statistical significant results. Its striking that the coefficient of determination has doubled when the model was better specified. Law seems to have an impact on remuneration. The fact that the LAW binary variable is based on the time of enactment, this variable measures the effect of time hence the overall market’s influence on remuneration.

The relation between financial performance and remuneration as literature suggests is confirmed in this study. The study used three variables. Two of them are based on accounting measures (ROA, TA) and the third is Tobin’s Q, which is based mainly on stock market value. Tobin’s Q is the only variable that in the final model is not statistical significant. This indicates that there is no connection of share premiums and remuneration. This is contrary to what in the Anglo-Saxon countries is the norm and the remuneration is based mainly on share premiums.

The negative sign of variable YEARF confirms the descriptive statistics and the theory that younger, with more concentrated ownership and quite possibly family firms have the trend to pay less. CG quality index has a negative effect on remuneration. This fact is consistent with the notion that better monitoring and control minimizes the ability of executives to dictate remuneration levels.

Overall the study has proven that remuneration levels in Greece are defined by a different set of factors than the ones in an Anglo-Saxon country. Fundamental financial measures of performance are more widely used. The age of firms and corporate governance quality have a catalytic impact on remuneration levels.

References 1. Baker, G.P., Jensen, M.C. and Murphy, K.J. (1988)

“Compensation and Incentives: Practice vs. Theory”, Journal of Finance, Vol. 63:No. 3, pp. 593-616.

2. Bebchuk, L. and Fried, J. (2003), “Executive Compensation as an Agency Problem”, Discussion

Paper Series No 3961, CEPR, available on line at http://www.ecgi.org/remuneration/documents/bebchuk_fried_03.pdf, (last access 18/12/2008).

3. Bebchuk, L. and Fried, J. (2004) “Pay without

Performance”, Harvard University Press, MA. 4. Becht, M. (1997), “Strong Stockholders, Weak Owners

and the Need for European Mandatory Disclosure, Executive Report”, European Corporate Governance

Network, London, October. 5. Cassidy, J. (2002), “THE GREED CYCLE (Brief

Article)”, The New Yorker, 09/23/2002 6. Conyon, M., Gregg, P. and Machin, S. (1995),

‘‘Taking care of business: executive compensation in

the United Kingdom’’, The Economic Journal, Vol. 105, pp. 704-14.

7. Conyon, M.J., Peck, S.I. (1998), "Board control, remuneration committee, and top management compensation", Academy of Management Journal, Vol. 41 No.2, pp.146-57.

8. Core, J.E., Holthausen, R.W., and Larcker, D.E. (1999), “Corporate Governance, Chief Executive Compensation, and Firm Performance.” Journal of

Financial Economics, Vol. 51, pp. 371-406. 9. Fremond, O. and Capaul, M. (2002), The State of

Corporate Governance: Experience from Country

Assessments, World Bank Policy Research Working Paper 2858, June.

10. Gregg, P., Machin, S. and Syzmanski, S. (1993), ‘‘The disappearing relationship between directors’ pay and corporate governance’’, British Journal of Industrial

Relations, Vol. 31, pp. 1-10. 11. Grossman, S. J., and Hart, O. D. (1982), “Corporate

Financial Structure and Managerial Incentives”, in: The

Economics of Information and Uncertainty, McCall, J.J. (ed.), University of Chicago Press.

12. Habib, M. A., and Ljungqvist, A. P. (2005), “Firm Value and Managerial Incentives”, Journal of

Business, Vol. 78, no. 6, pp. 2053-2093. 13. Hassan, S., Christopher, T., Evans, R. (2003),

"Directors' remuneration and firm performance: Malaysian evidence", Malaysian Accounting Review, Vol. 2 No.1, pp.57-67.

14. Hellenic Capital Market Commission (2000), Annual Report, Hellenic Capital Market Commission, Athens.

15. Hellwig, M. (1998), “On the Economics and Politics of Corporate Finance and Corporate Control”, Proceedings of Conference on Corporate Governance

of the Institute Catala de Finances Sitges, October 23-24.

16. Hirschman, A. (1970), “Exit, Voice, and Loyalty:

Responses to Decline in Firms, Organizations, and

States”, Harvard University Press, Cambridge, MA. 17. Jensen, M.C. and Murphy, K.J. (2004), Remuneration:

Where We've Been, How We Got to Here, What are the Problems, and How to Fix Them” Working Paper

no 44, available on line at http://papers.ssrn.com/sol3/papers.cfm ?abstract_id=561305 (last access 20/12/2008)

18. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., (1998), “Law and finance”, Journal of

Political Economy, Vol. 106, pp. 1113– 1155 19. Lazarides, Th.G. and Drimpetas, E. (2008),

“Evaluating Corporate Governance and Identifying Its Formulating Factors: The Case of Greece”, Working

Paper Series (December 18, 2008). Available at SSRN: http://ssrn.com/abstract=1317629

20. Letza, S. and Kirkbride, J., Sun, X., Smallman, C. (2008), “Corporate governance theorising: limits, critics and alternatives”, International Journal of Law

and Management, Vol. 50, No. 1, pp. 17-32. 21. Mavridis G. Dimitrios (2002), “Marital Status Aspects

Disclosed in Greek Interims Reports””, Equal

Opportunities International, Vol. 21, No 8, pp. 1-17. 22. Mertzanis V. H. (2001), “Principles of Corporate

Governance in Greece”, Corporate Governance, Vol. 9, No 2, pp. 89-100.

23. Minow, N., and Bingham, K., (1995) "The Ideal

Board", in Monks R., and Moniw, N., Corporate Governance, Blackwell Publishers, Cambridge, Massachusetts.


103

24. Muth, M.M. and Donaldson, L. (1998), “Stewardship Theory and Board Structure: a contingency approach”, Corporate Governance: An International Review, Vol. 6, No. 1, pp. 5-28.

25. Petra S. T. (2005), “Do outside independent directors strengthen corporate boards?”, Corporate Governance:

the International Journal of Business in Society, Vol. 5, No 1: 55-64.

26. Schulze, W.S., Lubatkin, M.H. and Dino, R.N. (2003), "Toward a theory of agency and altruism in family firms". Journal of Business Venturing, Vol. 18 No. 4, pp. 473-90.

27. Shleifer, A. and Vishny, R. (1989) “Management Entrenchment: The Case of Manager-specific Investments.” Journal of Financial Economics, Vol. 25, pp. 123-140.

28. Shleifer, A. and Vishny, R. (1997), “A survey of corporate governance”, Journal of Finance, Vol. 52, pp. 737- 783.

29. Shleifer, A., and Vishny, R., (1986), "Large Shareholders and Corporate Control." Journal of

Political Economy, Vol. 94, No 3, pp. 461-88. 30. Shleifer, A., and Vishny, R., (1989), “Management

Entrenchment: The Case of Manager-specific Investments.” Journal of Financial Economics, Vol. 25, pp. 123-140.

31. Spanos, L. J. (2005), “Corporate Governance in Greece: developments and policy implications”, Corporate Governance: the International Journal of

Business in Society, Vol. 5, No. 1, pp. 15-30. 32. Stulz, R., (1990), “Managerial Discretion and Optimal

Financing Policies”, Journal of Financial Economics, Vol 26, pp. 3-27.

33. Weimer, J. and Pape, J. (1999), “A Taxonomy of Systems of Corporate Governance”, Corporate

Governance: An International Review, Vol. 7, No. 2, pp. 152-166.


104

BAD DEBTS, OWNERSHIP CONCENTRATION, AND BOARD

COMPOSITION: EVIDENCE ON THE QUALITY OF CORPORATE

GOVERNANCE OUTCOMES IN CHINA’S LISTED NON FINANCIAL

COMPANIES

Yuan George Shan*, Ron P. McIver**

Abstract

This study analyses the relationships between performance metrics and the corporate control and governance characteristics of a sample of China’s listed non financial companies in order to assess the influence of corporate governance structures on the quality and independence of corporate decision making. We use a panel data set covering the years 2001 to 2005 comprised of a stratified sample of A, AB and AH non financial companies listed on China’s Shanghai and Shenzhen stock exchanges. We find that concentration of ownership, including state and foreign ownership, and board size and independence are significant factors in determining performance outcomes, and by association the quality and independence of corporate policy decisions, as measured in the form of firm bad debt to total asset and bad debt to receivables ratios. Our findings support claims of continued inadequacies in the operation and effectiveness of China’s institutions of corporate governance, especially with respect to the effectiveness of the supervisory board.

Keywords: Corporate Governance; Corporate Control; China

*School of Commerce, University of South Australia **School of Commerce, University of South Australia

1. Introduction Establishment of the sets of institutions dealing with macroeconomic stabilisation, social insurance, conflict management, property rights, and the regulatory environment for business is often given priority in transition economies (Rodrik, 2000). The priority given to these institutional structures reflects a concern to create an environment conducive to sustainable long-term growth under a post-transition market-oriented economic system. Institutions covering property rights and the legal and regulatory environment are particularly important in transition economies as these are enabling factors in the reform of state-owned enterprises (SOEs). Sound institutional choices assist in ensuring the separation of government from the management of SOEs and in the successful restructuring of SOEs in corporate form (Beim and Calomiris, 2001; Zhu, 1999; OECD, 2000; Jevons Lee, 2001). At issue are the forms taken by the corporate control and governance structures of these enterprises.

For China the core of the framework for corporate regulation consists of The Company Law (proclaimed December 1993), The Securities Law (proclaimed December 1998), and The Code of

Corporate Governance for Listed Companies in

China (The Code) (issued January 2002 by the Chinese Securities Regulatory Committee (CSRC) and the State Economic and Trade Commission (SETC)) (see endnote 1). Given the importance of this set of institutions it is not surprising that, as well as having established the core of the framework and also its establishment of the CSRC in 1992, China has been active in evolving its regulatory environment. Since 1992, over 300 laws and directives have been issued that relate to the securities and futures markets (Lin, 2004).

Market-based governance is considered a priority due to its positive association with productivity improvement and through this growth in real output. This reflects the impact of governance on the efficiency with which individual firms utilise resources internally (Tadesse, 2004). On the other hand, inappropriate institutional choices may allow former SOEs, privatised and listed, to engage in corporate governance practices associated with misuse or misappropriation of state and corporate assets. For China, the latter outcome is an apparent problem. Questions exist as to the quality and effectiveness of its institutional choices and the outcomes that they generate. Following the establishment of the Shanghai and Shenzhen stock exchanges in 1990 and 1992, respectively, agency


105

problems have emerged in relation to ownership structure and corporate governance in China. These include evidence of the ineffectiveness of supervisory boards of listed companies in China (Dahya et al., 2003; Lin, 2004), and also of the weak independence of boards of directors and impact on minority shareholders of high levels of ownership concentration (Lin, 2004).

The objective in this paper is to develop and test a set of hypotheses related to the impact of the corporate control and governance characteristics of listed, non financial Chinese companies on the quality and independence of corporate policy decision making. The first set of characteristics is the level of concentration in and type of ownership. The second set of characteristics relate to the composition and independence of the Board of Directors and Supervisory Board under China’s two-tier board structure. Of interest are the impact in these companies of high levels of ownership concentration, particularly government and foreign ownership, and expertise and independence of the boards. The quality and independence of corporate policy decisions is measured in terms of the sampled firms’ bad debt to total asset and bad debt to accounts receivable ratios. This study addresses a gap in the literature in relation to the impact on corporate governance outcomes in China of both high concentrations in ownership and board size and composition under China’s two-tier board system.

The structure of the paper is as follows. Section 2 provides a brief review of literature on the nature of institutions and on the institutions central to corporate governance in China, and also develops the set of hypotheses to be tested. Section 3 outlines the research methodology and describes the data. Section 4 presents the results of the analysis and a discussion of these results. Section 5 is comprised of a brief conclusion.

2. Literature and Hypotheses The study of governance is a study of incentives and contractual relations or, as per Williamson (2000), a study of the ‘play of the game’. Under consideration in the study of governance are the impact of such matters as board structure and compensation, accounting standards, and transparency and disclosure in corporate reporting. This places the study of governance as a sub-set within the broader field of study of the economic, political and social institutions that impact economic decisions and outcomes (see

endnote 2). These forms of institution impact on resource allocation and employment outcomes, the focus for neoclassical economics and agency theory. Thus the establishment of institutions appropriate to the efficient operation of markets is a principal determinant in the success of reforms in transition economies. The importance of institutional structures and choices is that they establish the boundaries

within which economic agents operate as they attempt to maximise the values of their utility or welfare functions. Institutions therefore define the set of economic decisions that are possible, and thus the outcomes that may be achieved by a particular economic system (North, 1990).

The difficulties for analysis implied by the suggested scope of the definition of institutions may be overcome through a focus on what institutions do. That is, what institutions achieve, more so than what is stated or intended, is the matter of interest. Institutions, including those of governance, may be seen as the structures that determine the outcomes available from interactions between multiple goal-oriented decision makers (Nelson, 2007). From this perspective institutions represent the set of rules that society has established to impose limitations on free behaviour by these decision makers (Redek and Sušjan, 2005). Thus, a review of particular economic outcomes, including outcomes resulting from corporate policy decisions, allows inferences to be drawn regarding the quality or success of specific institutional arrangements in limiting adverse behaviours (or, alternatively, in promoting beneficial behaviours).

In the case of China, the core of the institutional framework for corporate governance (gongsi zhili) is comprised of the Company Law of the People’s

Republic of China (The Company Law) (proclaimed December 1993, revised in 2005), the Securities Law

of the People’s Republic of China (The Securities

Law) (proclaimed December 1998, revised in 2005), and the Code of Corporate Governance for Listed

Companies in China (The Code) (issued January 2002 by the CSRC and the State Economic and Trade Commission (SETC), and also revised in 2005) (see

endnote 3). The Code is the primary government document specifically dealing with corporate governance. Its intended role is to provide a set of guidelines that allow companies and their investors to conduct a self-evaluation of whether (or not) good corporate governance is in place; a “measuring standard” specifying good practice (see endnote 4). Thus, for example, it requires all listed companies in China to “act in the spirit of The Code in their efforts to improve corporate governance” (The Code, 2005), rather than providing a legally enforceable piece of legislation or regulation.

Consistent with the Company Law, The Code outlines a number of requirements in relation to the corporate control and governance characteristics of listed Chinese companies. Chapter 2 of The Code, in particular, deals with the responsibilities of controlling shareholders with respect to the company and other shareholders. Its intent is that controlling shareholders act in the interest of both the company and minority shareholders, and be prevented from advantaging themselves at the cost of these other parties. Chapters 3 and 4 of The Code deal with the matters related to the board of directors and the supervisory board present under China’s two-tier


106

board structure. Chapter 3 covers the duties, responsibilities, composition and independence of the board of directors. Chapter 4 outlines requirements for the supervisory board, including roles, reporting responsibilities, and requisite skill set. Of primary importance are guidelines regarding the quality of expertise of members of each of the boards, the independence between the board of directors and the supervisory board, and the ability of the supervisory board to monitor the performance of the company and board of directors.

It is apparent that China has put in place a set of institutional structures that evidence a desire to create (or provide) an institutional environment conducive to good corporate governance. However, it is with respect to this environment that transition economies such as China’s frequently display weaknesses. These weaknesses come in the form of deficiencies in institutional capacity, and thus the potential for institutional failure. Institutional failure leads these institutions to either undertake or discharge the functions for which they are designed inappropriately (Šević, 2005). Poor clarification of ownership and control rights preclude the enforceability of contractual obligations. Poorly developed accounting standards and a lack of transparency and disclosure in corporate reporting reduce the effectiveness of corporate governance mechanisms in aligning incentives of managers with those of enterprise owners. Poor accounting and disclosure practices will allow recognition of problems in financial performance to be deferred or hidden (Šević, 2005). Such systemic deficiencies are likely to be particularly problematic for transition economies, as they act to prevent the “low-cost transacting and credible commitment” required to support the creation of efficient markets (as per North, 1997).

In the case of China, a number of potential deficiencies in institutional capacity may be identified. These relate to corporate ownership and control, the independence of the board of directors, and to performance of the supervisory board. Each may be recognised as relating to an aspect of the agency problem. In this case institutional failure implies that the agency problem, as it relates to corporate governance, has not been adequately addressed at these multiple levels (see endnote 5).

2.1. Ownership, Majority Shareholders and Corporate Control The typical listed company in China has five classes of ownership: state shares, legal person shares, employee shares, domestic minority shares, and foreign institutional and/or foreign investor shares (Sun and Tong, 2003; Wei et al., 2005; Firth et al., 2007; Wei, 2007). When considering the impact within the corporate governance context of these different classes of ownership concentration and composition are key aspects. The degree of ownership

concentration determines the distribution of power. Contrary to the widely accepted views of Berle and Means (1932), large corporations face problems in the separation of ownership and control because they are managed by controlling shareholders and not by their professional managers (La Porta et al., 1999). Concentrated ownership of companies may reduce managers’ freedom to take risks, make strategic decisions and take advantage of opportunities. High levels of concentration in ownership are expected to affect management incentives and corporate policy choices through the pressure that these investors can exert on managers (Brickley et al., 1988; Pound 1988; Bushee 1998). Thus while a group of shareholders with a large total share of the equity might be more effective at monitoring management, their powers must be restrained to prevent them taking advantage of other shareholders (Clarke, 1998). High ownership concentration provides both incentive and opportunity for controlling shareholders and managers to engage in expropriation (Morck et al., 1988; Shleifer and Vishny, 1997; La Porta et al., 1999).

In China, majority shareholders are typically very strong and individual minority shareholders are relatively weak. In many cases minority shareholders are regarded as speculators with an expectation of gaining a “free ride” based on the performance of the firm (Lin, 2004). Thus minority shareholders in China are unable to counter the influence of majority shareholders. Contrary to The Code, related-party transactions between controlling shareholders may be detrimental to minority shareholders, and controlling shareholders may act so as to advantage themselves at the cost of minority shareholders. This suggests that China’s corporate governance is potentially relatively ineffective in the matter of protecting minority shareholders’ rights, and the first of our hypotheses: H1: High levels of concentration in

ownership in listed Chinese firms will

be associated with poorer corporate

policy decisions and performance.

In addition to concentration in ownership, government ownership is a feature of the ownership structure of many listed companies in China. This reflects their history as state-owned enterprises prior to being listed (Xiang, 1998). Such state ownership has been associated with a negative impact on firm performance (Wei et al., 2005; Gunasekarage et al., 2007) (see endnote 6). In this case agency problems may arise due to the differences in objectives between state and non state shareholders. For example, in the case of the state (central or local), maintenance of employment may take preference over profitability. Thus our second hypothesis is that: H2: High levels of state ownership in listed

Chinese firms will be associated with poorer

corporate policy decisions and performance.

With respect to corporate policy decisions and performance our third hypothesis deals with the impact of high levels of foreign ownership in these listed Chinese companies. This is:


107

H3: High levels of foreign ownership in listed

Chinese firms will be associated with improved


That a high level of foreign ownership may be a factor in improving corporate governance and, more broadly, accounting standards, reflects that international investors have an incentive to push for improvements in these areas. This is in order to better guarantee their interests (Šević, 2005; Krzywda et al., 1995). Thus we conjecture that this pressure will, on average, have a positive effect on those firms in which foreign investors have significant control.

2.2. Board Independence and Expertise As noted above, China has adopted a two-tier board system as a means to promote better governance. This choice was made in the early 1990s partly because many enterprises and their directors were perceived to be engaging in questionable related-party transactions. The Code gives particular attention to two aspects of these boards. The first is the independence of directors on the board of directors. The second is the qualifications and knowledge of members on the supervisory board.

Since 2003 at least one-third of directors on the board of directors have been required to be independent. Independence is required from both the listed company that employs them and its major shareholders. It also requires that their role in the listed company is limited to that of independent director. Independence may be argued to be important due to its behavioural motivations. From this perspective, independent directors work in the best interests of the minority shareholders in order to maintain their good reputation in society (Fama and Jensen, 1983) (see endnote 7). This suggests that both larger boards and those with a higher proportion of independent directors will have more individuals with these incentives, improving corporate policy decisions. Thus our fourth set of hypotheses is that: H4A: An increase in the number of board members

that are independent in listed Chinese firms

will be associated with improved corporate

policy decisions and performance.

H4B: An increase in the size of the board in listed

Chinese firms will be associated with improved


Under China’s two-tier board structure the supervisory board has the responsibility of and duty for oversight of both the board of directors’ and senior managements’ performance. They are also required to protect the company’s and stakeholders’ rights and interests. Under The Company Law they have the power to investigate their company’s operating status without interference and to report directly to the CSRC and other related regulatory authorities. Xiao et

al. (2004) argue that several key characteristics of the supervisors determine the role of the supervisory board. These include the proportion of supervisors who are insiders and shareholders, and the

professional knowledge or work experience of the supervisors. With respect to these final characteristics, the professional knowledge and experience of supervisors would be prerequisite to an ability to identify issues related to financial and managerial performance. Dahya et al. (2003) also highlight the importance of the capacity of the supervisory board to its ability to fulfil its stated functions. In doing so they identify four types of behavioural roles that supervisory boards can engage in, depending on the independence and capacity of the board members (see

endnote 8). These roles are: honoured guest, friendly advisor, censored watchdog, and independent watchdog. If the supervisory board takes on the role of honoured guest, friendly advisor or censored watchdog, its annual supervisory board report is unlikely to provide useful information to minority shareholders and investors. Thus the role of independent watchdog requires that members on the supervisory board have the necessary capacity in terms of knowledge and experience to act with independence and expertise. Logically those supervisory boards that have a higher number of members with appropriate professional knowledge or work experience should be in better position to improve corporate policy decision making. Larger supervisory boards may also be better able to ensure that they have a combination of members with the requisite set of skills and/or experience. This leads to our final set of hypotheses: H5A: An increase in the number of supervisory

board members with professional knowledge

or work experience in listed Chinese firms will

be associated with improved corporate policy

decisions and performance.

H5B: An increase in the size of the supervisory

board in listed Chinese firms will be

associated with improved corporate policy

decisions and performance.

3. Research Methodology 3.1. Research Schema The hypothesized relationships between the variables in this study are depicted in Figure 1. The theoretical perspective underlying the relationships is agency theory as it relates to ownership structure and the composition of the two boards of listed Chinese companies. The empirical schema for this study identifies these two classes of factors as primary influences of the bad debt ratios (defined as the values of bad debts to total assets and bad debts to accounts receivables). Additionally, the time since listing of the firm (Firm Age) is used to moderate the ownership structure factors. The choice of Firm Age as the moderating variable reflects the recognition of several important features likely to be present in China’s privatisation process. Prior to listing significant improvements in the structure of the balance sheet and firm performance are required, given


108

requirements for profitability, especially for state-controlled flagship firms. This suggests that these firms may start in a good balance sheet position in relation to our dependent variables. Thus problems due to state control/influence may only become apparent with the passage of time. For firms with high levels of foreign ownership, however, this need not be the case. China’s focus on building a set of large,

internationally competitive companies, especially in key industries such as telecommunications, energy, transport, etc., may suggest that it has allowed weaker companies to come under foreign control. Thus a particular concern is that, should foreign ownership exert a positive influence, performance may only be impacted gradually over time.

Figure 1. Empirical Schema

The choice of bad debt ratios as indicators of the

quality of corporate policy decisions reflects a number of factors related to the core characteristics of bad debts. The first is the likelihood that this item will be impacted by management decisions. This reflects that the board of directors, in consultation with management, ultimately has responsibility for decisions on credit policy with respect to standards and terms. Thus, along with the firm’s management, it can determine corporate practice with respect to credit policy, as reflected in initial credit evaluation, ongoing credit monitoring and collection, and forgiveness of delinquency and default. Where problems arise in any of these areas, action would be expected from the supervisory board. The second is that as this item will be impacted by management decisions (i.e., internal decisions), this suggests that that while there may be common trends/cycles in bad debts over time, much of the variation in bad debts between different enterprises should largely reflect firm-specific influences and decisions. 3.2. Sample and Data This study focuses on non financial A-share firms listed on either the Shanghai Stock Exchange or the Shenzhen Stock Exchange. In order to test the effects of various types of ownership—that is, high levels of

state and foreign investor ownership—the sample of companies is divided into three groups: A-share, AB-share, and AH-share companies. A-share companies are companies that have issued A-shares only, and are listed on the domestic stock exchanges. AB-share companies are those that have issued both A-shares (see endnote 9) and B-shares, (see endnote 10) with an initial A-share offering. They are also listed on the domestic stock exchanges in China. However, AH-share companies are those that have issued both A-shares and H-shares, (see endnote 11) and have floated their shares simultaneously on the Hong Kong Stock Exchange and one of China’s two mainland stock exchanges.

A sample size of 120 companies was selected from the currently listed companies in China’s Shanghai SSE180 (see endnote 12) and Shenzhen SSE100

(see endnote 31) for the period from 2001 to 2005. This was achieved through use of a stratified sampling method. As shown in Figure 2, 46 companies

(see endnote 14) were randomly selected from the A-share group, and 42 companies were randomly selected from the AB-share group. There were only 32 companies listed on both the Hong Kong Stock Exchange and one of the two mainland Chinese stock exchanges. For this reason, all of these companies were selected for our sample.


109

H-Share firms

B-Share firms

AH-Share firms 32

selected for sample

AB-Share firms 42

selected for sample

A-Share firms 46 selected for sample

from Shanghai

SSE180 and Shenzhen SSE100

Figure 2. Sampling Frame

Table 1. Composition of Sample

Share Type No. of Sample Percentage of

Sample No. of observations Percentage of

observations A 45 38.46 191 35.37 AB 42 35.90 210 38.89 AH 30 25.64 139 25.74 Total 117 100.00 540 100.00

Finally, as shown in Table 1, the final sample of

non-financial companies consists of 117 companies listed on the Shanghai and Shenzhen stock exchanges (see endnote 15). Over the five-year period from 2001 to 2005 this has resulted in 540 observations (see

endnote 16).

3.3. Model Development and Variable Measurement According to the hypotheses and empirical schema discussed above, the theoretical model to be formed is as follows:

titiP

P

tiPti

FactorsnCompositioBoard

ModeratorFactorsStructureOwnershipBDRATIO

,,

1,,

)(

)(

εβ

βα

++

×+=

∑

∑=

(1)

The panel data regression models to be empirically investigated in this study are stated as follows:

titititi

titititi

tititititi

PROFSBTSBBSIZE

INDPAGEFORTOPFORTOP

AGESTOPSTOPTOPBDRATIO

,,9,8,7

,6,,5,4

,,3,2,1,

1010

101010

εβββ

βββ

βββα

+++

++∗++

∗+++= (

2) The variables are in each model comprised of

three types: one dependent variable, seven independent variables, and one moderating variable. The definition and measurement for each of the variables in this study are listed Table 2.

Table 2. Definition and Measurement of Variables

Variable Acronym

Definition Expected Sign

Measurement

Dependent: BDTA

BDRA

The bad debt ratio measured relative to total assets (BDTA) or accounts receivable (BDAR)

N/A Either

ti

ti

tiAssetsTotal

DebtsBadBDTA

,

,

, =,

or

ti

ti

tisReceivable Accounts

DebtsBadBDAR

,

,, =

.

Where: Bad Debts = total bad debts at the end of a reporting year; Receivables = value of accounts receivable at the end of a reporting year; Total Assets = book value of total assets at the end of a reporting year; i = sampled company; and t = year.


110

Table 2 continued Independent: TOP10 Overall ownership

concentration +ve Proportion of total shares held by the top 10 shareholders

STOP10 State ownership concentration

+ve Proportion of shares held by the state in those held by the top 10 shareholders

FORTOP10 Foreign ownership concentration

-ve Proportion of shares held by foreign owners from those held by the top 10 shareholders

INDP Independent directors -ve Number of independent directors on the corporate board BSIZE Board of directors size -ve Number of directors on the board TSB Supervisory board size -ve Number of members of the supervisory board PROFSB Professionalism of the

supervisory board -ve Number of supervisory board members with professional knowledge or work

experience Moderating: AGE Firm age N/A Years since initial listing

4. Results 4.1. Descriptive Statistics

Table . Descriptive Statistics on Variables

Variable Mean Median Min Max Standard Deviation Sample: 2001-2005 Observations: 540

BDTA 0.0347 0.0076 0.0000 1.3494 0.1129 BDAR 0.2609 0.0959 0.0000 13.1147 0.8403 TOP10 0.6550 0.6616 0.2145 0.9967 0.1627 STOP10 0.6486 0.7067 0.0000 1.0000 0.3078 FORTOP10 0.1408 0.0398 0.0000 0.5906 0.1670 INDP 2.9259 3.0000 0.0000 6.0000 1.3762 BSIZE 10.4130 10.0000 5.0000 19.0000 2.4217 TSB 4.5889 5.0000 2.0000 12.0000 1.6697 PROFSB 1.8593 2.0000 0.0000 5.0000 1.0385 AGE 6.7259 9.0000 0.0000 14.0000 3.4906

The descriptive statistics in Table 3 provide a

profile of the corporate governance characteristics of the listed A-share, AB-share and AH-share Chinese companies in our sample. First, the mean ownership concentration (i.e., TOP10) is 65.5 per cent, with a range of 21.45 per cent to 99.67per cent. This is consistent with previous studies by Xu and Wang (1999) and Deng and Wang (2006) that show high ownership concentration in listed companies in China. As argued by Deng and Wang (2006) this supports the potential for larger shareholders to dominate listed firms in China. Second, the mean concentration of state ownership in the top 10 company shareholders (i.e., STOP10) is 64.86 per cent, with a range from zero to 100 per cent. This supports the argument that the state has maintained a dominant role in the operation of many previously SOEs. Third, the mean concentration of foreign ownership in the top 10 company shareholders (i.e., FORTOP10) is only 14.08 per cent, with a range from zero to 59.06 per cent. Fourth, with respect to board composition, the mean number of independent directors (i.e., INDP) is 2.93, with a range of zero to six. The minimum reflects that appointment of independent directors was rare before the year 2002. However, effective from

2003 CSRC guidelines have required at least one-third of the board directors to be independent. Fifth, the mean number of professional supervisors (i.e., PROFSB) is 1.86, with a range from zero to five. The lower value of the range reflects the period prior to the 2002 issue of The Code by the CSRC, which requires that supervisors have professional knowledge or work experience in such areas as law and accounting. 4.2. Multivariate Analysis and Hypothesis Testing Generalized least squares (GLS) fixed effects methods are used in this study. A panel regression model (see equation (2)) was estimated using the three ownership structure variables (with two of them moderated by Firm Age) and the four board composition variables for each of the measures of the dependent variables. The possible existence of multicollinearity was tested. Gujarati (2003) argues that correlations between the independent variables should not be deemed harmful for multivariate analysis unless they exceed 0.8.


111

Table 4. Correlation Matrix for Independent Variables

STOP10 FORTOP10 TOP10 TSB PROFSB BSIZE INDP AGE STOP10

1.0000

FORTOP10

-0.3979**

1.0000

TOP10

0.1252**

0.4396**

1.0000

TSB

0.1720**

0.0322

0.1857**

1.0000

PROFSB

0.1185**

0.0830*

0.2721**

0.6160**

1.0000

BSIZE

-0.0116

0.1953**

0.0855*

0.4096**

0.2032**

1.0000

INDP

-0.0655

0.2018**

0.1535**

0.2483**

0.2237**

0.4517**

1.0000

AGE

-0.1407**

0.0294

-0.3722**

-0.0627

0.0113

-0.1424**

0.0672

1.0000

Note: * significant at the 5% level; ** significant at the 1% level.

As shown in Table 4, there are no correlations

between independent variables that reach this level. However, a certain degree of multicollinearity can still exist even when none of the bivariate correlation coefficients is very large. The reason is one independent variable may have a linear function with a set of several independent variables (Gujarati, 2003). Hence, multicollinearity is also tested for using the Variance Inflation Factor (VIF) (see endnote 17). The result, not reported in the tables, is that the largest VIF is 1.92 and that the VIFs of all other independent variables are below 1.86. Thus, there is no serious multicollinearity problem present in the regression model.

Tables 5 and 6 provide the panel regression results to test the five hypotheses. These reveal a high adjusted-R

2 of 0.63 for the panel regression using BDTA as the

dependent variable, and an adjusted-R2 of 0.35 for the panel regression using BDAR as the dependent variable. The F statistics for each of the regression models indicate that a statistically significant component of the variation in the chosen measure of the bad debt ratio is explained by variation in the set of independent variables. The discussion that follows examines the results in Tables 5 and 6 in terms of the five hypotheses established earlier.

Table 5. Panel Regression Results BDTA

Dependent variable: Bad Debt Ratio – BDTA

Sample: 2001 – 2005 Cross-sections: 117 Panel obsv: 540

Adjusted-R2: 0.63 R2: 0.72 F significance: 0.00 F-statistic: 8.01**

Independent variables: Expected signa Coefficient Standard error t-statistic Constant N/A -0.0016 0.0216 -0.0744 Overall ownership concentration – TOP10 + 0.0281 0.0131 2.1411* State ownership concentration – STOP10

– -0.0553 0.0247 -2.2349*

State ownership concentration moderated by firm age – STOP10*AGE

+ 0.0058 0.0021 2.7069**

Foreign ownership concentration – FORTOP10 + 0.0913 0.0290 3.1519** Foreign ownership concentration moderated by firm age – FORTOP10*AGE

– -0.0161 0.0029 -5.6107**

Independent directors on the corporate board - INDP – -0.0021 0.0005 -4.2001** Directors on the corporate board – BSIZE – 0.0036 0.0016 2.2228* Supervisors on the supervisory board – TSB – -0.0009 0.0015 -0.6056 Qualified and experienced supervisors on the supervisory board – PROFSB

– 0.0025 0.0020 1.2572

Notes: a Giving consideration to the arguments associated with the use of the moderator variable (AGE). * Statistically significant at the 5% level. ** Statistically significant at the 1% level.


112

Table 6. Panel Regression Results BDAR

Dependent variable: Bad Debt Ratio – BDAR Sample: 2001 – 2005

Cross-sections: 117 Panel obsv: 540 Adjusted-R2: 0.35 R2: 0.50

F significance: 0.00 F-statistic: 3.30** Independent variables: Expected sign Coefficient Standard error t-statistic Constant N/A -0.4247 0.2545 -1.6688 Overall ownership concentration – TOP10 + 0.7163 0.2407 2.9755** State ownership concentration – STOP10

– -1.0485 0.1705 -6.1502**

State ownership concentration moderated by firm age – STOP10*AGE

+ 0.1517 0.0198 7.6741**

Foreign ownership concentration – FORTOP10 + 0.3903 0.1954 1.9977* Foreign ownership concentration moderated by firm age – FORTOP10*AGE

– -0.0671 0.0118 -5.6683**

Independent directors on the corporate board - INDP – -0.0365 0.0180 -2.0302* Directors on the corporate board – BSIZE – 0.0381 0.0132 2.8932** Supervisors on the supervisory board – TSB – 0.0099 0.0150 0.6604 Qualified and experienced supervisors on the supervisory board – PROFSB

– -0.0364 0.0142 -2.5695*

Notes: a Giving consideration to the arguments associated with the use of the moderator variable (AGE). * Statistically significant at the 5% level. ** Statistically significant at the 1% level.

The first hypothesis (H1) states that a higher

level of ownership concentration is associated with poor corporate policy decisions and performance. In this specific case poor corporate performance is defined in terms of a higher level of the bad debt ratio. This negative effect on firm performance in this area arises from poor corporate policy decisions that result from the agency conflict between majority and minority shareholders. The ownership concentration measure is assumed to reflect the distribution of power within in a firm. The results of Tables 5 and 6 show a positive relationship between ownership concentration (TOP10) and our alternative measures of the bad debt ratio. This is consistent with the perspective of Shleifer and Vishny (1997), that agency problems involve expropriation from minority by majority shareholders. They refer to this case as “tunnelling”, and determine that it is likely to be a significant problem in emerging market economies. Thus, we find support for the first hypothesis.

The second hypothesis (H2) implies that where a greater percentage of shares are held by the state sector this is associated with a higher level of the bad debt ratio. This reflects that the corporatisation of former SOEs in China via share issue has not effectively dealt with the agency problems associated with public ownership (Chen, 2004). In the case of higher levels of state control (STOP10) we find that, in contrast to our initial hypothesis, the coefficient is negative. This would suggest that rather than increase the bad debt ratio higher levels of state ownership reduce it. However, as argued above, consideration needs to be given to the financial requirements required of SOEs prior to listing in China. Thus, in considering our initial hypothesis we also focus on the coefficient on the level of state control combined with our moderator variable (STOP10*AGE). For this variable the coefficient for the impact over time of

high levels of state ownership on the bad debt ratio is positive. As newly listed former SOEs will start with a relatively clean bill of financial health, we would expect a relatively low initial level for the bad debt ratio, and thus potentially a negative coefficient on STOP10. Given this good start, it is important to our second hypothesis that high levels of state ownership are associated with an increase in the level of the bad debt ratio over time (STOP10*AGE). We therefore find support for our second hypothesis, based on the impact of our moderation variable, and accept it. This suggests that agency problems still exist within our sample firm set due to a misalignment between shareholder and state objectives for these firms.

The third hypothesis (H3) implies that where there is a greater percentage of shares held by foreign investors (FORTOP10) this will be associated with a lower level of the bad debt ratio. The expectation is that, in the case of China, improving corporate governance and, more broadly, accounting standards, is in the interests of international investors. However, we find that the coefficient on FORTOP10 is positive, suggesting that foreign ownership increases the level of the bad debt ratio. Again, as when considering the impact of high levels of state ownership, we find that when we focus on the share of foreign ownership in the top ten shareholders is moderated by firm age (FORTOP10*AGE), the coefficient has the expected negative sign. This lends support to our third hypothesis, which we cautiously accept (see endnote

18). We thus suggest that rather than having an immediate impact on corporate governance, the impact of foreign ownership occurs progressively over time.

The fourth set of hypotheses (H4A and H4B) argue that the larger the number of independent directors on the board and the larger the corporate board the lower will be the bad debt ratio. The results


113

in Tables 5 and 6 indicate that there is a negative relationship between the alternative measures of the bad debt ratio and the number of independent directors, supporting the first of these hypotheses. This adds weight to the agency argument of Fama and Jensen (1983) that independent directors are motivated to work in the best interests of shareholders in order to maintain their good personal reputation. However, when considering board size, we the opposite results to that hypothesized for the coefficients in each regression. Rather than having a beneficial impact on the bad debt ratio, it appears that larger boards worsen corporate policy making being associated with a higher level of the bad debt ratio. While this latter result may reflect that the requirement that at least one-third of board members has only been in place since 2003, an issue for further consideration and research, we must reject the second of our fourth set of hypotheses.

The fifth set of hypotheses (H5A and H5B) state that the higher the proportion of supervisory board members with relevant professional knowledge or work experience and the larger the supervisory board the lower will be the bad debt ratio. The argument in the first case is that the supervisory board will require high professionalism from its members to effectively carry out its role of overseeing the performance of the corporate board and management and protecting stakeholders’ rights and interests. However, the coefficients for supervisory board size (TSB) in Tables 5 and 6 are not significant. Additionally, while the coefficient on the skills of the supervisory board (PROFSB) is of the hypothesized sign and statistically significant in the case where BDAR is the dependent variable, this is not the case for the coefficient where BDTA is the dependent variable. This clearly suggests that further research into the matter is required, as is consideration of chosen performance metrics. We thus conclude that based on our results, the effectiveness of supervisory board in improving corporate governance may be questioned. This is consistent with Dahya et al. (2002), who find that because of the transitional nature of the Chinese economy supervisory board effectiveness is limited. It may be speculated that supervisory boards of listed companies in China have tended to become “censored watchdogs” in the words of Dahya et al. (2003), during a period when rapid corporate expansion and the dominance of the corporate board has occurred. Thus, given the conflict between the results of our two alternative panel regressions, we reject our fifth set of hypotheses. 5. Conclusions The paper has argued that the priority given to establishment of institutional structures reflects their importance in establishing an environment appropriate to a post-transition, market-oriented economic system. Sound institutional choices with

respect to the legal and regulatory environment are particularly important to the reform of SOEs, as they assist in the separation of government from SOE management, restructuring of SOEs, and thus the effectiveness of corporate governance frameworks in the transition economy. Assessment of these institutions requires that we consider whether they assist (or are effective) in producing outcomes consistent with their stated intent.

In the case of China the corporate governance structures charged with shaping corporate decision-making behaviour consist of The Company Law,

The Securities Law, and The Code (the last being a code of good practice rather than a legal requirement). Thus a key issue to address in China is whether these corporate governance structures have led to good corporate governance outcomes, or potential deficiencies in the capacity of these institutions may be identified. These outcomes relate to: corporate ownership and the exercise of control by majority shareholders at the expense of minority shareholders; the independence of the board of directors and the quality of their decisions; and to performance of the supervisory board in protecting the company’s and stakeholders’ rights and interests. Failure in any of these areas suggests that corporate governance structures have not properly addressed an aspect of the agency problem.

In this context this paper has examined the relationship between corporate policy decisions, as measured by our proxies for performance, and the corporate control and governance characteristics of listed, non financial Chinese companies. The characteristics explored related to both ownership and board structure under China’s two-tier board system. The corporate control characteristics are the level of concentration in and type of ownership, particularly high levels of government and foreign ownership, and thus potential for abuse of a majority shareholder position. With respect to governance characteristics, as reflected in the composition of companies’ boards, there are the matters of board size, expertise and independence, the latter being specific matters dealt with under China’s corporate governance structures. To represent the quality of corporate policy decisions we chose two alternative measures of the bad debt ratio. These were the bad debt to total asset and bad debt to accounts receivable ratios. The choice of performance metrics related to bad debts reflect that the sample firms’ boards and management have direct influence over a range of practices that influence this variable. These include the ability to determine corporate practice with respect to credit policy, as reflected in initial credit evaluation, ongoing credit monitoring and collection, and forgiveness of delinquency and default.

To assess the impact of these characteristics within the Chinese institutional structure we used an unbalanced panel data set covering the years 2001 to 2005. This was comprised of a stratified sample of


114

observations for 117 A, AB and AH non financial companies listed on the Shanghai and Shenzhen stock exchanges. We hypothesized that due to agency problems that high levels of concentration in firm ownership, and high levels of state control of enterprises, would be associated with poorer quality corporate decision making and thus higher levels of the bad debt ratio. High levels of foreign ownership, high levels of independent directors on the corporate board, and high levels of skill and experience on the supervisory board were hypothesized to decrease bad debt ratios.

Our results suggest that concentration of ownership in general, and high levels of state ownership in particular, are associated with an increase in the bad debt ratio. However, in the latter case it is apparent that, rather than state dominated firms starting with a poorer balance sheet position, state influence has a negative impact over an extended period. Such an outcome suggests that in some cases state objectives rather than those of non state shareholders may dominate in the decision making process. Similarly, consistent with the likely objectives of foreign investors, high levels of foreign ownership are found to impact the bad debt ratio over an extended time; in this case, acting to reduce the bad debt ratio.

Board size and independence are found to be significant factors in determining the ratio of bad debts to total assets for these listed Chinese firms. Independence of the board was shown to be a significant factor in reducing the bad debt ratio. However, rather than a larger board leading to improved performance, we find that increases in the size of the board are associated with higher bad debt ratios. We have argued that this may reflect that it is only relatively recently (i.e., in 2003) that requirements regarding board independence in China were established. Thus this warrants further exploration of the panel data set and its properties, and must be recognised as one of the limitations of this study. Thus future research in the area should seek to increase the size and scope of the sample utilised, in order to address concerns related to the sample size and the length of period covered.

In common with previous research we have failed to find support for the importance of the size of the supervisory board in corporate policy decision making. However, with respect to the impact of the professionalism of the supervisory board on corporate policy decision making we have produced conflicting results. While one of our regression models suggests that no statistically significant impact is present, in the other the qualifications and professionalism of the board had a significant and negative effect on the level of the bad debt ratio. While rejecting support for the importance of this factor, due to the conflict in our results, we recognise this as another limitation of this study. Again we recognise the need to further explore the panel data set, and choice and properties of performance metrics used in this study.

Overall our results are suggestive of the need for China to continue to address the underlying effectiveness of its corporate governance framework. Given our mixed results we suggest that it must continue to act to ensure that the supervisory board’s effectiveness is enhanced. This is in order that it becomes an independent watchdog (as per Dahya et

al., 2003). Furthermore, it must ensure that its corporate governance model addresses the issue of majority shareholder influences on firm decision making. This is especially the case with the strong links that still appear to remain between the state and formerly state-owned enterprises.

References

1. Beim, D.O. and Calomiris, C.W. (2001), Emerging

Financial Markets, McGraw-Hill/Irwin, New York, NY.

2. Berle, A. and Means, G. (1932), The Modern

Corporation and Private Property, Transaction Publishers, New Brunswick.

3. Brickley, J., Lease, R.C. and Smith Jr., C.W. (1988), “Ownership Structure and Voting on Antitakeover Amendments”, Journal of Financial Economics, Vol. 20, pp. 267–291.

4. Bushee, B.J. (1998), “The Influence of Institutional Investors on Myopic R&D Investment Behaviour”, The

Accounting Review, Vol. 73, pp. 305–333. 5. Chen, J.J. (2004), “Corporatisation of China’s State-

Owned Enterprises and Corporate Governance”, Corporate Ownership and Control, Vol. 1 No. 2, pp. 82-93.

6. China Securities Regulatory Commission (2005), Company Law of the People’s Republic of China (revised edition), www.csrc.gov.cn.

7. China Securities Regulatory Commission (2005), Securities Law of the People’s Republic of China

(revised edition), www.csrc.gov.cn. 8. China Securities Regulatory Commission and State

Economic and Trade Commission (2005), Code of Corporate Governance for Listed Companies in China (revised edition), www.csrc.gov.cn.

9. Clarke, T. (1998), “The Stakeholder Corporation: A Business Philosophy for the Information Age”, Long

Range Planning, Vol. 31 No. 2, pp. 182-194. 10. Dahya, J., Karbhari. Y. and Xiao, J.Z. (2002), “The

Supervisory Board in Chinese Listed Companies: Problems, Causes, Consequences and Remedies”, Asia

Pacific Business Review, Vol. 9 No.2, pp. 118-137. 11. Dahya, J., Karbhari, Y., Xiao, J.Z. and Yang, M.

(2003), “The Usefulness of the Supervisory Board Report in China”, Corporate Governance: An

International Review, Vol. 11, No.4, pp. 308-321. 12. Deng, X. and Wang, Z (2006), “Ownership Structure

and Financial Distress: Evidence from Public-listed Companies in China”, International Journal of

Management, Vol. 23 No.3, pp. 486-502. 13. Fama, E.F. and Jensen, M.C. (1983), “The Separation

of Ownership and Control”, Journal of Law and

Economics, Vol. 26 No. 2, pp. 301-325. 14. Firth, M., Fung, P.M.Y. and Rui, O.M. (2007),

“Ownership, Two-tier Board Structure, and the Informativeness of Earnings: Evidence from China”,


115

Journal of Accounting and Public Policy, Vol. 26 No. 4, pp. 463-496.

15. Gujarati, D.N. (2003), Basic Econometrics, 4th edn, McGraw-Hill, New York, NY.

16. Gunasekarage, A., Hess, K. and Hu, A.J. (2007), “The Influence of the Degree of State Ownership and the Ownership Concentration on the Performance of Listed Chinese Companies”, Research in International

Business and Finance, Vol. 21, pp. 379-395. 17. Jevons Lee, C.-W. (2001), “Financial Restructuring of

State Owned Enterprises in China: The Case of Shanghai Sunve Pharmaceutical Corporation”, Accounting Organizations and Society, Vol. 26, pp. 673-689.

18. Krzywda, D., Bailey, D. and Schroeder, M. (1995), “A theory of European accounting development applied to accounting change in contemporary Poland”, The

European Accounting Review, Vol. 4 No. 4, pp. 625–657.

19. La Porta, R., Lopez-de-Silanes, F. and Shleifer, A. (1999), “Corporate Ownership Around the World”, Journal of Finance, Vol. 54 No. 2, pp. 471-517.

20. Lin, T.W. (2004), “Corporate Governance in China: Recent Developments, Key Problems, and Solutions”, Journal of Accounting and Corporate Governance, Vol. 1 No. 1, pp. 1-23.

21. Mallin, C.A. (2007), Corporate Governance, 2nd edn, Oxford University Press, Oxford.

22. Morck, R., Shleifer, A. and Vishny, R.W. (1988), “Management Ownership and Market Valuation: An Empirical Analysis”, Journal of financial economics, Vol. 20 No.1/2, pp. 293-315.

23. Nelson, R.R. (2007), “Institutions and Economic Growth: Sharpening the Research Agenda”, Journal of

Economic Issues, Vol. 41 No. 2, pp. 313-323. 24. North, D. (1990), Institutions, Institutional Change and

Economic Performance, Cambridge University Press, Cambridge.

25. Organisation for Economic Co-Operation and Development (2000), China in the Global Economy: Reforming China’s Enterprises, OECD Economics: Emerging Economies in Transition. Paris: OECD.

26. Organisation for Economic Co-operation and Development (2004), OECD Principles of Corporate Governance. Paris: OECD.

27. Pound, J. (1988), “Proxy Contests and the Efficiency of Shareholder Oversight”, Journal of Financial

Economics, Vol. 20, pp. 237–265. 28. Qiang, Q. (2003), “Corporate Governance and State-

Owned Shares in China Listed Companies”, Journal of

Asian Economics, Vol. 14, pp. 771–783. 29. Redek, T. and Sušjan, A. (2005), “The Impact of

Institutions on Economic Growth: The Case of Transition Economies”, Journal of Economic Issues, Vol. 39 No. 4, pp. 995-1027.

30. Rodrik, D. (2000), “Institutions for High Quality Growth: What They Are and How to Acquire Them”, NBER Working Paper Series no. 7540.

31. Shleifer, A. and Vishny, R.W. (1997), “A Survey of Corporate Governance”, Journal of Finance, Vol. 52 No. 2, pp. 737-782.

32. Šević, Ž. (2005), “Emerging Markets: Preferences, Risks, Performance...”, Economic Change and

Restructuring, Vol. 38, pp. 1-10. 33. Sun, Q. and Tong, W.H.S. (2003), “China Share Issue

Privatization: The Extent of its Success”, Journal of

Financial Economics, Vol. 70 No. 2, pp. 183-222.

34. Tadesse, S. (2004), “The Allocation and Monitoring Role of Capital Markets: Theory and International Evidence”, Journal of Financial and Quantitative

Analysis, Vol. 39 No. 4, pp. 701-730. 35. Wei, G. (2007), “Ownership Structure, Corporate

Governance and Company Performance in China”, Asia Pacific Business Review, Vol. 13 No. 4, pp. 519-545.

36. Wei, Z., Xie, F. and Zhang, S. (2005), “Ownership Structure and Firm Value in China's Privatized Firms: 1991-2001”, Journal of Financial and Quantitative

Analysis, Vol. 40 No.1, pp. 87-108. 37. Williamson, O. (2000), “The New Institutional

Economics: Taking Stock, Looking Ahead”, Journal of

Economic Literature, Vol. 38, pp. 595-613. 38. World Bank (2004), Country Policy and Institutional

Assessment Index, www.worldbank.org. 39. Xiao, J.Z., Yang, H. and Chow, C.W. (2004), “The

Determinants and Characteristics of Voluntary Internet-Based Disclosures by Listed Chinese Companies”, Journal of Accounting and Public Policy, Vol. 23 No. 3, pp. 191-225.

40. Xu, X. and Wang, Y. (1999), “Ownership Structure and Corporate Governance in Chinese Stock Companies”, China Economic Review, Vol. 10 No. 1, pp. 75-98.

41. Zhu, T. (1999), “China’s Corporatization Drive: An Evaluation and Policy Implications”, Contemporary

Economic Policy, Vol. 17 No. 4, pp. 530-539. Endnotes

1 Each of these laws/codes was revised in 2005. 2 As well as governance, the definition of institutions usually covers polity, the judiciary and legal system, economic policy, rules, regulations, etc. For example, the Country Policy and Institutional Assessment Index developed by the World Bank provides an indication of the diversity of the political, social and economic structures defined as institutions. It is comprised of 16 components grouped under four broad categories: economic management; structural policies; policies for social inclusion/equity; and public sector management and institutions. Economic management includes macroeconomic management, fiscal policy, and debt policy. Structural policies cover trade policy, financial sector regulation, and the business regulatory environment. Policies for social inclusion/equity cover matters dealing with gender equality, the equity of public resource use, building human resources, social protection and labour, policies and institutions for environmental sustainability, and property rights and rule-based governance. Public sector management and institutions refers to the quality of budgetary and financial management, efficiency of revenue mobilization, quality of public administration, and transparency, accountability, and corruption in the public sector (World Bank, 2004). 3 To this may be added a variety of other laws, including the Audit Law (1994), Accounting Law (1999), etc. 4 The Code is based on the OECD (2004), OECD Principles

of Corporate Governance. 5 While examining the development of the Chinese capital markets, Qiang (2003) also recognises the presence of multiple agency problems. Many of these are caused by direct and indirect government influence on corporate governance matters in China’s listed companies.


116

6 Gunasekarage, Hess and Hu (2007) also find weak indication of a negative effect for high ownership concentration. 7 While independent from the company as defined in The

Code, the effectiveness of the corporate board in China may differ from that of Western corporate boards. This is because of the close connection between controlling investors and the central or provincial government. The government may influence the appointment of directors and senior managers, and even interfere with the decision-making of a listed firm (Firth, Fung and Rui (2007). 8 Dahya, Karbhari, Xiao and Yang (2003) utilise qualitative research based on interviews of supervisory board members to gain these insights into the behaviour and functioning of supervisory boards in China. 9 A-shares are common stock issued by mainland China firms, subscribed and traded in RMB, listed on the mainland stock exchanges, and are reserved for trading by Chinese citizens. The A-share market was launched in 1990 in Shanghai. 10 B-shares are issued by mainland China firms, traded in foreign currencies, and listed on the mainland stock exchanges. The B-share market was launched in 1992 and was restricted to foreign investors before 19 February 2001. 11 H-shares are securities of companies incorporated in mainland China and nominated by the Chinese Government for listing and trading on the Hong Kong Stock Exchange, being quoted and traded in HKD. There are no restrictions on holdings by international investors. 12 Shanghai SSE180 Index was created by restructuring and renaming the SSE30 Index. Through scientific and objective

methods it selects constituents that best represent the market. The SSE is a benchmark index reflecting the Shanghai market and serves as a performance benchmark for investment and a basis for financial innovation. 13 The Shenzhen SSE100 is a benchmark index reflecting performance in the Shenzhen market and serves as a performance benchmark for investment and as a basis for development of financial innovations. 14 These samples are randomly selected from A-share companies included in the Shanghai SSE180 and the Shenzhen SSE100 after removing dual listed companies (these being either AB-share companies or AH-share companies). 15 One company found to be a financial company and two companies listed after 2005 have been removed. This results in 117 companies being present in the final sample. 16 45 observations for which the data was incomplete have been excluded, resulting in 540 observations. 17 The critical value of the VIF to test for multicollinearity is 10. Gujarati (2003) suggests that there is no evidence of multicollinearity unless the VIF of a variable exceeds 10. All values used in this study were well below this critical level. 18 This caution recognises that additional information is required to properly explain why firms with high levels of foreign ownership may initially be expected to have higher levels of bad debts relative to total assets.


117

AUDIT COMMITTEES IN NIGERIA

Ogbuagu Ekumankama and Chibuike Uche*

Abstract A new requirement in the Nigerian Companies and Allied Matters Act (CAMA) of 1990 is the introduction of audit committees as an additional layer of control and certification in the bid to make annual accounts of public corporations more acceptable and reliable. This paper reviews the law and practice of audit committees in Nigeria. It argues that for audit committees to become more useful in the Nigerian context there is need for changes to be made in both its law and practice. Key areas of concern include the need to: determine and codify the qualification for membership of the committee given its technical nature; allow appropriate remuneration for committee members, and; the determination of appropriate membership tenure for such committees. The above review is necessary if audit committees are to be in a position to effectively perform their oversight functions aimed at improving the quality and information content of corporate financial reports. Keywords: directors, audit committee, control *Department of Banking and Finance, University of Nigeria, Enugu Campus, Enugu, Nigeria Ogbuagu Ekumankama is currently a Doctoral (M.Sc/ PhD) Student in the Department of Banking and Finance at the University of Nigeria Enugu Campus. He holds a Master of Business Administration Degree in Banking and Finance from the same University. This article is mainly based on his research project during his MBA programme. Chibuike Uche is Professor of Banking and Financial Institutions in the same department.

Introduction A requirement of the Nigerian Companies and Allied Matters Act (CAMA) of 1990 is the introduction of audit committees as an additional layer of control and certification in the bid to make annual accounts of public corporations more acceptable and reliable.1 Prior to the 1990 Act, the only statutory requirement for the certification of annual accounts of public corporations in Nigeria was the provision that such accounts be audited by external auditors.2 The idea of appointing external auditors arose in the quest to find more efficient ways of promoting accountability in complex organizations where management interests could differ from shareholder interests. Initially, the focus was on explicitly spelling out the characteristics of auditors in order to ensure their independence and competence.3 The law usually stipulates that external auditors should be appointed by shareholders and report to shareholders at annual general meetings. The independence of external auditors was further enhanced by the fact that their reports were treated as professional opinions and thus attracted some degree of liability in the event it is shown that they have been negligent in the conduct of their duty (Sasegbon, 1991, pp. 588-9).

Over time, however, various accounting and reporting scandals have led to corporate failures and embarrassments. Examples include the celebrated case of the Mckesson and Robbins (USA) in the late 1930s, and the failure of the Atlantic Acceptance Corporation (Canada) in 1965 (Cf. Lee and Stone, 1997, p.99; and Okaro, 2001, p155). This has led to

the questioning of very concept of auditor independence. Although the law usually provides that auditors should be appointed by shareholders and report to such shareholders in annual general meetings, the reality is somewhat different. In practice, annual general meetings are no more than rubber stamps for board decision on such matters. Within the board itself, executive directors usually have an upper hand since they deal with auditors on a day to day basis. Under such circumstances, the ability of such external auditors to remain truly independent, especially if there is need to express reservations about management’s accounting policies, is whittled down. The idea thus developed that accountability will be enhanced if a subcommittee of the board: audit committee, comprising only of independent directors, be appointed to act as an arbiter between external auditors and management.4 The assumption is that such a committee is more likely able to protect the interest of shareholders. This has been the guiding principle behind the establishment and codification in the laws of the audit committee requirement all over the world. The Audit Committee requirement was enshrined in the Nigerian Companies and Allied Matters Act in 1991. Despite the fact that this provision has been in existence for more than fifteen years, its utility value especially with respect to enhancing the information value and credibility of financial accounts remain suspect (Okaro, 2001, p.157).

This paper reviews the law and practice of audit committees in Nigeria. It argues that while its introduction in Nigeria is a welcome development


118

there is need for the review of the existing laws governing the appointment, remuneration, duration of tenure and qualifications of audit committee members in the country. The above review is necessary if audit committees are to be in a position to effectively perform their oversight functions aimed at improving the quality and information content of corporate financial reports. To achieve its aim, the remaining part of this paper is divided into three sections. Section One examines the theory and origins of the idea of audit committees while Section Two critiques the law and practice of audit committees in Nigeria. Section Three concludes the paper

The Theory and Origins of Audit Committees

Theoretically, there is widespread agreement that the separation of ownership from control (the principal- agent problem) is the economic explanation for auditing (Lennox, 2003, p.5). Principal-agent problems mainly arise in companies where management is distinct from ownership. As far back as 1776, for instance, Adams Smith vividly explained this concept. He essentially argued that Joint stock companies are usually managed by directors who are usually subject to the control of the proprietors who seldom understand in detail the activities of the company. They are thus usually satisfied with the periodic dividends that are declared by the directors. The total exemption from both trouble and risk beyond a limited point offered by joint stock companies tends to encourage investors to prefer it over investments in private businesses which offer no such buffers. Joint Stock companies thus became the preferred investment option. Despite this preference, Adam Smith conceded that directors of joint stock companies, being the managers of other peoples’ money, rather than their own, could not be reasonably expected to watch over it with the same vigilance with which the partners of private enterprises watch over their own (Spira, 2002, p. 2).

The above description is no doubt the foundations of the agency problem in shareholder manager relationships. There are however other extensions of this problem. Agency problems can also arise when managers remuneration are based on accounting performance. This could provide the incentive for managers to be engaged in income management, which subsequently impacts on the quality of market information which is critical for investment decisions. Along these lines, it has been asserted that:

When manager incentives are based on their companies’ financial performance, it may be in their self-interest to give the appearance of better performance through earnings management. In many companies, managers are compensated both directly (in terms of salary and bonus) and indirectly (in terms of prestige, future promotions and job

security) depending on a firm’s earning performance relative to some pre-established benchmark. This combination of management’s discretion over reported earnings and the effect this earnings have on their compensation leads to a potential agency problem. Beyond the management compensation problem, earnings management may impact investors by giving them false information. Capital markets use financial information to set security prices. Investors use financial information to decide whether to buy, sell or hold securities. Market efficiency is based upon the information flow to capital markets. When the information is incorrect, it may not be possible for the markets to value securities correctly. To the extent that earnings management obscures real performance and lessens the ability of shareholders to make informed decisions, we can view earnings management as an agency cost (Xie et al, 2003, p. 297).

The above scenario is even made more complex when executive directors of a company dominate the board of directors. This is especially so given the fact that it is usually the responsibility of the directors to prepare financial statements which should reflect a true and fair view of the operations of the company during the financial year.5 A conflict thus arises when directors act as both shareholders and managers of a publicly quoted company. It can, for instance be argued that higher levels of stock ownership motivate directors to artificially boast reported performance. This brings them on direct conflict with external auditors who are supposed to be independent experts whose main role is to certify the credibility of financial statements to shareholders. Based on the above, owning large stockholdings in a company will weaken the performance of audit committee members. An interpretation of the above is that all things being equal, audit committee members who own less of the company’s stock will be more likely to resist managerial attempts to manipulate financial statements and give the appearance of better performance through earnings management. Such members are also more likely to resist any attempt by management to dismiss an external auditor following the issuance of a going concern report (Carcello and Neal, 2003a, p.96).

Based on the above theory, it is not surprising that the idea of audit committees have continued to gain increasing prominence over time. In some countries, the emergence of audit committees has followed some form of financial crisis or scandal. In the United States of America, for instance, the origin of the emergence of the audit committee has been


119

linked to the McKesson and Robbins Inc. incident in the late 1930s (DeZoort, 1997, p.211; Olowokure and Nnadi, 1989a, p.18). The Securities and Exchange Commission reacted to this incident by issuing Accounting Series Release Number 19 in 1940. This essentially recommended that in order to assure auditor independence, a committee should be selected from non executive board members to nominate auditors and arrange details of engagement for such auditors (Williams, 1977, p.71).

Another country where the advent of Audit committees have been crisis induced is Canada. There, the 1965 collapse of the Atlantic Corporation Limited led to a rethink of financial practices and corporate governance. The Royal Commission subsequently set up to investigate the failure of the Corporation recommended that an audit committee consisting of not less than three directors of a company, the majority of whom should be non executives, should review the companies financial statements before approval by the board of directors. The committee was also required to confer with auditors (Collier, 1992, p.20). Australia is yet another example of a country where the idea of audit committees emerged after a financial incident. In 1979, the country’s Corporate Affairs Commission endorsed, after an inquiry into the affairs of Gollin Holdings, the principle that companies should be encouraged to have audit committees (Collier, 1992, p.24).

Some other countries have simply adopted the concept of audit committees partly based on the American influence and partly as a means of reinforcing the financial credibility of their publicly quoted companies. Examples of countries in this category include United Kingdom and Singapore.6 The establishment of audit committees in Nigeria also falls, in the main, under this category. It has, for instance, been suggested that:

The factors which motivated the establishment of audit committees in the US and Canada are not present in Nigeria. There… [is] no evidence yet of company scandals, directors and auditors have not been subject of litigations, and disclosures in company accounts are laid down by law. However, it seems that the investing public in this country is getting skeptical about the credibility of financial reports including the auditors’ reports and the performance of directors. This lack of faith in the directors was manifested recently by the formation in Lagos of a body known as ‘Shareholders Solidarity Association’… In a television interview, the chairman of the new body indicated that his organization plans to monitor the activities of companies in which its members hold

shares. This is an indictment on the directors because the shareholders are indirectly accusing them of paying inadequate attention to the affairs of the companies. It also shows distrust of the auditors’ reports (Olowokure and Nnadi, 1989b, p.31).

Despite the above noises from this new body of shareholders, the real pressure for the establishment of audit committees came from the submission of a Nigerian accountancy body to the Nigeria Law Reform Commission in 1988).7 In otherwords, the origin of the Audit committees in Nigeria can be traced to the review of the Nigerian Companies Act in 1988. This was in compliance with the Nigerian Law Reform Act of 1979 which require the Nigerian Law Reform Commission to periodically review Nigerian laws (Section 5). According to the Report of the Reform of the Nigerian Company Law, the idea of enshrining audit committees in our laws was mooted in a memorandum submitted by an accountancy body. The body was essentially concerned that auditor’s report which were usually brief, as prescribed by the then existing company law did not usually deal with matters crucial to the future of the company which ought to be brought to the attention of shareholders.8 As a remedy, it was thus suggested that auditors be allowed to deal directly with the chairman or other non executive members of the board on such matters. The same accountancy body further suggested that public companies should be made to appoint audit committees made up of representatives of directors and shareholders. The auditors should then be made to report to such audit committees on the audit of the accounts in addition to reporting to the shareholders at Annual General Meetings. The need for an additional independent layer of persons to oversee the auditing process and act as an intermediary between auditors and management was explained thus:

The present practice whereby the auditor deals with the executive members of the Board who are part of management and report on management matters to the same members of the Board has proved unsatisfactory and with many pitfalls. Evidence abounds where auditors report weaknesses or malpractices by members of the board … [to] the same board (Nigerian Law Reform Commission, 1991, p.212).

There was thus emphasis, from inception that audit committees be manned by non executive directors who are supposed to be sufficiently independent in order to meaningfully act as arbiter between the external auditors and the management (Cotter and Silvester, 2003, p.213; McMullen and Raghunandan, 1996, p.80). It was essentially on the basis of the above recommendation that the Law Reform Commission recommended that provisions be made for the appointment of an audit committee in


120

every public company in the emergent 1991 companies act. The Law and Practice of Audit Committees in Nigeria As a direct consequence of the work of the Nigerian Law Reform Commission, the Companies and Allied Matters Act was promulgated in 1990. In line with the recommendations of the Commission, the Law provided for establishment of an audit committee by each public company in Nigeria. Specifically, the audit committee is charged with the following responsibilities: to ascertain whether the accounting and reporting policies of the Company are in accordance with legal requirements and agreed ethical practices; to review the scope and planning of audit requirements; review the findings on management matters in conjunction with external auditor and departmental responses thereon; keep under review the effectiveness of the Company’s system of accounting and internal control; make recommendations to the board in regard to the appointment, removal and remuneration of the external auditors of the company and; authorize the internal auditor to carry out investigations into any activity of the company which may be of interest or concern to the committee (Section 359 (6)).

Section 259(4) further requires the audit committee to examine the auditors report and make recommendations thereon to the annual general meeting as it may deem fit. It is the above provision that appears to be the basis of an audit committee report which is found on the accounts of all public companies in Nigeria. A standard Audit Committee Report states thus:

In compliance with Section 359(6) of the Companies and Allied Matters Act 1990, we have (a) Reviewed the scope and planning of the audit requirements (b) Reviewed the external Auditors’ Memorandum of Recommendations on Accounting Policies and internal controls together with management responses (c) Ascertained that the accounting and reporting polices of the Company for the year ended 30th June, 2004 are in accordance with legal requirements and agreed ethical practices. In our opinion, the scope and planning of the audit for the year ended 30th June 2004 were adequate and the Management Responses to the auditors’ findings were satisfactory (Guinness Nigeria PLC, 2004, p. 27).

Such reports raise a number of concerns. First, it raises a fundamental legal question. It could for instance be legitimately asked if such reports can be relied upon by shareholders and investors and if so whether audit committee members can be held liable for the opinions so expressed in such reports? This

question is particularly pertinent given the fact that the Companies and Allied Matters Act of 1990 imposed civil liability on external auditors whose reports and procedures are reviewed by the audit committee, should they be found to be negligent in conduct of their duty (Section 368). Despite the above, the CAMA does not explicitly extend such liability to audit committee members. This fundamentally questions the very essence of such audit committees and the utility value of their reports.

Based on the above, it is not surprising that the 1999 Blue Ribbon Committee recommendation, aimed at improving the effectiveness of audit committees in the United States, that audit committees’ report to shareholders should specifically describe the procedures it performed and the conclusion that the financial statements are in accordance with the Generally Accepted Accounting Principles (GAAP) in all material respects was opposed. One such opponent was KPMG which explicitly stated that:

We do not believe, inspite of the recommended financial literacy requirements, that audit committee members, in accordance with their oversight role, should be asked to conclude that the financial statements are presented in accordance with GAAP. Management, as preparers of the financial statements, and auditors, as professionally trained accountants, are in a position to make such statements…. [W]e believe that SEC should require disclosure of audit committee responsibilities and activities. Extending these disclosures to reporting on the financial statements not only would discourage audit committee membership, it also will be inappropriate given the oversight role of the committee … We expect the SEC to adopt new rules to require an audit committee report to shareholders. The SEC informally has proposed to modify the recommendation to eliminate the reference to GAAP and instead substitute a reporting on “accurate, full and fair disclosure.” It strikes us that the new proposal is potentially worse than the original one in that GAAP at least provides a context in which to make a judgment (KPMG, 1999, p.4).

In Nigeria, the matter is further complicated by the fact that the 1990 CAMA specifies no qualification for audit committee members. The implication is that persons who have little understanding of financial reports could actually be elected into such an important committee. Given the fact that audit committees can only be as good as its members, such appointment turns the entire concept


121

of audit committees on its head.9 It is in recognition of such lapses in the law that some regulatory agencies, in their codes of best practices, which are not enforceable in law, have advised on some basic qualifications for audit committee membership. For instance, the Code of Corporate Governance in Nigeria, coauthored by the Securities and Exchange Commission and the Corporate Affairs Commission recommends that “members of audit committees should be able to read and understand basic financial statements”(2003, p.13). The Code of Corporate Governance for Banks recently released by the Central Bank of Nigeria also recommends that “some of them [audit committee members] should be knowledgeable in internal control processes” (Central Bank of Nigeria, 2006, p.19).

Despite these improvements, the above recommendations, even if enshrined in the statute does not go far enough. Given the technical nature of their assignment all members of this committee should be knowledgeable enough in accounting in order to be effective and make their opinion credible. Along these lines, it has been asserted that:

In view of the complex accounting and auditing issues faced by audit committees, it has also been recommended that committee members should have some degree of financial knowledge… The Public Oversight Board… suggests that the effectiveness of the audit committee is primarily affected by the expertise of its members in the areas of accounting, financial reporting, internal controls and auditing. The BRC [Blue Ribbon Committee] … recommends that an effective audit committee should comprise at least three members all of whom should be financially literate… and at least one of whom should have accounting and management expertise. This is defined as past employment experience in accounting or finance, a professional certification in accounting or comparable experience... Such expertise is regarded as important if the audit committee is to effectively carry out its role of overseeing the work of both external and internal auditors.10

Another provision of the 1990 Companies and Allied Matters Act with respect to audit committees relates to the composition of its membership. Section 359(4) specifically states that such a committee shall consist of an equal number of directors and representatives of shareholders of the company subject to maximum number of six members. This section further stipulates that members of such audit committees shall be subject to reelection annually and shall not be entitled to remuneration.

There are a number of controversies surrounding the above provisions. Take for instance the issue of composition of membership of the audit committees. Right from the conception of this idea at the law reform stage, there was opposition to the composition. According to the Report on the Reform of the Nigerian Company Law:

We note some opposition at the workshop to the representation of shareholders on the committee. We think that such a representation is an important investor- protection devise which should be encouraged for the sake of better understanding between the management and the investors (Nigeria Law Reform Commission, 1991, p.212).

Such opposition is not surprising. The convention in most parts of the world is for audit committees to be seen as a sub committee of the board whose members are independent directors of the board.11 The inclusion of non board members in the committee is therefore unusual. Despite this, the inclusion of non board members in audit committees is a welcome development. This is because it has rightly been argued that the idea that some directors could be truly independent could actually be a ruse. Along these lines, it has been argued that:

The expectation that certain directors will be semi-detached, orient themselves to interests outside the board, and actively safeguard the shareholder, may be somewhat optimistic. It underestimates the ability of boards to reproduce themselves in their own images electing people like themselves, and the incorporation and partial weakening of independence that follows from socialization into a powerful boardroom culture.12

The above view is even more pertinent in a developing country like Nigeria with underdeveloped capital markets. In several cases, non executive directors of publicly quoted companies are substantial shareholders in such companies. Owning large shareholdings in a company no doubt has the potentials of impairing the performance of audit committee members (Carcello and Neal, 2003a, p.96). Audit committee members that have strong economic ties to the company are likely to view financial reporting issues from a perspective similar to that of management. Such audit committee members would, for instance, like management prefer that their companies going concern problems are not discussed in the pages of the accounts (Caecello and Neal, 2003b, p.291). Under such circumstances, non board members, who do not have substantial financial investments in the company, are likely to be more useful in ensuring that audit committees are effective.


122

Despite the above endorsement, the method of electing these outside members and the duration of their tenure give reasons for concern. The requirement that such members be elected yearly at annual general meetings essentially means there is no guarantee of tenure for such outside members. It is for instance possible for such outside members to be replaced yearly during the Annual General Meetings. In this kind of scenario, the most experienced members of the Committee will therefore be the board nominees who are rarely changed. Although the law requires that audit committee members be elected at annual general meetings, the practice is that boards usually agree on its nominees before the annual general meetings. The situation for shareholder members of the audit committees is further worsened by the fact that boards may indeed be in a strong position to influence those who are elected to the Committee. Along these lines, it has been argued that:

Although Section… [359] Sub-section 3 of the CAMA provides for the appointment of an equal number of directors and shareholders as members of an audit committee of a public company but it seems the director members are the one determining the activities of the committees even when a shareholder is chairman of the committee. Instead of trying to see how to better the lot of the shareholders what you see as major attributes of the shareholders’ members of the committee is how to perpetrate themselves in office as member of the committee. The annual election-reelection of the shareholders members of the committee has not only polarized their groups but has resulted in situation in which some members have become rubber stamp of the management – a matter of you rub my palm and I rub your back.13

Under such circumstances persons who are likely to be elected into audit committees will be men of little substance and integrity who are willing to sacrifice shareholder goals for personal gains. The fact that CAMA does not specify any minimum qualification or experience requirement for audit committee members ensures the availability of a steady army of volunteer shareholders who are willing to deal with the board and compromise the interest of shareholders.14 This position is further reinforced by the absence of any obvious liability on the part of the audit committee members. The result is that audit committee practices in Nigeria have, in the main become a ritual. The authors are unaware of any audit committee report in Nigeria that has expressed reservations about the accounting and auditing processes of any quoted company in Nigeria since the codification of the audit committee requirement in Nigerian laws in 1990.15

Despite the above analysis, there is need to define clearly the tenure of both board and non board members of the audit committee. This will at the very least promote certainty and allow for the development of constructive change strategy that will enable the injection of new ideas into the audit committee process. Furthermore, recruitments and exit from audit committees could be made more systematic in order to ensure continuity. Based on the above, there is need to strike a balance between continuity and rotation. Along these lines, it has been suggested that:

In theory, rotation of membership is desirable on two main counts, first to strengthen the independence of the committee and second to spread the responsibility and experience of audit committee work among as many directors as possible. The requirement to rotate membership should not however be allowed to interfere with the committee’s effectiveness. It will take time for an audit committee to learn its job properly and each member will have more to contribute once he or she has the experience of several years service. A suitable compromise might be to rotate both chairmanship and membership every three years on a staggered basis, but to allow the chairman and individual members to be reappointed on expiry of their terms of office if the board believes that this additional continuity will strengthen the audit committee’s effectiveness (Buckley, 1979, p.31).

In addition to the above, we recommend that no member of the audit committee shall serve for more than two terms consecutively. This will, at the very least help prevent possible collusion that familiarity between executive directors and audit committee members may bring. Given the ranking of Nigeria as one of the most corrupt countries in the world, this constitutes a real danger. In fact, a worrying innovation in the Company and Allied Matters Act of 1990 is Section 287 which permits directors of companies to receive unsolicited gifts as ex-post gratification from persons who have had dealings with the company. Such gifts should however be declared before the board and recorded in the minute books of the directors. The problem here, however, is that for a company which can be taken as a going concern and dealing in a particular line of business, ex-post gratification, even if unsolicited, could well amount to an ex-ante bribe for a future contract. Furthermore, it can be argued that directors are likely to be more favorably disposed, with regard to future dealings, towards companies that come to say “thank you” without being solicited than those companies that rightly believe that directors are fiduciary officers of the company whose judgment should not be clouded by such gifts. The requirement that recipient board members disclose such gifts in minute books also makes little sense. Most board decisions need the approval of the entire board. Under such circumstances, it mat not be illogical for the benefiting company to gratify the entire board. In such a case, disclosure among fellow directors also amount to secret profits (Uche, 2004, pp.69-70). The


123

fact that minutes of board meetings are not normally available to shareholders further reinforces this position.

In fact, such a provision is a threat to the independence of both the board and audit committee members. It is for instance possible for auditors to regularly say “thank you” to both audit committee members and if need be the members of the entire board. Such an action will no doubt negatively impact on the judgment of audit committee member in the conduct of their oversight audit and accounting function. Specifically, it could impair their objectivity especially in their statutory responsibility of making recommendations to the board in regard to the appointment, removal and remuneration of the external auditors of the company.

Another area of concern is the requirement that audit committee members should not be remunerated in Nigeria. This is rather puzzling. At a very basic level, the same Companies Act allows the remuneration of Board members. If an audit committee is a sub committee of the board as is the convention in most parts of the world, it can meaningfully be argued that sub committee work should not attract additional compensation. Assuming this is the thought process of the promulgators of the Act, then it makes sense not to remunerate the representatives of the Board who are on the audit committee. The same however cannot be said for the representatives of the shareholders. To suggest that they should employ their expertise and time towards a venture mainly aimed at enhancing the outlook, credibility and performance of business enterprises motivated mainly by profits is absurd. A likely economic consequence of the above rule is that there will be a shortage of supply of competent audit committee members. Admittedly, the evidence in Nigeria suggests otherwise. A speculative explanation for this variance is that management, despite the explicit regulation on the issue use covert ways to remunerate audit committee members. Another possible explanation is that there are other intrinsic benefits derived from being a member of an audit committee. Either way, the independence of the audit committee members is compromised. Whatever the reason may be, the codification of non remuneration of audit committee members is an abnormally that should be corrected. On the basis for remunerating audit committee members, it has been suggested that:

In addition to the remuneration paid to all non executive directors… each company should consider the further remuneration that should be paid to members of the audit committee to recompense them for the additional responsibilities of membership. Consideration should be given to the time members are required to give to the audit committee business, the skills they bring to bear, and the onerous duties they take on, as well as the value

of their work to the company. The level of remuneration paid to the members of the audit committee should take into account the level of fees paid to other members of the board. The chairman’s responsibilities and time demands will generally be heavier than the other members of the audit committee and this should be reflected in his or her remuneration (Financial Reporting Council, 2003, p.8).

Appropriate remuneration of audit committee members will at the very least help stimulate the interest of persons with the required qualifications, experience and integrity in such audit committees.

Conclusion

From the above analysis, it can be concluded that the codification of the requirement for the establishment of audit committees as an additional layer of corporate accountability control in Nigeria is a welcome development. This no doubt has the potentials of improving corporate governance and promoting corporate accountability in publicly quoted companies in the country. Mere codification of the audit committee requirement is however not enough to ensure the attainment of the objective of the audit committee goal (Rhagunandan, 1994, p.318). In Nigeria, for instance, there is still a wide variance between the above potentials and the actual practice of audit committees. Despite the codification of the audit committee requirement for publicly quoted companies in Nigeria in 1990, there is little evidence that this has made any meaningful impact on corporate accountability in Nigeria. It has for instance been argued that:

In fact, in most cases, the committees have become parts and parcel of management. Management report, through which external auditors communicate their findings to the shareholders are usually not well treated by the committee, leading to a situation in which those directing the affairs of the company continue business as usual. Until the bubble burst in Lever Brothers Nigeria Plc, now Unilever Nigeria Plc, in 1997 over a N1.2 billion overstated profit of the past, the audit committee was giving its management clean bills. Also the audit committee of National Oil and Chemical Plc (NOLCHEM) never indicted the management of the company for any wrong-doing, even when it was almost run aground, until a new core investor went there to reposition it. Even in companies whose accounts have been qualified by external auditors, the kind of audit


124

committee’s report you see are the kind that tend to toe the line of the management- that is, the committees are usually satisfied with management’s responses to auditors’ management report. Also, the audit committees of some companies which have been declared as nonperforming in the stock market never deemed it fit to raise any alarm before the companies reached such level.16

For audit committees to become more useful in the Nigerian context there is need for changes to be made in both its law and practice. Some areas of concern include the need to: determine and codify the qualification for membership of the committee given its technical nature; allow appropriate remuneration for committee members, and; the determination of appropriate membership tenure for such committees.

References

1. Al-Twaijry, A, Brierley, J. and Gwilliam, D. (2002),

An Examination of the Role of Audit Committees in Saudi Arabian Corporate Sector, Corporate

Governance, Volume 10, Number 4, pp.288-297. 2. Archambeault, D and DeZoort, F.T. (2001), Auditor

Opinion Shopping and the Audit Committee: An Analysis of Suspicious Auditor Switches, International

Journal of Auditing, Volume 5, .pp 33-52 3. Armitage, J. and Bradley, R (1994), Audit Committees,

The Nigerian Accountant, January/ March, pp.3-5. 4. Buckley, R (1979), Audit Committees: Their Role in

UK Companies (London, Insdtitue of Chartered Accountants of England and Wales).

5. Carcello, J.V. and Neal, T.L. (2003a), Audit Committee Characteristics and Auditor Dismissals following “New” Going Concern Reports, The

Accounting Review, Volume 78, Number 1, pp.95-117. 6. Carcello, J.V. and Neal, T.L. (2003b), Audit

Committee Independence and Disclosure: Choice for Financially Distressed Firms, Corporate Governance, Volume 11, Number 4, pp.289-299.

7. Carson, E (2002), Factors Associated with the Development of Board Sub-Committees, Corporate

Governance Volume 10, Number 1, pp4-18 8. Central Bank of Nigeria (2006), Code of Corporate

Governance for Banks in Nigeria Post Consolidation (Abuja, Central Bank of Nigeria)

9. Cotter, J. and Silvester, M (2003), Board and Monitoring Committee Independence, Abacus, Volume 39, Number 2, pp.211-232.

10. Collier, P (1992), Audit Committees in Large UK

Companies (London, Institute of Chartered Accountants in England and Wales)

11. Companies Decree of 1968 (Lagos, Government Press).

12. Companies and Allied Matters Act, CAP 59 of the Laws of the Federation of Nigeria 1990, Volume III (Lagos, Federal Ministry of Justice).

13. Deli, D. and Gillan, S. (2000), On the Demand for Independent and Active Audit Committees, Journal of

Corporate Finance, Volume 6, pp.427-445.

14. Dezoort, F. T. (1997), An Investigation of Audit Committees’ Oversight Responsibilities, Abacus, Volume.23, Number.2, pp. 208-227.

15. DeZoort, F.T. (1998), An Analysis of Experience Effects on Audit Committee Members’ Oversight Judgment, Accounting, Organizations and Society, Volume 23, Number 1, pp.1-21.

16. Financial Reporting Council (2003), Audit Committees

Combined Code Guidance: A Report and Proposed

Guidance by an FRC Appointed Group Chaired by Sir

Robert Smith (London, Financial Reporting Council). 17. Goodwin, J (2003), The Relationship Between the

Audit Committee and The Internal Audit Function: Evidence from Australia and New Zealand, International Journal of Auditing, Volume 7, pp. 263-279

18. Guiness Nigeria Plc. (2004), Annual Report and

Financial Statements.

19. The Nigerian Law Reform Commission Act [CAP 313 of 1990], Laws of the Federation of Nigeria, Volume XVIII (Lagos, Federal Ministry of Justice)

20. Idornigie, P.O. (1999), A Critical Appraisal of The Establishment of Audit Committee for Public Companies, Nigerian Law and Practice Journal Volume 3, Number 2, pp 112-119.

21. Klein, A. (2002a), Audit Committee, Board of Director Characteristics and Earnings Management, Journal of

Accounting and Economics, Volume 33, pp.375-400. 22. Klein, A. (2002b), Economic Determinants of Audit

Committee Independence, The Accounting Review, Volume 77, Number 2, pp. 435-452.

23. KPMG (1999), 1999 Audit Committee Summer Update

24. KPMG (2003), Audit Committee Roundtable High-

lights, (Audit Committee Institute, Spring). 25. Lee, T and Stone, M (1997), Economic Agency and

Audit Committees: Responsibilities and Membership Composition, International Journal of Auditing, Volume 1, Number 2, pp. 97-116.

26. Lennox, C (2003), Opinion Shopping And The Role of

Audit Committees when Audit Firms Are Dismissed:

The US Experience (Edinburgh, The Institute of Chartered Accountants of Scotland).

27. Mautz, R. K and Neumann, F.L (1970), The Effective Corporate Audit Committee, Harvard Business

Review, November-December 1970, pp. 57-65 28. McMullen D. A and Reghunadan, K. (1996),

Enhancing Audit Committee Effectiveness,. Journal of

Accountancy, August, pp.79-81 29. Nigeria Law Reform Commission (1991), Report on

the Reform of Nigerian Company Law (Lagos, Nigerian Law Reform Commission)

30. Nnadi, G.S (1999), Making Audit Committees Effective, ICAN News, October/ December, Volume 4, Number.4, pp. 6-9

31. Olowokure, T.O. and Nnadi, G.S (1989a), A Case for Audit Committees in Nigeria, The Nigerian Accountant (January/ March. pp18-21)

32. Olowokure, T.O. and Nnadi, G.S (1989b), A Case for Audit Committees in Nigeria, The Nigerian Accountant

(April/June1989. pp29-32) 33. Okaro, S. C. (2001), Empowering the Audit Committee

for more Effective Role in Corporate Governance, Nigeria, Journal of the Management Sciences

[Awka].(pp.154-158) 34. Oyejide, T. and Soyibo, A. (2001), Corporate

Governance in Nigeria, Paper Presented at the Conference on Corporate Governance (Accra, Ghana, 29-30 January 2001).


125

35. Palmer, R. (1977), Audit Committees- Are they Effective? An Auditors View, The Journal of

Accountancy, September, pp.76-79 36. Raghunandan, K. and McHugh, J. (1994), Internal

Audiors’ Independence and Interaction with Audit Committees: Challenges of Form and Substance, Advances in Accounting, Volume 12, 313-329.

37. Raghunandan, K, Reid, W. and Rama, D. (2001), Audit Committee Composition, “Grey Directors” and Interaction with Internal Auditing, Accounting

Horizons, Volume 15, Number 2, pp.105-118). 38. Sasegbon, D (1991), Nigerian Companies and Allied

Matters Law and Practice, Volume 1 (Lagos, DSc Publishing Limited).

39. Securities and Exchange Commission and Corporate Affairs Commission (2003), Report of the Committee

on Corporate Governance of Public Companies in

Nigeria (Lagos, Securities and Exchange Commission and Corporate Affairs Commission).

40. Spira L.F. (1998), An Evolutionary Perspective of Audit Committee Effectiveness, Corporate

Governance, Volume.6, Number 1, pp.29-38. 41. Spira L.F. (2002), The Audit Committee: Performing

Corporate Governance (Massachusetts, Kluwer Academic Publishers).

42. Spira, L. (2003), Audit Committee: Begging the Question, Corporate Governance, Volume 11, Number 3, pp.180-188.

43. ThisDay Newspaper (Various Issues) 44. Uche, C. (2004), Ethics in Nigerian Banking, Journal

of Money Laundering Control, Volume 8, Number 1, pp.66-74.

45. Williams, H. (1977), Audit Committees-the Public Sector’s View, The Journal of Accountancy, September, pp.71-74.

46. Xie, B., Davidson III, W. and DaDalt, P. (2003), Earnings Management and Corporate Governance: The Role of the Board and the Audit Committee, Journal of

Corporate Finance, Volume 9, pp.295-316.


126

РАЗДЕЛ 3 КОРПОРАТИВНОЕ УПРАВЛЕНИЕ

В ЮЖНО-АФРИКАНСКОЙ РЕСПУБЛИКЕ

SECTION 3 CORPORATE GOVERNANCE PRACTICES: SOUTH AFRICA

DO PORTFOLIO MANAGERS IN SOUTH AFRICA CONSIDER HUMAN

BEHAVIOUR ISSUES WHEN MAKING INVESTMENT DECISIONS OR

ADVISING CLIENTS?

Strydom JSG*, Van Rooyen JH**

Abstract The efficient market hypothesis is based on the assumption that individuals act rationally, processing all available information in their decision-making process. Prices therefore reflect the appropriate risk and return. However, research conducted regarding the ways that investors arrive at decisions when faced with uncertainty, has revealed that this is in fact not always the case. People often make systematic errors, the so-called cognitive biases, which lead them to less rational behaviour than the traditional economic paradigm predicts. These cognitive biases have been found to be responsible for various irregular phenomena often observed in financial markets as(turbulence or, volatility, seasonable cycles, "bubbles", etc. Behavioural finance attempts to explain some of the changes in the financial markets that can not be explained by the efficient market hypothesis. This research reviews some results from the behavioural finance and other related literature. A survey was also done to determine whether the most prominent portfolio managers in South Africa are aware of behavioural finance issues/models and consider the influence of cognitive issues when making investment decisions or giving advice to clients. Keywords: Behavioural Finance, Herding, Overconfidence, Aversion, Risk * University of Stellenbosch, South Africa **University of Stellenbosch, South Africa

Introduction Behavioural finance is the paradigm where financial and capital markets are studied using models that are not as narrow as those based on the expected utility theory and arbitrage assumptions. Specifically, behavioural finance has cognitive psychology as a main building block. Cognitive refers to how people think. Much has been dealt with in literature documenting how people make errors in the way

they think. People are often overconfident, they put too much weight on recent experience, they tend to overreact or follow other players in the market. The American sub-prime issue may well be proof of many of the behaviours as it relates to cognitive issues, that we would expect of investors. The mere fact that so many hedge funds were affected by the problems created by it underlines how much herding is taking place in the financial markets and also how much overconfidence there may be amongst


127

managers. Part of the way these managers act or the decisions they take has to do not only with herding but greed. Their actions may very well not always be as rational as they would make them out to be.

Many anomalies have been observed in securities markets all around the world. The size effect and the value effect in stock markets are among the most common. Researchers who have been trying to reveal factors underlying some of the anomalies can be divided into at least two schools. One school consists of proponents of traditional finance theory. As shown in Fama and French (1993), they try to resolve seemingly irregular phenomena by generalizing equilibrium models. The most important feature of this school is their insistence that all of the players in the market are rational. Rationality means that they conform to the assumptions of expected utility theory, and that prices reflect all of the information (in terms of risk and return) available in the market (von Neumann and Morgenstein, 1947).

As a result, modern finance has as a building block the Efficient Market Hypothesis (EMH). The EMH is based on the premise that competition between investors seeking abnormal profits drives prices to their "correct" value. The EMH does not assume that markets can foresee the future, but it does assume that investors make unbiased forecasts of the future.

The other school consists of researchers opposed to the idea of fully rational market players. They insist that the level of investors' rationality is seriously limited. Simon (1955) called it bounded rationality, and was sceptical regarding the assumption made by expected utility theory that decision makers are completely rational. Simon proposed instead the construction of decision-making models that do not assume perfect rationality on the part of decision-makers. Building on Simon's work, Tversky and Kahneman (1974) introduced the idea of heuristics. This idea means that people tend to use rules of thumb when making a decision due to their lack of ability to process information rationally and I or to time pressures, i.e. they have to make many decisions in a limited time. A few years after they presented the idea of heuristics, Kahneman and Tversky (1979) proposed prospect theory, an alternative decision-making model to expected utility theory.

In the last twenty years several radical psychology-oriented theories made their appearance in the financial world, which have significantly changed the way we see pricing of securities. Those theories in many ways contradict the traditional financial framework, where security analysis is based solely on fundamental information. That is, information concerning the company, industry, sector or the economy as a whole. By incorporating "behavioural" ideas into the modelling framework, it will lead to a more realistic and accurate representation of financial phenomena.

The purpose of this research is to investigate the extent to which investor behaviour and their investment decision-making processes are influenced by one or more irrational factors. An attempt is made to explain these apparent market irrationalities by using a fundamentally psychological approach, also referred to as behavioural finance. This, in effect, challenges the traditional notion that people are thought of as rational economic agents that do not engage in financial decisions based on emotion.

Some light could be shed on the question of whether a psychological approach could be used to help understand and explain eccentric and irrational investor behaviour within capital and financial markets by focusing on the following topics for discussion throughout this research: • To what degree are institutional investors in

South Africa aware and subject to certain cognitive biases when evaluating investments for their clients?

• What are some of the most important and predictable behavioural or cognitive biases that we know of?

• How do these biases influence and characterise financial and economic agent's behaviour?

Methodology

Most, of the information to be used in this research will be qualitative in nature. A questionnaire was sent to some of the largest and most influential brokerage houses and portfolio managers in South Africa.

Due to the similarity of the problems/issues that many of the asset managers/portfolio managers/brokers may face when doing business in today’s investment environment, it would not be unreasonable to claim this research to be a fair representation of what is experienced by managers in South Africa as a whole. Some differences in emphasis may, however, occur in different regions due to the differences in the wealth of clients. However, the underlying issues researched should not be different. Brief overview of literature Few mainstream finance theorists argue that individuals cannot behave in an irrational way and that the homo economicus is a gross simplification that does not describe any human being. At the same time, economists normally maintain that the functioning of markets may be well described and predicted "as if' agents were all homo economicus. The analysis of the functioning of markets is the core task of economics, and economics does not deal with the psychology of economic agents as an objective per se, but only with the market implications of it (Mas-Colell, 1999).

In essence, the debate around behavioural and


128

mainstream finance revolves around the "as if' hypothesis. Most of the debate concerns whether prices set on speculative and highly competitive and developed markets are "rational" or whether a pricing error arises. Both behavioural and mainstream finance theorists agree that studying these markets is important. The fundamental problem, however, is that no agreement can be reached on the very definition of "rationality". Rubenstein (2000) goes on to describe various types of market rationality. He distinguishes between markets that are maximally rational, rational, and minimally rational. Markets are defined as "maximally rational" if all investors are rational. If markets were maximally rational, investors would probably trade relatively infrequently and make extensive use of index funds.

The concept of rationality maintained by mainstream finance theorists is normally ons of attempting to beat the market. Initially, the publication of the paper by De Bondt and Thaler (1985) - according to whom the stock market displays a systematic tendency to overreact to any type of news - seemed to deal a blow to market rationality as proposed by the traditional, mainstream financial theories everyone were so familiar with. However, in subsequent years several instances of market under-reaction were also detected. This has led Fama (1998) to claim that over- and under-reaction anomalies are simply due to chance, and that market efficiency prevails on average.

Moreover, Fama (1998) stressed that most anomalies are fragile and do not withstand a closer scrutiny. Today, there seems to be almost a consensus that the market is rational most of the time. The most solid proof of this is that portfolio managers and in general active investment strategies, do not outperform passive investment strategies, especially when transaction costs are considered (Malkiel, 1995). In this beat-the-market sense, mainstream finance seems to have resisted the "attack" by behaviouralists (as Thaler, 1999b).

It is important to stress, however, that market irrationality in the beat-the-market sense is not necessarily inconsistent with the idea that anomalies are a persistent and systematic behaviour of individuals and may lead to a pricing bias. It simply signals that it is not easy to make money out of these anomalies, for example because there are limits to arbitrage activity (Schleifer and Vishny, 1997). Mullainathan and Thaler (2000) and Barberis and Thaler (2001) pointed out that it is impossible to arbitrage away many instances of "irrationality", simply because there is no speculative market on such matters or because arbitrage is risky. Thus, a

pricing bias term might be impossible to arbitrage away, and the existence of a pricing bias is fully compatible with rational expectations and random walk behaviour of asset prices.

Most advocates of behavioural finance contend that the beat-the-market definition of market rationality is too narrow and not relevant from a welfare perspective (Barberis and Thaler, 2001). The ultimate function of the financial market is not to allow agents to speculate over future movements in prices, but rather (over time) to allow them to allocate consumption in their lifetime in an optimal manner and to allocate funds to the most attractive or productive investment opportunities. There is very little research on whether behavioural biases lead to misallocations of capital and to lower economic growth in the longer run, despite the obvious importance of this matter. Misallocation of capital due to biases does in all likeliness take place. It is, however, not in the scope of this research.

The controversy about whether markets are rational will still be with us in the years to come. This is unlikely because, as Fama (1998) pointed out, market efficiency is per se un-testable. In fact, testing the hypothesis that the market is efficient requires a model of expected returns, which is actually tested together with the hypothesis. Only the evidence that it is possible to systematically beat the market would be a certain way to discredit the hypothesis of market efficiency. Thus far, behavioural finance has failed to provide such evidence.

Whether the alleged influence of behavioural biases on financial markets calls for a policy changes or not is addressed by Daniel et al (2002). According to these authors, governments are likely to be affected by behavioural biases as well, with the difference that they would not be subject to the powerful disciplinary force of competition. Thus, their involvement in setting market prices would probably be counterproductive. At the same time, governments could make investors more aware of their psychological biases and of the incentives that others have to exploit them, creating some room for policy intervention in terms of reporting rules and disclosure. Sources of Bias and Behavioural Models Figure 1 represents the structure of behavioural finance (Toshino, 2004). This particular structure considers that there are several sources of bias underlying market anomalies. Aspects of the model is dealt with in the sections that follow.


129

Sources of Biases Behavioural Models

Bounded Rationality 1. Imprecision of memory 2. Selective Bias of Information 3. Imprecision of judgment

Heuristics 1. Representativeness 2. Availability 3. Anchoring

Time constraints

Prospect Theory 1. Value function 2. Weighting function

Emotional Factors 1. Overconfidence 2. Aversion to Loss or Regret

Mental Accounting

Social Factors 1. Exposure to Market Sentiment 2. Herding 3. Avoidance of Cognitive

Dissonance

Market Anomalies

Figure 1. Structure of behavioural finance

Source: Toshino (2004)

Sources of Bias Firstly, as already described, the concept of bounded rationality was first proposed by Simon (1955). This concept implies that human behaviour is not always rational, as assumed by the expected utility theory, and the EMH, which is the basis for traditional finance theory. Human judgment, such as selection among several alternatives, is generally made based on past memory and newly collected information (Figure 2). Simon (1955) suggested that human behavior could be subject to biases at any of three stages in the decision-making process: recalling memories; selecting information; making judgments.

The second source of bias is time constraints. Human beings are very busy and have to continuously make various decisions. Consequently, they cannot afford to spend a lot of time trying to make optimal decisions. Thirdly, emotional factors can be a source of bias in human judgment. In

particular, overconfidence and regret aversion are included in biases, which could lead to market anomalies. Overconfidence suggests that investors overestimate their ability to predict market events. Overconfidence may be increased with market experience. Because of overconfidence, investors often take risks without receiving appropriate returns. Another effect of overconfidence is overtrading, which can lead to poor investment decisions and excessive transaction costs.

Among the emotions that determine the individual investor's perception of risk is an aversion to losses. The idea that investors are not risk-averse but loss-averse is one of the main tenets of behavioural finance. While this distinction may seem trivial, it implies that investors will increase their risk (uncertainty), in an attempt to avoid the probability of loss.

Figure 2. The human decision-making process

Source: Toshino (2004)


130

Fourth is social factors (see the sources of bias shown in Figure 1). Human beings tend to create a variety of societies, and to act as members of each society. Meanwhile, they are subject to some kinds of social bias such as exposure to market sentiment, herding, and avoidance of cognitive dissonance.

Market sentiment means the general atmosphere of bullishness or bearishness in the market. For example, when market sentiment is very bullish, investors would like to purchase securities even though they are mostly overvalued. The South African stock markets experienced this situation for a number of years and is now in a downward phase.

On the other hand, herding is the human tendency to act similarly, following other people's behaviour. To act differently when other people are acting uniformly is mentally very difficult. This is especially true of the American sub-prime issue. Although this research is not about this aspect, it needs to be mentioned that many hedge funds got involved in this market through securitised loans, probably because many others were doing it. Managers probably did not really think about the risk of negative gearing should defaults take place on a large scale.

Lastly, the avoidance of cognitive dissonance refers to the human tendency to try to be consistent in one's behaviour. Having once expressed a positive or negative opinion about something, people find it difficult to change their position even though they have discovered reasons for doing so. This is closely related to the perseverance belief, which indicates that people are unlikely to change their opinions even when new information becomes available. At least two effects appear to be at work. First, people are reluctant to search for evidence that contradicts their beliefs. Second, even if they find such evidence, they treat it with disproportionate scepticism.

For example, if an analyst has written a report with a buy recommendation for a certain stock, he or she may feel some mental pressure not to express a negative view about the stock even when its issuing firm has announced some unfavourable news. Behavioural models Some researchers have presented behavioural models reflecting the sources of bias described so far that could be more directly associated to market anomalies. Heuristics is one of the earliest behavioural models presented in Tversky and Kahneman (1974), which is considered the decisive work in the field of behavioural finance. Heuristics refers to the human tendency to try to intuitively solve problems with limited information by using rules of thumb, even when people could derive better answers with more time and information. Although this kind of decision-making rule is generally regarded as an effective way to deal with daily incumbencies, it can lead to systematically biased decisions. Tversky and Kahneman (1974) presented

three types of heuristics: representativeness; availability; anchoring.

Representativeness heuristics is the human tendency to judge person A as belonging to a group X if A has any representative feature of that particular group. People consider that which they find easier to imagine, more probable.

Availability heuristics is the human tendency to consider that more familiar things happen more often. Accidents and homicides are generally considered to happen quite often because they receive a lot of coverage in the mass media. People usually take care not to get involved in such affairs since they are a familiar risk. At the same time, people tend to disregard the risk of sicknesses such as diabetes because they are largely ignored by the news media.

Anchoring refers to the tendency to consider a random available number as a starting point for estimating the true value of an unknown matter. The random number can affect the resulting estimate. The effects of anchoring are pervasive and robust and are extremely difficult to ignore, even when people are aware of the effect and aware that the anchor is ridiculous. One of the most common effects of anchoring is under reaction, where people fail to react to new information quickly enough.

On the other hand, Kahneman and Tversky (1979) presented prospect theory that was intended to be an alternative model to expected utility theory. Before describing prospect theory however, it is useful to briefly comment on expected utility theory.

This was originally developed by von Neumann and Morgenstein in 1947, and describes how people behave if they follow certain requirements of rational decision-making. The most important concept behind expected utility theory was developed in 1738 by Bernoulli and is that of declining marginal utility. This implies that people will be risk averse. Specifically, Bernoulli argued that the following graph could represent the value of money (Kahneman and Tversky, 1979):

Prospect theory was developed by two Israeli psychologists, Daniel Kahneman and Amos Tversky. This model was based on five experimentally established aspects of human nature (Kahneman and Tversky, (1979): 1. People tend to evaluate alternatives not by their

ultimate asset value but by how far the alternatives depart from a reference value.

2. People tend to be risk averse when making a profit, but reckless when suffering from a loss.

3. People tend to weigh a loss of a certain quantity more than a gain of the same quantity.

4. People tend to value 100% certain things much higher than merely probable things.

5. People tend to overvalue the chances of a scenario succeeding when the probability is very small.


131

Graph 1. Declining marginal utility

Source: Kahneman and Tversky (1979)

The fundamental issue underlying prospect theory is best illustrated using the results of one of their experiments: They asked people to choose between two alternatives: 1. An 80% chance of winning $4 000 and a 20%

chance of winning nothing 2. A 100% chance of receiving $3 000 Even though alternative 1 has a higher expected outcome of $3 200, 80% of people chose alternative 2. These people were therefore risk averse as suggested by expected utility theory. Then Kahneman and Tversky offered the

following choices: 3. An 80% chance of losing $4,000 and a

20% chance of losing nothing 4. A 100% chance of losing $3,000

Even though the expected loss of $3,200 is bigger under alternative 3, 92% of people chose to gamble. When the choice involves losses people appear to become risk seeking and not risk averse.

Prospect theory provides an explanation for this asymmetry of how people make decisions. Prospect theory replaces the concept of decreasing marginal utility based on total wealth with a concept of value defined in terms of gains and losses relative to a reference point. This is illustrated in Graph 2 below. Unlike expected utility theory, prospect theory suggests that decisions like the ones above will depend on how a problem is presented or "framed".

Graph 2. Asymmetric value function of the Prospect Theory

Source: Kahneman and Tversky, 1979

If the initial position is viewed as a gain (the choice between alternatives 1 and 2) then the value function is concave, representing the risk aversion of the decision maker. If the initial position is defined such that an outcome is viewed as a loss, (the choice between alternatives 3 and 4) then the value function

will be convex, representing the risk-seeking nature of the decision maker (Kahneman and Tversky, 1979).

Finally, the idea of mental accounting was presented by Thaler (1985). It is the human tendency to set up a local account and try to get an optimal


132

value for each account.

Review with regard to South African institutional investors

In this research various important cognitive biases

are pointed out. For the sake of conducting this research, involving South African portfolio managers, the scope will include overconfidence and loss aversion only. By confining the research to these two anomalies only, it will be possible to obtain a clearer picture of the current situation in South Africa.

If too much information is gathered, the questionnaire will take too long to complete which may reduce the response rate substantially. Overconfidence and loss aversion were chosen as the two main factors that may materially affect the South African financial markets leading to a negative impact on investors. Attention is also paid to the extent to which investment professionals are aware of the behavioural finance issues and if aspects of this is ever considered when evaluating their investment decisions or dealing with clients.

Data sample

The survey was conducted in August through September 2007. Questions were grouped into three parts: the personal profile of the fund manager, some questions regarding their awareness regarding behavioural finance, and a few questions related to overconfidence and loss aversion. The questionnaire was sent to 19 prominent fund management companies within South Africa, all employing numerous individual fund managers, analysts, and other individuals involved with the management process of their assets.

Due to the fact that the sample was small, a response rate of at least 50% was sought. A response rate of 58% was achieved. Of the respondents, 91 % were male, and all held bachelor degrees (73%) or masters degrees (27%). They listed their current positions as follows: fund manager (54.5%), senior fund manager (27.3%), chief investment officer or

chief executive officer (18.2%). About eighty percent of the respondents had been working in the asset management environment for at least 6 years. Around 54.5% were at the age of 36 to 45 years, while 18% were under 35 and 27% were over 45 years old.

Findings

Keeping in mind the objective of determining the degree to which institutional investors in South Africa believe the Efficient Market Hypothesis to be a proper theory for understanding financial markets, and to what extent they are aware of the concept of behavioural finance, some simple questions regarding the matter were asked. Firstly, all the respondents indicated that they were familiar with the EMH, but only three respondents believed that South African markets were efficient at least to a degree (see Table 1). Out of these three respondents, all indicated that the weak form of the hypothesis was relevant for South Africa. Likewise, all respondents pointed out that they were familiar with the concept of behavioural finance, while all believe that it definitely does hold merit, even if at least to a degree. This was an interesting observation since the respondents indicating that they believed South African markets to be weak form efficient, also indicated that the behavioural finance paradigm definitely does hold merit. Perhaps the reason for this is that some individuals are not entirely convinced that the traditional financial framework may be so fundamentally flawed, but they certainly see the reasoning behind behavioural finance.

All respondents pointed out that they were aware of one or more biases caused by human behaviour and that they do in fact take them into account most of the time (if not always) when evaluating investment decisions for their clients. It may be important to note however, that the responsibility of educating the client with regard to these behavioural biases also fall on the fund manager. The biases apply equally to the client and the manager him-/herself.

Table 1. Results on awareness and consideration of behavioural biases

Question asked Responses %

Yes 11 100.00 Familiar with EMH

No 0 0.00

Yes 2 18.18

Sometimes 1 9.09 Believe SA markets to be efficient

No 8 72.73

Strong 0 0.00

Semi-strong 0 0.00 If yes, form of efficiency

Weak 3 27.27


133

Table 1 continued

Yes 11 100.00 Familiar with concept of BF

No 0 0.00

Yes 10 90.91

Some merit 1 9.09 Believe BF concept to hold merit

No 0 0.00

Yes 11 100.00 Aware of cognitive anomalies

No 0 0.00

Yes 9 81.82

Consider cognitive anomalies Mostly 2 18.18

No 0 0.00

Findings regarding overconfidence In order to establish the existence of overconfidence in the forecasts of institutional investors, they were required to give their one-month and one-year forecasts of the All Share Index (ALSI) and Dow Jones Industrial Average (DJIA). They were also asked to estimate a percentage change in this regard, given the current index levels as well as their 90% confidence range. The average lower bound returns, upper bound returns, and the respective medians were calculated to get an idea of the central values for every forecast.

The forecasts of South African institutional investors showed a clear positive predisposition. Although the lower bound and upper bound returns were similar in terms of absolute value for the one-month forecast of the DJIA, the average of the upper bound returns for the one-month forecast of the ALSI was larger by around 23% than that of its lower bound returns in terms of absolute value. The discrepancy was much larger for one-year return forecasts. The difference was more than 55% for the one-year return of the DJIA, while it was more than 113% for that of the ALSI.

These results were consistent with expectations as follows:

Firstly, institutional investors were generally

optimistic in their market forecasts. This occurrence is consistent with the notion that they are more sensitive to positive market news, or subject to a bias in selecting information, which is one of the aspects of bounded rationality.

Secondly, the over-/under confidence was more significant for the domestic market than for the foreign market. This result is consistent with the notion that investors tend to undervalue the risk of familiar investment products (availability heuristics). This is especially important if one considers where the South African economy will be going in the immediate future. Managers tend to overvalue the performance and underestimate the risk inherent in the markets.

Thirdly, over-/under confidence was much greater when the forecasting period was longer and there was greater uncertainty. However, it must be noted that a theoretical explanation for what appears to be a connection between forecasting period and the confidence in forecasts is still outstanding. It is reasonable to assume that investors may perhaps request a higher premium for more risky investments with a longer holding period (more uncertain).

Table 2. Results: confidence in forecasting

Lower bound Upper bound

Target of forecast

Median Mean Median Mean

Difference between lower and upper bound

All Share Index

One-month forecast

-6

-7.00

8

8.64

23.40%

One-year forecast -6 -8.73 16 18.64 113.50%

Dow Jones Industrial Average

One-month forecast -5 -5.09 5 5.18 1.70%

One-year forecast -9 -8.91 14 13.82 55.10%

The conclusion that can be drawn from the

above is that institutional investors were mostly confident about forecasting their own market and the

optimism was stronger for the domestic market. Also, when the forecasting period was longer they tended to be more confident. This kind of uniformity


134

may result from the fact that most institutional investors are doing business based on similar information, e.g. economicIpolitical news as well as market reports from analysts. As a result, they may tend to make similar market forecasts. This result is consistent with the herding concept.

Findings regarding loss aversion

In order to determine if institutional investors in South Africa are also subject to loss aversion, the following questions were posed:

"Please assume that you can make a bet with an even chance of making a profit or loss. If the loss is ten thousand rand, what would be the minimum profit that you would require in order to make a bet?"

"What would be the minimum profit if the loss were one million rand in the above question?"

Fewer than 20% of the respondents answered that they would make a bet even if it were a fair game, or the expected outcome of the bet were zero. The majority required a premium. The requirement was 1 to 2 times - they required that the profit be more than the loss, while less than twice the loss. There were even respondents who required a profit of more than 10 times the loss. This confirms the existence of investors with obvious loss aversion.

The means of the answers were 46.4 thousand rand for a loss of ten thousand rand and 7 million rand for a loss of one million rand, more than twenty fold for the larger loss and almost five times for the smaller expected loss.

In order to avoid the influence of these outliers, the medians were calculated, which turned out to be R18 000 return for a loss of ten thousand rand, and R3.5 million return for a loss of one million rand. These figures imply that institutional investors are loss averse, requiring gains of several times the bet when faced with a 50-50 chance of a loss.

The medians were calculated for different categories of loss averse respondents. The following results were obtained:

Firstly, loss aversion was stronger among females and respondents with spouses than among males and respondents without spouses, respectively. Secondly, older respondents displayed stronger aversion to loss than their younger counterparts. Thirdly, academic background and the type of portfolio under management were not related to the extent of loss aversion. Finally, loss aversion was stronger for senior fund managers than for junior ones, but weaker for CIOs and CEOs than for their subordinates.

Table 3. Results: loss aversion

Required gain

Ten thousand rand

One million rand

% %

<1 time 0 0.00 0 0.00

1 time 2 18.18 1 9.09

1 to 2 times 5 45.46 4 36.37

2 to 5 times 2 18.18 2 18.18

5 to 10 times 1 9.09 1 9.09

>10 times 1 9.09 3 27.27

Total 11 100.0 11 100.00

Mean R46 400 R6.97 million

Median R 18 000 R3.5 million

Table 4. Results: loss aversion by ascription

Personal details Number of responses Average bet for R10 000

loss Average bet for R1 million

loss Gender Male 10 R42 500 R6.5mil Female 1 R85 000 R12 mil

Marital status

Married 8 R49 000 R7.7 mil

Not married 3 R39 500 R5.1 mil


135

Table 4 continued

Age (years)

<35 2 R15 000 R1.35 mil 36 - 45 6 R19300 R5.9mil >45 3 R121 500 R12.9 mil Academic background Bachelor 8 R46 000 R8.5mil Masters 3 R47 500 R8.7mil Position held Fund manager 6 R37 800 R6.3mil Senior FM 3 R86 700 R12 mil CEO or cia 2 R11 750 R1.4 mil Financial asset Stocks 9 R46 700 R7.0mil Bonds 2 R44 750 R6.9mil

Although the scope of this study did not include all sources of bias and all behavioural models, important biases were detected among institutional investors and professionals in the investment management industry in South Africa.

To briefly summarize the findings, the results confirmed the existence of: 1. High confidence in market forecasting by

institutional investors 2. The use of availability heuristics to

underestimate the risk of more familiar markets 3. Herding behaviour due to the uniformity of the

information on which institutional investors base their forecasts

4. A loss aversion tendency where investors feel much more pain from losses than they feel joy from gains for the same amount involved. It is apparent that confidence was stronger for a

longer forecasting period. The reason for this may be due to many different reasons which may not so obvious. However, the observations shown above may also be subject to bias. The prevailing economic conditions in South African securities markets certainly influence the outlook for markets. The same research conducted in a downward phase in the economy may deliver different results. However, this does not mean that one can not depend on the research, it just means that the outcome of the research has to be interpreted in the relevant context. This also means that a model may be developed which may be used to interpret results. However, this can be the subject for further research.

As far as loss aversion is concerned, prospect theory also implies that people start to take greater risks when suffering from a loss. People tend to gamble, trying to recover some of the losses and in this sense then increase risk. Summary and Conclusion A lot of research has been conducted on variability in the financial markets. Behavioural finance as a field is fundamentally another attempt to try and explain

some of the noise that influence the markets and that affect the pricing of instruments and impacts on the return that we may expect from a portfolio. Understanding why people make certain decisions is important from an investment management perspective. It also helps us understand how markets will react to information. We also will be better equipped to advise clients when we understand more of the psychological aspects of human behaviour. We can not ignore the fact that emotion is more often than not, part of how we make decisions and it may, depending on the special circumstances, cause us to make irrational decisions.

Critics of behavioural finance typically support the efficient market hypothesis. They contend that behavioural finance comprises merely a collection of explained anomalies rather than being a true branch of finance and that these anomalies will eventually be priced out of the market or somehow be explained.

The survey of the South African managers concerning two of the emotional factors namely over/under confidence and loss aversion was carried out. The confidence aspect showed how different managers view the market. There seems to be a distinct bias which may have various sources. What is important about this is that the bias eventually affects the managers/markets negatively or positively. However, only knowing that it is there is unfortunately not enough. Experts/managers need to be able to somehow measure bias and use the outcome of the measurement to make more realistic forecasts/take more realistic investment actions, with other words, de-bias our forecasts/investment actions. At the same time managers/experts should be able to formulate better investment policy statements on behalf of their clients. Clients also need to understand something about the bias that they may be subject to and have realistic expectations.

The second emotional factor included in the survey was aversion to loss. The results show that people are a lot more emotional about losses and conversely require substantially more return on investment compared to a loss of a certain amount.


136

This confirms that institutional investors are loss averse.

This research reiterated that there is a human aspect that attaches to the business environment. The scholars that support the EMH can not argue away the fact that people make the markets function and the very people that do that are subject to certain biases. If the JSE’s ALSI falls by a thousand points from one day to the next due to sudden news of the prospect of high inflation in the coming months, no values actually changed from the one day to the next. It is the emotional aspect and the perceptions of people that changed due to the release of new economic information. These people therefore place lower values on the shares due to what is expected in the months to come. Suggestion for further research Similar, more comprehensive research may be undertaken per region/country over different economic cycles on a regular basis capturing especially the social aspects as regards the sources of bias mentioned in Figure 1. Setting up a database over time with factors that may be used as behavioural adjustments to financial magnitudes in the investment environment my help remove some of the bias. Although it may not be the ultimate solution to counter some of the volatility experienced in the financial markets, it is certainly more meaningful than ignoring the human element of investor actions. With frequent use and further research this field may eventually take its rightful place in the investment industry.

References

1. Barberis, N. and Thaler, R. (2001). A survey of behavioural finance. Working paper.

2. Daniel, K., Hirshleifer, D. and Teoh, S. (2002).

Investor psychology in capital markets: evidence and policy implications. Journal of Monetary Economics, 49, pp.139-209.

3. De Bondt, W. and Thaler, R. (1985). Does the stock market overreact?, Journal of Finance, 40, pp.793-805.

4. Descartes, R. (1637). Discourse on the Method of Rightly Conducting the Reason, The Great Books of the Western World, Volume: Descates and Spinoza.

5. Fama, E. (1998). Market efficiency, long-term returns and behavioural finance, Journal of Financial Economics, 49, pp. 283-306.

6. Fama, E. and French, K. (1993). Common risk factors in the returns on stocks and bonds, Journal of Financial Economics, 33, pp.3-56.

7. Kahneman, D. and Tversky, A. (1979). Prospect theory: An analysis of decision under risk, Econometrica, 47, pp.263-291.

8. Malkiel, B. (1995). Returns from investing in equity mutual funds. Journal of Finance, 50, pp.549-572.

9. Mas-Colell, A. (1999). The future of general equilibrium. Economic Review. 1, pp.207-214.

10. Mullainathan, S. and Thaler, R. (2000). Behavioural Economics. NBER working paper.

11. Rubinstein, M. (2000). Rational Markets: yes or no? The affirmative case, Financial Analyst Journal, 56( 3), pp.15-28.

12. Shleifer, A. and Vishny, R. (1997). The limits of arbitrage., Journal of Finance, 52(1), pp.35-55.

13. Simon, H. (1955). A behavioral model of rational choice. Quarterly Journal of Economics, 69, pp.99-118.

14. Thaler, R. (1985). Mental accounting and consumer choice. Marketing Science, 4, pp.199-214.

15. Thaler, R. (1999b). The end of behavioural finance. Financial Analyst Journal. 55(6), pp.12-17.

16. Toshino, M. (2004). Effectiveness of behavioural finance. Journal of Economics and Business. 190, pp.15-31.

17. Tversky, A. and Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, pp.1124-1131.

18. Von Neumann, J. and Morgenstein, O. (1947). Theory of games and economic behavior, Princeton.


137

LAYER HISTOGRAM PATTERNS IN FINANCIAL TIME SERIES

Van Zyl-Bulitta, VH*, Otte, R**, Van Rooyen, JH***

Abstract This study aims to investigate whether the phenomena found by Shnoll et al. when applying histogram pattern analysis techniques to stochastic processes from chemistry and physics are also present in financial time series, particularly exchange rate and index data. The phenomena are related to fine structure of non-smoothed frequency distributions drawn from statistically insufficient samples of changes and their patterns in time. Shnoll et al. use the notion of macroscopic fluctuations (MF) to explain the behaviour of sequences of histograms. Histogram patterns in time adhere to several laws that could not be detected when using time series analysis methods. In this study special emphasis is placed on the histogram pattern analysis of high frequency exchange rate data set. Following previous studies of the Shnoll phenomena from other fields, different steps of the histogram sequence analysis are carried out to determine whether the findings of Shnoll et al. could also be applied to financial market data. The findings presented here widen the understanding of time varying volatility and can aid in financial risk measurement and management. Outcomes of the study include an investigation of time series characteristics, more specifically the formation of discrete states. Keywords: Histogram, Layer, Pattern, Volatility, Discrete, States * University of Stellenbosch ** University of Stellenbosch *** University of Stellenbosch

Introduction and statement of the problem One of the most important magnitudes in the financial environment that we need to understand is the volatility of price changes. Volatility of price changes is unfortunately not directly observable in the financial markets. Market volatility reflects different events, happening in succession or synchronously. Volatility increases with increasing time scale as information about the future is either inaccurate or unknown. If we can enhance our understanding of volatility, we may be able to accurately forecast price changes and successfully manage financial risk and therefore maximize shareholder wealth. The prediction of price movements is not only important for devising a profitable trading strategy, but also for creating warning mechanisms to protect against or balance abrupt changes that would otherwise lead to losses. To learn about the timing of price changes of any magnitude could aid in managing trading processes aimed at valuing assets more accurately.

Many different models have been developed in recent years in an attempt to understand and model volatility. Fat tails and volatility clustering are two stylised facts that are often encountered in financial time series analyses. Apart from determining or estimating the volatility of a magnitude, an attempt is also made to find patterns in time series that may repeat themselves over the course of time. We often calculate the standard deviation based on historic price changes. However, this figure is certainly only at best a rough estimate of future volatility.

This study focuses on further extending the knowledge of the nature of the mechanisms that drive price changes. Volatility represents a measure of the way financial asset prices change. If financial markets record large declines of asset prices in short periods, one refers to a crash. The international financial system is highly linked, which often results in financial crises spanning more than one economy. Objective of the study The main objective of this study is to apply the method of histogram pattern analysis to financial data, searching for the phenomena found by Shnoll et al. in their investigations of stochastic processes from the natural sciences. This research describes layer histogram formation based on the work of Shnoll et al., applied to a financial time series. Two layer histogram methods are applied to a South African Rand (ZAR) data set of one minute frequency from 2 months.

The study includes an empirical analysis of a foreign currency exchange rate where the objective is to evaluate aspects of its distributional structure through compiling non-smoothed layer histograms of the currency dataset.

Data used One minute frequency exchange rate data of the South African Rand (ZAR) against the American Dollar (USD) have been obtained from Reuters for the time period covering 25 August 18:17 until 14:46 on 25


138

October 2006. The data set size comprises 38 006 values. Overview of approaches to modelling financial time series Financial markets are influenced to a large extent by human reactions of players in the marketplace to news that enter the market during trading times. This is especially true of the foreign exchange markets since they are influenced to a large extent by economic activities and trade between countries and generally by economic conditions prevalent in a country.

Many different models have been developed over a number of years in an attempt to encapsulate those variables that drive price changes. In the financial markets, volatility is not the result of a single magnitude that changes but rather represents the outcome of many different events and processes that happen simultaneously. Volatility is an important underlying magnitude that affects financial markets and is usually measured by the standard deviation. However, this is a very simplistic representation in one figure from a complex and dynamic system, where different players interact on different time scales under the influence of information feedback.

A new research field known as econophysics emerged from the vast increase in data availability since the 1990s coupled with the similarities of financial markets to stochastic processes known in the natural sciences. It spans a wide set of approaches to modelling and understanding the dynamics of financial markets, such as statistical analysis of the time evolution of asset prices and microscopic trading models. (Paul and Baschnagel, 1999, 131-132) Roehner also described stock markets from the viewpoint of statistical physics as an "open, out of equilibrium system" (2005, xiii), where different sorts of particles interact and where rules change over time.

Some of the more well known models developed in an attempt to understand financial time series include generalised autoregressive conditional heteroskedasticity ((G)ARCH) models. GARCH modelling builds on advances in the understanding and modelling of volatility in the last decades (Bollerslev 1986, Engle: 1982). It considers excess kurtosis (i.e., fat tail behaviour) and volatility clustering. These are two important characteristics of financial time series. It provides relatively accurate forecasts of variances and covariances of asset returns through its ability to model time-varying conditional variances. GARCH models may be applied to diverse fields such as risk management, portfolio management and asset allocation, option pricing, foreign exchange, and the term structure of interest rates.

In general, randomness implies incomplete knowledge of the process on which it is based. One assumes in constructing a financial model of price changes that useful information relevant for the future may be obtained from the patterns and frequencies of

past price changes. Furthermore, one may assume in this context that these frequencies reflect some intimate mechanism of the markets themselves. In such a case one may hope that these frequencies would remain stable over the course of time. (Bouchaud and Potters, 2003, 1-3)

The statistical approach to financial markets is based on the notion that whatever evolution takes place, it will do so sufficiently slowly, so that what happened in the past is relevant for predicting the future. However, this ‘weak stability’ hypothesis is often quite erroneous and particularly unreliable in times of financial crises. Hence the statistical description of financial fluctuations is imperfect, but nevertheless helpful to describe risks. While the prediction of future returns on the basis of past returns is much less justified, the amplitude of possible price changes - and not their sign - is to a certain extent predictable. (Bouchaud and Potters, 2003, 1-3) This amplitude reflects the intensity of price changes - in other words their volatility. Forecasting volatility in the financial environment is challenging as volatility cannot be observed directly in the market place (Mantegna and Stanley, 2000, 57, 76). At best, even a

posteriori, volatility is only approximately available. The volatility process is non-trivial and several stylised facts arise from its behaviour (Zumbach et al., nd, 1). Over the years, various researchers have developed several discrete-time and continuous-time volatility models. In particular, the modelling of volatility is based on the following stylised facts (Fasen et al., 2006, 108):

• time variation, • randomness, • heavy tails, • volatility clustering on high levels (long

memory of the volatility). The different modelling approaches take one or

more of these stylised facts into account. Bouchaud and Potters (2003, 122) noted that volatility fluctuations are a multiscale phenomenon. Therefore, the dynamics of volatility cannot be properly characterised on a single time scale. As far as volatility clustering is concerned, Zumbach et al. (nd, 1) pointed out that because of the slow decay of autocorrelations this clustering occurs on all time horizons. Several models have been developed to study one specific aspect of the behaviour of stock price movements, namely the clustering of volatility. Besides volatility clustering, other interesting properties of prices have also been described. These include jumps and downfalls in prices, heavy tails of price distributions, and their long memory. Non-linear models evolved from these findings, as they cannot be understood in the framework of linear models (Schiryaev, 2000, 152). Bouchaud and Potters (2003, 130) described how price-volatility correlations lead to anomalous skewness and volatility correlations induce anomalous kurtosis, i.e. fat tails. The failure of linear models to capture important characteristics of


139

real time series also led some researchers to consider chaotic modelling for financial markets.

Forecasting financial time series involves an element of uncertainty. Merely a fraction of the information about future price evolution can be known, based on the stochastic nature of financial systems. A longer prediction time horizon implies more inaccurate results and any prediction thus becomes more difficult due to greater uncertainty.

It is often assumed that the evolution of stock prices is driven by a noise sequence. Gaussian white noise is usually used as a first approximation. This distributional assumption is central to many financial applications, such as portfolio risk management and option pricing. Assuming a Gaussian distribution has many important consequences and comforts. Empirical research, however, does not confirm this assumption. The Gaussian distribution does not incorporate important findings such as heavy tails, asymmetry and excess kurtosis. (Stoyanov and Racheva-Jotova, 2004, 299-300)

Stable non-Gaussian distributions have been proposed as an alternative. As the name suggests, they have a desirable stability property as well as domains of attraction. Their use is complicated by their lack of closed-from expressions for probability density functions and cumulative distribution functions as well as their infinite second moment. According to Stoyanov and Racheva-Jotova (2004), the use of an infinite variance model for bounded financial asset returns seems inappropriate. However, since any empirical distribution possesses finite variance, infinite variance distributions may seem inappropriate for any application. Large deviations occurring in stock market price changes suggest that “any statistical theory based on finite-variance distributions is impossible to predict accurately” (Stoyanov and Racheva-Jotova, 2004, 300).

Shnoll referred to two aspects of the nature of fluctuations in analysing the similarity of histogram shapes, namely the histogram fine structure and the periodic recurrence of histogram shapes. Long-term investigations of various time series measurements, led Shnoll (2006) to believe that the laws that may be discovered by examining the fine structure of histograms, are not captured by traditional time series analysis methods. These phenomena have been found in the study of biological, chemical and physical stochastic processes. Stochastic processes and their properties are also at the core of financial modelling today. Shnoll et al. pointed out that accepted statistical methods, based on the central limit theorem, are not suitable for a histogram fine structure analysis. These techniques do not consider the fine structure of distributions, and they are insensitive to the particular shape of histograms. As Shnoll et al. explained, statistical techniques “overlook” (Shnoll et al., 1998, 1034) the fine structure, since they have been developed for different purposes.

Shnoll et al. (2000) formed histograms from an insufficient number of measurements and focused on

their fine structure. In contrast to analysing smooth histograms, which Shnoll et al. (1998, 1026, 1033) view as artefacts, their analysis focuses on empirical distributions that have only been smoothed a few times in succession so as to not destroy the extremes present in the distributions. Shnoll posed the question: Given a histogram pair that passed a test of similarity based on a visual comparison, what is their time distance and is there a time period that can be expected to occur more often than expected due to chance? According to this reasoning some predictive power may be gained if a particular histogram were known to reoccur periodically.

Shnoll and Mandelbrot referred to the concept of probability. Shnoll et al. (1998, 1035) state that the concepts of probability and stochasticity by themselves do “not yet predetermine the answer to the question concerning the distribution of fluctuations” (1998, 1035). According to Shnoll et al. these two concepts are closely associated with the concept of chaos. In this context they claimed that a distinction should be made between those types of chaos that differ in their distributions of fluctuations. On the one hand, the probability of fluctuation may fall monotonically with its magnitude, which they agreed is the real (or ideal) chaos. On the other hand they suggested that another chaos may be invented in which the distribution of fluctuations will be non-monotonic, corresponding to the histograms they present. Similarly, Mandelbrot proclaimed the usefulness and even necessity of recognising the existence of several distinct states of randomness and random and non-random variability. He denoted these distinct states of randomness as mild, wild and slow variability. (Mandelbrot, 1999, 2-3) Layer Histograms according to Shnoll Shnoll et al. originally applied the method to data from stochastic processes other than financial. The noise sequences that resulted from measurements of the original data resemble white noise. Thus, no structure or repeatable patterns would be expected. Financial data used in this study is inherently different to chemical reactions or radioactive decay. It is more likely that patterns may be revealed.

In their research regarding layer histograms and histogram patterns in time, Shnoll et al. involved as a first step the conversion of the pointwise time series to a series of successive histograms. Shnoll et al. then analysed the properties of histogram shapes in time, where each histogram is compared to each other histogram on an individual basis. (Panchelyuga and Shnoll, nd, 1) (Shnoll et al., 1998) (Shnoll et al., 2000, 207) Their goal was to verify the “fairly high probability of similar fine structure of distributions governing the results of simultaneous measurements of any processes in each time interval” (Shnoll et al., 2000, 205). For the layer histogram, the position in time is not the central focus of attention. The fine structure of the layer histogram provides information


140

on the shape of the distribution. More likely values would be represented by several pronounced peaks in the distribution, and less likely ones by troughs. Shnoll et al. refer to polyextremity of the distributions exhibiting alternating peaks and troughs, and thus several (poly) extremes.

Distributions with one extreme value can result from smoothing procedures. Shnoll et al. show in their work how after successive smoothing of histograms the distributions become bell-shaped. However, Shnoll et al. (1998, 1026, 1033) state that the smooth distributions may be regarded as artefacts.

Layer histograms represent one of several phenomena found by Shnoll et al. The phenomena that are named under the general term 'macroscopic fluctuations effect' are the near neighbour effect, synchronism, and monthly and annual recurrence of similarly shaped histograms. Our focus lies on the discrete states that become apparent in layer histograms. Other phenomena of the MF effect are not dealt with in this study. The focus is on the formation of discrete states in a financial data set by investigating the fine structure of unsmoothed layered histograms.

Discrete States in Unsmoothed Distribution: Peak and Trough Formation in Layer Histograms Layer Histogram Construction Details Before the layer histograms are calculated, the time series should be standardised. Figure 1 shows the data in raw format. The outliers are then limited to the specified threshold determined as a multiple of the series standard deviation. The effect of the thresholding may be seen in the higher outermost bins where the outliers are grouped. This ensures that bins represent the same sub-ranges for all layer histograms. Bins represent the frequency of occurrence of particular values within a certain subrange of the minute currency data. After standardisation, the subperiods are formed and layer histograms calculated.

Figure 1. ZAR/USD raw time series points with

missing data on weekends

Layer histograms were drawn from ZAR/USD exchange rate data. Table 1 below illustrates that, if the constant increment method is used, each successive histogram is drawn on the total number of data points from the first time series point. Figure 2 illustrates the daily ZAR recordings and from what subsets of the data layer histograms are drawn for each period. Figure 3 illustrates the input data for the layer histogram construction using the increment doubling method.

The fine structure of a data set may be represented by layered histograms. Shnoll found that in the construction of layered histograms, peaks and troughs emerge as more measurements are added. A layer histogram is composed of frequency distributions of a stepwise increasing number of observations. At each step, a fixed number of measurements are added and a histogram of all previous values and the new values is calculated. This step is repeated until the entire data set is exhausted.

Since each new subset contains the previous subset, each bin height of the new histogram will be equal to or larger than the respective bin of the previous histogram, thus giving rise to layers of histograms when all histograms are graphed on the same figure.

The first method, also used by Shnoll, adds a constant number of measurements (constant increment method) to the previous subset of values.

Table 1. Illustration of time series data points used for the constant increment method of layer histogram

construction


141

Figure 2. Constant increment method of layer histogram construction

Figure 3. Increment doubling method of layer histogram construction

Figure 2 illustrates how the time series are divided into successively larger subsets using the first 2 400 values of the ZAR time series. The number of values added to each new subperiod is determined by the increment size. Two methods are used for the incrementing.

The second approach (increment doubling method) deviates from the Shnoll layer histogram construction method. Instead of using constant increments at each step, the increment size is doubled. Thus, the doubling of time series subsets at each step of the layer histogram formation leads to a notable decrease in the number of calculated layer histograms as compared to the constant increment method.

Having described the data that enters the histogram calculation and two possible incrementing approaches we now describe the histogram construction parameters given the data. Instead of normalising the entire data set before subsets are formed, an alternative approach to construct layer

histograms would be first to normalise the data at each step before a layer histogram is calculated. However, this would lead bin ranges to vary as more measurement points are added because the range varies as subsets change.

Especially for the first layer histograms a changing data range changes the bin ranges. Because of the influence of a changing range on the subdivision into bin ranges, the same data point may be counted in different bins for different layer histograms. This approach was not followed, since a consistent bin range subdivision from the first to the last layer histogram was chosen. To illustrate the layered histograms and detect whether discrete states form, it is necessary for the bins to refer to the same interval range at each step.

The convergence of histogram bins to their final assignment makes it necessary to specify bin ranges before any layer histogram could be calculated. Otherwise the bin assignment will only be converging


142

to the final assignment as determined by the outlier threshold once a section that includes the top and bottom outlier occurs. Thus the programme calculating the layer histograms specifies the bin edges to ensure that bin ranges are fixed from layer histogram one, instead of a gradual convergence of bins to their final values. If an outlier to the top and an outlier to the bottom occurred in the first section from which a layer histogram is calculated, the bins would be fixed at their correct value from the beginning, also without the prior bin assignment.

If this procedure of fixing bin widths from step one is not followed in constructing layer histograms and the outliers do not occur in the first section, bins

will shift around for the first few steps. This shifting continues until outliers are reached for the first time fixing the maximum bin range.

Polyextremes in a subset of data set

The representation of each time series in sequential layers is now provided. Figure 4 shows the first five layer lines of the ZAR layer histogram using the constant increment method. This illustration shows that the polyextremes already become apparent for a subset of the ZAR data from a few days. Compare the subset layer histogram (Figure 4) to the entire layer histogram figure (Figure 5).

Figure 4. Constant increment method applied to the first 5 layer histograms (layer

histogram construction parameters: 400, 40, 2)

Figure 5. Constant increment method applied to all ZAR data (layer histogram construction parameters: 400, 40, 2)

The sensitivity of the structure of successively

drawn layers to parameter settings is illustrated in the next section. These parameters include the number of increments that are added after each layer line, the

number of bins, and the outlier limitation threshold. Figures 5-10 show that the formation of peaks and troughs grows more pronounced as layer lines are added and they show the sensitivity of the fine


143

structure to changing the number of bins into which measurements are classified. Thus this study shows that the discrete formation effect found by Shnoll for other stochastic processes is also present in financial data sets.

Comparison of Methods

Two methods of layer histogram construction were employed. The original method also used by Shnoll was complemented with another method of incrementing. This serves to test the effect of the incrementing procedure on the final layer histogram shape. Since the addition of a constant number of increments gradually decreases the relative contribution of the latest time series section that is added to the layer histogram shape, the same relative number or measurements has also been used for incrementing in the increment doubling method.

The study expands further on the method employed by Shnoll et al. (1998, 126-1028). It could be expected that the fine structure of layered histograms would grow more pronounced as measurements are increased because of the way Shnoll et al. propose to increment the number of measurements for successive layer histograms. For the method employed by Shnoll et al., the addition of the same constant number of measurements at each step will likely develop the fine structure that is there already.

As data are added, each new increment becomes less influential because the number of added values is constant in absolute terms. The distribution then necessarily represents a similar fine structure to the previous histogram. The presence of fine structure will now be tested for using an alternative layer histogram construction method.

Instead of adding the same absolute number of measurements each time, the second method constructs layer histograms by adding the same relative number of measurements at each step. This

means that at each step the number of measurements used is doubled, resulting in an exponential growth of the subset size until the entire data set is included. The original method's subsets grow linearly, and therefore the method is referred to as constant increment method in contrast to the increment doubling method.

The number of increments influences the layer histogram shape less than the number of bins or the outlier limitation, i.e. the outlier limitation and the number of bins determine the layer histogram end result the most. The variations of layer histograms resulting from different parameter choices and different construction methods make it apparent that there is a fine structure in the data. Furthermore, as Shnoll et al. found in their data, peaks and troughs also emerge in financial layer histograms.

It is important to understand that there is no one correct layer histogram. Different shapes and structures may be detected, where the parameter choices play an important role. The first method calculated increments taking much smaller steps, and thus more layer lines were drawn. These represent incrementally less additional information. For the construction method that doubled the increments of previous steps, much less layer lines were necessary.

The sensitivity to parameter settings, as well as the layer histogram construction method, are compared using the ZAR data set. Different settings for the number of bins are illustrated. In doing this, the discrete state structure becomes apparent at different resolutions. The number of increments is kept unaltered among the data sets, namely 400 for ZAR layer histograms. Outlier limitation is kept the same for the ZAR data to focus the illustration on the sensitivity of the final shape to the number of bins used. The effect of a change in the outlier limitation is the reassignment of bin widths and possibly a decrease in the number of empty bins as the total bin range becomes smaller.

Figure 6. Increment doubling method applied to all ZAR data (layer histogram construction parameters: 400, 40, 2)


144

Figure 7. Constant increment method (parameters: 400, 80, 2)

Figure 8. Increment doubling method (parameters: 400, 80, 2)

Figure 9. Constant increment method (parameters: 400, 160, 2)


145

Figure 10. Increment doubling method (parameters: 400, 160, 2)

Figure 6 shows the same parameter settings as

Figure 5 using the increment doubling method. Figures 7 to 8 show a finer resolution (80 bins) and Figures 9 to 10 show layer histograms using 160 bins. It can be detected that more peaks and troughs emerge as the number of bins is increased. The three major polyextremes at about 13, 21 and 30 is apparent in all illustrations. Thus, besides the most likely value (centre extreme), in a time series of 2 months we found two more likely states that the process occupies. When looking back again at Figure 2, the three polyextremes already become apparent after applying this method to only 2 days data. Layer histograms consider time information (date and time) of time series values deterministically in presenting the distributional structure of the data. The subsets of each layer histogram are increased by counting the subsequent number of values to be added, which disregard their position in time. A particular layer histogram counts the frequency distribution of values in respective bins and ignores the sequential information as any histogram does. It may be expected by construction that the information that is added to sequential layer lines represents new information occurring later in time.

Summary

We find that discrete states form in the financial data, similar to that found by Shnoll et al. in the data that were used by them. Both methods for constructing layer histograms discussed above show similar structures forming in the data. These structures are represented by polyextremes growing more pronounced.

The unsmoothed distributional character of the entire data set was illustrated via layer histograms. The fine structure of the entire distribution and the effect of polyextremity became apparent. The shape of a layer histogram differs according to the parameter

choice used to calculate it. The construction parameter choices are the number of increments chosen to calculate each successive layer histogram line, the number of bins into which these measurements are classified, and the outlier limitation threshold.

The method Shnoll et al. applied to construct layer histograms was implemented and extended. Both methods showed that the polyextremes present in the histogram of the entire data set appear already in small subsets, i.e. after a few layer histogram lines were drawn. The polyextremes shown by both layer histogram construction methods coincide approximately. Previously it was shown (van Zyl, 2007) that the increment size has a mild effect on the final layer histogram shape, while the choice of the outlier limitation threshold and the number of bins substantially affect the layer histogram shape. The outlier limitation directly influences the bin assignment.

Too many polyextremes - especially when they appear jagged - and empty bins may be an indication that too many bins have been chosen. To calculate a meaningful layer histogram that provides insight into the statistical properties of the process, the relation of the number of bins to the increment size needs to be well chosen. The bin width could be controlled by the outlier limitation threshold.

If polyextremes are present in the layer histogram of the entire data set, as well as in the first layer lines, this means that information from a smaller subset of the data could provide information relevant for a longer time scale.

Several (poly-) extremes of a layer histogram provide more detailed information about the expected value of the process. In the ZAR example, there are 3 more likely states for the entire 2 months, which are already apparent after the data of a few days has been represented by a layer histogram. These states are the polyextremes, namely bins 13, 20, and 30. This


146

information could be made useful for the purpose of prediction.

Layer histograms focused on distributional characteristics of price changes without taking their time explicitly into consideration. Discrete state formation was detected, as predicted by Shnoll. Shnoll's method of constructing layer histograms was applied first, and a new method was proposed to investigate the fine structure of these distributions. It was shown that discrete states also form and fine structure also emerges from the use of this alternative method.

It must be remembered that the choice of parameters may substantially influence the final shape of histograms. These parameters include outlier limitation, the selection of the number of bins from which histograms are constructed, not so much the number of measurements.

Conclusion

The research underlines the fact that, according to the histogram analysis, the currency values as contained in the financial time series, are the result of non-random events/actions of participants in the market place. The patterns that emerge from the layer histograms analysis reflect the cumulative actions of dealers. The analysis as is carried out in this research attempts to find patterns that may be used to better manage exposure to market changes to maximise profit and eventually shareholder’s wealth. The research further illustrates that the findings of Shnoll et al. (applied in the natural sciences field) can also be applied to financial market data.

Suggestions for further research

This research draws attention to the variation inherent in financial data sets as well as preferred states of a financial process. Traditional statistical techniques describe different aspects of the variation in price changes. For this reason visual comparison of histogram shapes was attempted by Shnoll et al. to show a different aspect of time series variation in an attempt to bypass some of the inherent weaknesses of traditional statistical analysis techniques.

Further research may include developing intelligent computer systems that will search for such patterns in data sets and present the results in such a way that management may be able to use it for risk management and profit maximisation purposes. Artificial intelligence systems may be helpful in this respect.

The other phenomena found by Shnoll et al. can also be built upon, namely the synchronous variation

of price changes, the near neighbour effect and the recurrence of similar histogram shapes after one week, one month, or one year. The findings of Shnoll show two specific year periods where histogram shape recurrence probability peaks. The methods used to detect the phenomena could be automated and comparison methods could use a variety of appropriate distance measures.

References

1. Bouchaud, J.-P., Potters,M., , and Meyer,M. (2003). Apparent multifractality in financial time series. arXiv:cond-mat/9906347 v1 23 Jun 1999.

2. Bouchaud, J.-P. and Potters, M. (2003). Theory of

financial risk and derivative pricing - from statistical

physics to risk management. Cambridge University Press, Cambridge, United Kingdom, 2nd edition.

3. Fasen, V., Klüppelberg, C., and Lindner, A. (2006). Extremal behavior of stochastic volatility models. In Shiryaev, A. N., Grossinho, M. R., Oliveira, P. E., and Esquivel, M. L., editors, Stochastic Finance, pages 107–155, Bellingham, Washington. Springer.

4. Mandelbrot, B. B. (1999). N1. In Multifractals and 1/f

noise - Wild self-affinity in Physics, Selecta Volume N, pages 1–35. Springer, New York.

5. Mantegna, R. N. and Stanley, H. E. (2000). An

Introduction to Econophysics - correlations and

Complexity in Finance. Cambridge University Press, Cambridge, United Kingdom.

6. Paul, W. and Baschnagel, J. (1999). Stochastic

processes - from physics to finance. Springer, Berlin, Heidelberg.

7. Roehner, B. M. (2005). Patterns of speculation - a

study in observational econophysics, Cambridge University Press, Cambridge.

8. Schiryaev, A. N. (2000). Essentials of stochastic

finance - facts, models, theory, volume 3 of Advanced

series on statistical science and applied probability. World Scientific, Singapore.

9. Shnoll, S. E., Kolombet, V. A., Pozharskii, E. V., Zenchenko, T. A., Zvereva, I. M., and Konradov, A. A. (1998). Realization of discrete states during fluctuations in macroscopic processes. Physics

Uspekhi, 10(41):1025–1035. Uspekhi Fizicheskikh Nauk.

10. Stoyanov, S. and Racheva-Jotova, B. (2004). Numerical methods for stable modeling in financial risk management. In Handbook of computational and

numerical methods in finance. 11. Van Zyl, V. H. (2007). Searching for histogram

patterns due to macroscopic fluctuations in financial time series, Master thesis, University of Stellenbosch

12. Zumbach, G., Pictet, O. V., and Masutti, O. (n.d.). Genetic programming with syntactic restrictions applied to financial volatility forecasting. Retrieved on 26.09.2006 from: http://www.olsen.ch/research/ working papers.html.


147

THE EFFECT OF PROCUREMENT STRATEGIES OF MILLING

COMPANIES ON THE PRICE OF MAIZE

W. Rossouw*, J. Young**

Abstract Since ultra poor South Africans spend up to a fifth of their income on maize alone, the demand for this commodity is price-inelastic, i.e. consumers have no choice but to absorb price increases. As such the success of procurement strategies from milling companies will ultimately have a direct impact on the financial well-being of the poor. Even though derivative instruments are available to use as counter against market fluctuations, the price risk management success of groups with a concern on SAFEX suggests that this is not achieved as yet, ultimately to the detriment of consumers. The view exists that markets are efficient and the return offered by the futures exchange cannot consistently be outperformed. This paper argues the exact opposite, since the use of the proposed futures/options strategies result in returns superior to that of the market. Keywords: Price-risk, Futures contracts, Options contracts, Momentum strategy, Maximum price strategy, Indexed strangle strategy *PO Box 1187, Welkom, South Africa, 9460 Phone: +27 57 352 2805 Mobile: +27 82 490 5059 Fax: +27 57 352 7874 E-mail: [email protected] **University of South Africa PO Box 52 185, Wierda Park, Centurion, Pretoria, South Africa, 0149 Phone: 27 12 429 3725 Mobile: 27 83 307 6265 Fax: 27 12 429 3552 Email: [email protected] and/or [email protected]

1. Introduction Agriculture is of inestimable value to South Africa since it is a major source of job creation and plays an important role in earning foreign exchange. The most significant contribution of agriculture, and especially maize, remains its ability to provide food for the nation. For a number of decades grain prices were determined by government legislation until this policy of grain pricing was amended by the Marketing of Agricultural Products Act 47 of 1996. The introduction of this Act deregulated the agricultural sector and resulted in the value of soft commodities being determined by the futures price of the underlying asset as traded on the South African Futures Exchange (SAFEX). This required market participants to adapt to a volatile environment resulting in uncertainty and financial losses. Since ultra poor South Africans spend up to a fifth of their income on maize alone, the demand for this commodity is price-inelastic, i.e. consumers have no choice but to absorb price increases. As such the success of procurement strategies from milling companies will ultimately have a direct impact on the financial well-being of the poor.

The performance of price-risk management strategies, as implemented by stakeholders in the futures market was identified and the performance calculated against a relevant benchmark. In addition, a structured approach to price-risk management was investigated via three alternative strategies and in turn compared against a benchmark.

This paper aims to identify the success by which participants in the soft commodity futures market mitigate price risk of maize by means of derivative instruments. The derivative price risk management strategies will be benchmarked against the return offered by the market. Successful risk mitigation with predictable returns could serve as motivation for organisations to develop similar risk management strategies with the goal of realizing below-average procurement prices to the ultimate benefit of consumers. This paper is divided into the following topics: • Theoretical background on derivative

instruments. • Price-risk management performance of

stakeholders in the futures market. • Background and application of proposed price-

risk management strategies.


148

• Evaluation of risk management strategies versus a benchmark.

2. Theoretical background on derivative instruments According to Bodie, Kane & Marcus (2002:15), derivatives are financial arrangements between two parties whose payments are based on, or derived from, the performance of some agreed-upon benchmark. Derivative instruments can be based upon an underlying asset entailing one of the following: • Currencies • Commodities • Government or corporate debt • Home mortgages • Equity shares • Interest rates 2.1. Distinction between derivative

instruments

2.1.1. Futures contracts

“A futures contract is a notional commitment to take delivery (purchase) or to make delivery of (sell) a given quantity of a specific instrument on a specified future date at a price determined at the time of taking out the contract” (Valsamakis et al 2003:267). Since futures are exchange traded contracts, the following characteristics are standardized: 1. The asset-type; 2. quantity of the asset; 3. quality of the asset; and 4. the future maturity date.

2.1.2. Options contracts

The fundamental difference between an option contract and futures contracts is the higher level of flexibility inherent in options contracts. This is due to the holder of an option having the right, but not the obligation, to enter into the underlying futures contract. (Madura 2000:66.). Two types of option contracts exist in the futures market, namely: • Call options. A call option provides the buyer

(long position holder) with the right, but not an obligation, to buy an asset for a certain price by a specific date.

• Put options. A put option provides the buyer (long position holder) with the right, but not an obligation, to deliver an asset for a certain price by a specific date

3. Price-risk management performance

of stakeholders in the futures market The feasibility and success of price risk management models applied on futures prices by speculators and hedgers can only be determined once its performance has been compared to the returns offered by the market or by alternative risk management models.

The specific calculation of the market return or alternative risk management model against which the performance of the particular risk management strategy is measured is an important consideration in the evaluation of a strategy. The concept underlying the evaluation of the performance of risk management strategies is the comparison of net prices achieved by these strategies versus the returns offered by similar active strategies or the passive market. This benchmarking serves as an objective standard of performance (Irwin, Good, Martines-Filho & Batts 2006-03:2). In its simplest form, benchmarking involves comparisons (Brigham, Daves & Gapenski 1999:80).

Benchmarking, according to external benchmarks, is based upon the efficient-market theory. This entails that markets are rational, all-knowing and that competition among participants in the marketplace will immediately eliminate all possible arbitrage opportunities available through the exploitation thereof. (Irwin et al 2006-02:29-30.)

For the purposes of this paper the benchmark is described as the average price of a single commodity, or group of commodities, on a specific date or over a pre-determined period. In this context the benchmark used, should measure the average SAFEX white maize price for July delivery over the contract-lifetime for the processing company who follows no active hedging strategies. The average price is determined in order to reflect the returns of a naïve strategy, hedging equal amounts of the commodity every day over the duration of the contract. This is consistent with research already done on this subject (Irwin, Good, Martines-Filho & Hagedorn 2005:27-31).

The pioneering work in this field of performance measurement was done by Irwin, Good, Martines-Filho and Hagedorn (2003) for the AgMas project at the University of Illinois. Every price-risk management recommendation from over twenty professional trading companies was recorded since 1994. A comparison was made between the net results of every recommendation from the individual companies and the benchmark average price constructed from the daily closing prices over the contract lifetime. The results indicate that only one professional trading company managed to outperform the simple average benchmark (by less than 7% on average). In other words, less than 5% of the professional trading companies managed to outperform the futures market.

Another finding from the research indicated that the net advisory prices vary substantially between companies, with differences of up to 70% on the realized futures price. The conclusion is made that markets are efficient and no additional profits can be made through risk management strategies. (Irwin et al 2003.)

Thorough research has been done on the forecasting ability of speculators in the soft commodity futures market. The earliest findings on


149

the performance of speculators in the grain futures market were published in 1934. Stewart (1934:415-433) made a detailed analysis of 9000 accounts of a nationwide brokers’ firm for the period 1925 to 1934. These accounts reflected exclusively speculative transactions in grain futures. The most striking finding of this research project was that nearly 75% of speculators lost money. More concerning, however, is that the entire sample highlighted losses six times the value of total gains.

In 2001, 67 years after Stewart published his findings, Wang (2001:929-952) released the results of his study on the predictability of returns in the futures market. It showed that large speculators in the futures market are still unable to accurately predict price movements.

Locally, the futures market was stunned when trustees of a pension fund stated their intention to sue WJ Morgan, a South African futures broker, for losses of R1.4 billion sustained due to over-exposure on SAFEX. The decision to expose funds of R2.7 billion to the derivatives market was taken by WJ Morgan on the basis of expectations of a continued rise in maize prices (The Star 2003).

The inability of producers to effectively manage their exposure to adverse price movements is highlighted in the well-known fact that two thirds of producers short the futures market in the bottom third of the price range (Decision Commodities 2006).

The inability of processors to effectively manage the risk of volatile prices has been well documented and for the purposes of this paper a non-probability purposive sample will be used with reference to the procurement companies and the trading year over which the hedging results will be evaluated. The accurate and available historical price and volatility information are the main reasons as to why a secondary data analysis is chosen as the most appropriate data collection method. The evaluation of the price-risk management performance of processors will be done through an investigation into the procurement results of African Products and Tiger Brands. These two processing companies are used for the purposes of this study since they are recognized as two of the main role players in the procurement market for the following reasons:

• African Products consume close to 7% of the average annual maize crop (Tongaat Hullett 2003:4).

• Tiger Brands is recognized as being among the four biggest milling companies in South Africa (Chabane 2003:6-7).

In order to determine the ability of African Products and Tiger Brands to successfully eliminate volatile futures prices, the year with the biggest maize price movement (2003) will be used in order to explore the impact of big price movements on hedging strategies.

African Products

Tongaat-Hulett is a Group consisting of four closely linked and focused businesses. One of these

businesses is African Products, Africa’s largest manufacturer of starch and glucose. Its five mills consume in excess of 600 000 tons of maize annually, i.e. nearly 7% of the average South African maize crop.

In the chairman’s statement for the 2003 financial year, Cedric Savage blamed expensive maize procurement costs as one of the main reasons contributing to a headline loss of R93 million, down from a 2002 headline profit of R380 million. During the year African Products followed its long established strategy with a focus on price stability. A characteristic of this procurement strategy was that the impact of price increases was nullified. The strategy simply consisted of going long the futures market. As the market came down from levels of R2000/ton late in 2002 to below R800/ton in 2003, African Products incurred huge valuation losses on the procurement contracts. This forced Peter Staude, Chief Executive Officer, to comment in Tongaat Huletts’ 2003 Annual Report that a new model of maize procurement is needed. (Tongaat Hulett 2003.) Tiger Brands

In the Group results for the year ended 30 September 2002, the following comment is made: “The effects of sharply higher grain prices were mitigated by the benefits of an effective procurement programme, which resulted in the group being able to source its grain requirements at below market prices.”

(Tiger Brands 2002.) In the subsequent months the futures market was

characterized by a sharp drop in white maize futures prices. This forced the Chief Executive Officer of Tiger Brands to make the following contrasting comment in the 2003 Annual Report: “High priced maize stocks carried over from last year, volatile maize prices and the stronger rand impacted on operating income which declined by 5% to R1.9 billion.” (Tiger Brands 2003.)

After much praise for the hedging strategy in the previous year, a comment was made in the 2003 Annual Report that new hedging strategies will be introduced to provide for better hedging against volatile commodity prices. This underlines the absence of a hedging strategy with a predictable outcome that is able to beat the average market price. 4. Background and application of

proposed price-risk management strategies

In the previous section the conclusion was made that neither hedgers nor speculators are able to outperform the returns offered by the market, in accordance with the efficient market hypothesis. The average price was consequently chosen as benchmark against which the results from the proposed risk management strategies will be compared.

The three proposed price-risk management strategies are based on a core/satellite model. This


150

framework is used in order to develop a strategy with the potential to add additional value above an average price, where the main strategy is a risk-controlled position, with an active structure aiming to add additional value. 6.1. Momentum strategy Erb and Harvey (2005:3) define the Momentum strategy as a method of pursuing above average returns by investing in commodity futures with positive past price movements. This is consistent with the description of Spurgin (1999:1) which states that the Momentum strategy involves buying the underlying asset that is rising in price and selling the asset when prices are falling. For the purposes of this study a drop in prices will not result in a short position being taken, as the underlying hedge position should result in delivery of the commodity being taken. Therefore the proposed Momentum strategy will hold exclusively long positions, similar to the Goldman Sachs Commodity Index (Spurgin 1999:1). The long-only Momentum strategy for the purposes of this paper is ultimately defined as buying the underlying commodity in the event of an increase in prices (or no change in prices), whilst a drop in the commodity price results in no action being taken.

Whilst sufficient literature on Momentum in equity markets exist (Carhart 1997:57-82; Johnson 2004:585-608) there does not seem to be general agreement as to the reasons why this strategy is successful in its application. In terms of the proposed Momentum strategy the view of Johnson (2004:585-608) seems to be the most likely reason for potential hedging success on SAFEX. He argues that the returns achieved by such a strategy are a payoff for

taking more risk than merely buying the Average Price Index.

The Momentum strategy, to be applied on historical data, possesses the following distinctive features: • Every trading day that July white maize prices

increase (or remain unchanged), a long position will be taken.

• Every trading day that July white maize prices drop, no position will be taken on SAFEX.

• The first trading day on which July white maize prices increase after a drop in prices, long positions will be taken. The number of long positions entered into should equal the sum of the number of trading days since the last trading day on which prices closed higher. The July contract is used in the evaluation of the

proposed strategy, as it is the most liquid futures contract available on SAFEX. Although grain processors do not wish to take delivery of a year’s stock all at once, the resultant long position can be rolled forward to the delivery month in which the grain is required. Rolling a position forward consists of going short the July contract against the long position obtained from the Momentum strategy, and immediately going long the desired contract month (Hull 2002:458).

For the purposes of the evaluation of the strategy, the assumption is made that a single daily position taken on SAFEX entails one futures contract (100 metric tons) traded at its closing price. Since processors are naturally much shorter the market, the daily number of futures contracts taken as position on SAFEX can be adjusted according to individual needs.

Table 1. Summary of Momentum strategy applied on July white maize data from 2001 to 2006

* Based upon a processor procuring 600 000 metric tons of maize annually

The results indicate that the benchmark index is outperformed by the Momentum strategy in five of the six years under review, as the realized procurement price is lower than the Average Price Index. The extent by which the Momentum strategy beats the market varies substantially, from R0.79/metric ton in 2001 up to R11.86/metric ton in 2003.

6.2. Maximum price strategy Hull (2002:461) defines an exotic option simply as “a nonstandard option”. He states that the price and volatility of plain vanilla options are determined by an exchange, whereas financial engineers develop exotic options to be sold at a price not necessarily related to prices quoted by the market. He further argues that an exotic product comes about due to a number of factors. These include a specific need for a hedging product in the market and to reflect the user’s view on

Year Average price Momentum strategy Value gained (R/ton) Total value gained*

2001 739.40 R 738.61 R 0.79 R 473,597.26 R 2002 1,254.24 R 1,252.48 R 1.76 R 1,057,304.03 R 2003 1,400.50 R 1,388.64 R 11.86 R 7,113,240.42 R 2004 1,086.16 R 1,084.39 R 1.77 R 1,060,166.59 R 2005 810.74 R 807.77 R 2.97 R 1,779,472.85 R 2006 958.11 R 959.72 R -1.60 R -962,400.85 R


151

potential future price movements (Hull 2002:394). He identifies an Asian option as “an option with a payoff dependent on the average price of the underlying asset during a specified period” (Hull 2002:456).

In their research on options available in the futures market for hedging purposes, Hagedorn, Irwin, Good, Martines-Filho, Sherrick and Schnitkey (2003:3-5) describes new generation contracts as products which use automated pricing rules, discretionary marketing, options strategies, or a combination of all three in order to achieve an average hedge price. They classify new generation contracts into three basic categories, namely automated pricing contracts, managed hedging contracts and combination contracts.

The Maximum price strategy is an exotic option strategy since it possesses no standard features and come about due to a specific need expressed by risk averse hedgers in the futures market. The strategy may be mistaken for an Asian option, but since its payoff is not dependant on an average price, it should rather be classified as a managed hedging contract. It complies with the definition of a managed hedging contract (Hagedorn et al 2003:4) since a specific volume of the underlying commodity is hedged over a predetermined period with a fixed maximum price.

Taking all of the above into account, the Maximum price strategy can be defined as an exotic and managed hedging strategy which guarantees a maximum procurement price. The benefit of daily price movements lower than the maximum price will lead to a reduction in the final procurement price of the commodity, whilst daily price movements higher than the maximum price will have no adverse impact on the price at which the underlying commodity is bought.

The characteristics which distinguish the Maximum price strategy from alternative exotic options can be summarized as follows: • On the trading day on which the strategy is

initiated the buyer of the underlying commodity is guaranteed a maximum procurement price.

• The total volume of maize to be hedged is divided into the number of trading days over which the strategy will be applied on futures prices; i.e. the daily volume.

• Every trading day on which the daily price is higher than the maximum price the daily volume of maize will be purchased at the maximum price. Should the price of maize be lower than the maximum price on option expiration, the long position holder will receive the benefit of the lower price, as on option expiration, for the total number of trading days over which the maize price traded higher than the maximum price.

• Every trading day, on which the price of maize is lower than the maximum price, the daily volume of maize will be purchased at the lower price.

The average of the daily volume of long positions

taken over a predetermined strategy period will result

in the final procurement price. The practical implementation of this strategy is structured as follows: • On the date of commencement, at-the-money call

options are bought for the total volume of grain to be hedged. This results in a guaranteed maximum procurement price.

• The volume of grain to be hedged is divided into the number of trading days from the date of commencement to option expiration. This is known as the daily volume.

• Every trading day between the date of commencement and option expiry on which the daily price is lower than the maximum price, the daily volume of call options are sold and replaced by a daily volume long futures position. By going short the option (which is now an out-of-the money call option), the net cost of the strategy is reduced.

• Every trading day between the date of commencement and option expiry on which the daily price is higher than the maximum price, no action is taken. The daily volume of grain is hedged by the call option at the maximum price, which is the strike level of the call option. This call option is in-the-money.

• Should the daily price of the underlying commodity be higher than the maximum price (strike level of option) on date of option expiry, the call-options will automatically become long futures positions. In the event of a daily price lower than the maximum price on the date of option expiry, long futures positions will be taken to the extent of the daily volume multiply by the total number of trading days over which the daily price traded higher than the maximum price.

• The average price of the daily volume of grain hedged over the strategy period will result in the procurement price for the underlying commodity. By adding the premiums and broking fees of the call options to the realized procurement price, the net hedged price can be compared to the market benchmark.

• Since historical data on maize prices and volatility are available, the Black-Scholes model will be used to calculate the historical prices of options. The Maximum price strategy will be initiated on the day on which volatility are first published, since this represents the first trading day on which options can be traded. The July contract is used in the evaluation of the

proposed strategy, as it is the most liquid futures contract available on SAFEX. Although grain processors do not wish to take delivery of a year’s stock all at once, the resultant long position can be rolled forward to the delivery month in which the grain is required. Rolling a position forward consists of going short the July contract against the long position obtained from the Maximum price strategy,


152

and immediately going long the desired contract month (Hull 2002:458).

For the purposes of the evaluation of the strategy, the assumption is made that the daily volume on SAFEX entails one futures contract (100 metric tons) traded at its closing price. Since processors are naturally much shorter the market, the volume of

grain to be hedged can be adjusted according to individual needs. Once the value of the long call option is below R1/ton, no effort will be made to go short the option on days when the daily price is lower than the maximum price, since the broking cost will exceed the premium benefit of the option.

Table 6.2. Summary of Maximum price strategy applied on July white maize data from 2001 to 2006

Year Average Price Index Maximum price strategy Value gained (R/ton) Total value gained* 2001 R 739.40 R 713.29 R 26.11 R 15,663,209.30 2002 R 1,254.24 R 1,102.03 R 152.21 R 91,325,589.74 2003 R 1,400.50 R 1,291.70 R 108.80 R 65,280,000.00 2004 R 1,086.16 R 1,021.03 R 65.13 R 39,076,852.94 2005 R 810.74 R 787.55 R 23.19 R 13,913,010.75 2006 R 958.11 R 777.24 R 180.87 R 108,523,477.20

* Based upon a processor procuring 600 000 metric tons of maize annually

The results indicate that the benchmark index is outperformed by the Maximum price strategy in all of the years under review, as the realized procurement price is lower than the Average Price Index.

The extent by which the Maximum price strategy beats the market varies substantially, from R23.19/metric ton in 2005 up to R180.87/metric ton in 2006. There are two reasons for the inconsistency in the rand value by which the Average Price Index is outperformed: • A bigger price movement over the duration of the

contract results in a higher rand value by which the Average Price Index is outperformed.

• The level at which a long position in call options is entered into.

6.3. Indexed strangle strategy In their study on information flows in financial markets, Berchtold and Norden (2005:1147-1172) analyzed two types of information flows, namely return information and volatility information. Whereas return information embodies the knowledge of informed investors on whether prices will increase

or decrease, volatility information entails the lack of knowledge on the direction of market movements. The Indexed strangle strategy is consistent with volatility information flows, since the future direction of market movements cannot be predicted. As such the Indexed strangle will aim to provide additional value to an average price index on both upward and downward price movements.

Table 6.3 confirms a trend among historical volatility movements for the July white maize contract. The following important conclusions can be made from this information: • Volatility is low at commencement of the

contract. • Volatility increase over time from

commencement and reach a peak over December/January. This is due to the high levels of uncertainty in the maize market during planting time.

• From January up until option expiration volatility decrease. The lower volatility is brought about by higher levels of supply and demand certainty.

Table 3. 10-day Volatility 2002-2007

Period from: Period to: 2002 2003 2004 2005 2006 Average

01-May 10-May 38.0 38.0 11-May 20-May 38.0 38.0 21-May 30-May 24.0 38.0 31.0 31-May 09-Jun 24.0 38.0 31.0 10-Jun 19-Jun 24.0 37.0 30.5 20-Jun 29-Jun 25.1 30.8 37.0 31.0 30-Jun 09-Jul 26.4 31.0 26.0 36.1 29.9 10-Jul 19-Jul 25.3 31.0 26.4 33.9 29.1 20-Jul 29-Jul 27.0 34.6 30.8 33.0 31.3 30-Jul 08-Aug 27.5 36.5 36.0 33.0 33.3

09-Aug 18-Aug 25.8 36.7 37.7 33.0 33.3 19-Aug 28-Aug 24.7 36.3 38.9 33.5 33.3 29-Aug 07-Sep 25.4 37.2 41.0 40.6 36.1 08-Sep 17-Sep 26.6 38.0 40.3 48.5 38.3 18-Sep 27-Sep 28.0 37.5 41.4 49.9 39.2


153

Table 3 continued 28-Sep 07-Oct 28.4 36.7 41.5 49.8 39.1 08-Oct 17-Oct 27.5 37.9 42.2 47.5 38.8 18-Oct 27-Oct 29.0 38.7 41.8 47.8 39.3 28-Oct 06-Nov 32.2 40.4 44.7 48.3 41.4 07-Nov 16-Nov 31.9 41.8 47.0 45.6 41.6 17-Nov 26-Nov 31.1 40.5 48.4 48.7 42.2 27-Nov 06-Dec 30.3 37.6 51.2 49.7 42.2 07-Dec 16-Dec 33.2 42.8 51.9 46.5 43.6 17-Dec 26-Dec 30.8 49.8 50.1 47.2 44.5 27-Dec 05-Jan 39.3 29.1 52.1 52.3 53.3 45.2 06-Jan 15-Jan 39.1 27.4 52.0 52.3 47.5 43.7 16-Jan 25-Jan 38.8 29.0 51.0 46.7 41.1 41.3 26-Jan 04-Feb 35.6 36.7 45.5 47.6 38.4 40.8 05-Feb 14-Feb 37.2 42.6 46.8 48.4 40.0 43.0 15-Feb 24-Feb 33.2 39.7 45.1 42.1 35.7 39.2 25-Feb 05-Mar 32.8 37.7 41.6 35.6 37.3 37.0 06-Mar 15-Mar 36.2 42.1 44.5 39.6 34.1 39.3 16-Mar 25-Mar 37.1 43.8 44.9 40.9 30.2 39.4 26-Mar 04-Apr 38.8 37.5 41.5 38.7 28.4 37.0 05-Apr 14-Apr 35.4 37.5 41.4 38.4 24.7 35.5 15-Apr 24-Apr 29.9 34.0 39.0 38.8 25.8 33.5 25-Apr 04-May 26.5 31.5 35.3 35.6 24.7 30.7 05-May 14-May 26.8 37.0 32.9 34.0 23.6 30.9 15-May 24-May 23.5 48.0 32.0 35.6 25.3 32.9 25-May 03-Jun 23.1 46.6 28.6 38.5 22.7 31.9 04-Jun 13-Jun 24.1 46.1 37.8 41.5 23.1 34.5 14-Jun 23-Jun 24.2 47.0 38.6 39.1 26.9 35.2

As discussed, one of the main variables

determining the price of an option is volatility. Lower volatility leads to a lower premium being paid for an option, whilst high volatility levels will result in a higher premium being paid for an option on an underlying futures contract. The opportunity therefore is to go short volatility (sell options) over the 10-day period on which volatility historically peaks. Since the direction of market movement cannot be accurately predicted, a short position in both put and call options should be taken. Although market movements tend to be less aggressive after the historical volatility peak, the strike level of the options, in which a short position is taken, should be out-of-the money. This is commonly known as a short option strangle.

Hull (2002:13) defines a strangle strategy as a position being taken in a put and call option with the same expiration date and different strike prices. He states that a short strangle position will be entered into if large price movements are possible. Maximum profit occurs when the underlying futures price on expiration date is trading between the strike prices of the options sold.

The proposed Indexed strangle strategy will therefore aim to profit from volatility trends, in addition to an average price. This will be implemented in the following way: • An average long futures price will be realized by

buying equal volumes of futures contracts on a daily basis over the whole of the contract lifetime.

• Since volatility tends to peak annually over the period 27 December – 5 January, short strangles

will be implemented during this period for the total number of tonnages to be hedged via the average price strategy. This volume of grain divided by the number of trading days between 27 December and 5 January will determine the daily number of short strangles to be entered into. Should the mentioned formula not result in a round hundred number, it will be rounded off and on the last trading day the resultant strangles will be entered into. The rule to be used is that an option will not be sold if the premium of the option is lower than R1/ton, since this is the breakeven value to offset broking fees.

• The strike of the call options to be sold will equal the SAFEX futures price for the July white maize contract as on 27 December (or the first trading day thereafter) plus 40%. The strike of the put options to be sold will equal the SAFEX futures price for the July white maize contract as on 27 December (or the first trading day thereafter) minus 40%. The resultant strike level will be rounded off to the nearest twenty rand interval. It is important to consider the effect of the short

options. • If, on option expiry, the July white maize futures

contract closes higher than the put option strike and lower than the call option strike the total amount of the option premiums will be realized as profit and deducted from the average long futures price.

• If, on option expiry, the July white maize futures contract closes higher than the call option strike, short futures contracts will be assigned against the short call options. These short futures


154

contracts will offset the long futures contracts entered into through the realization of an average price. As such no futures position will exist and the difference between the long- and short futures levels plus the option premiums will be the resultant cash flow per ton.

• If, on option expiry, the July white maize futures contract closes lower than the put option strike, long futures contracts will be assigned against the short put options. As such a double-up of tonnages will arise. In this instance, the average long position price

level can be calculated as follows: AP = (LF + PS - OP)/2

where AP = average price of long futures

position (per ton)

LF = average of daily long futures

position (per ton)

PS = strike level of short put options

OP = sum of option premiums (per ton)

Example:

Average daily long futures position = R1

235/mt

Strike level of short put options = R1 000/mt

Call option premium = R50/mt

Put option premium = R40/mt

AP = (R1 235 + R1 000 – R90)/2

= R1 072.50/mt

Table 4. Summary of Indexed strangle strategy applied on July white maize data from 2001 to 2006

Year Average Price Index Indexed Strangle Strategy Value gained (R/ton) Total value gained*

2001 R 739.40 R 713.94 R 25.46 R 15,273,209.30

2002 R 1,254.24 R 1,200.85 R 53.39 R 32,033,589.74

2003 R 1,400.50 R 1,393.89 R 6.61 R 3,966,000.00

2004 R 1,086.16 R 1,028.64 R 57.52 R 34,510,852.94

2005 R 810.74 R 764.25 R 46.49 R 27,893,010.75

2006 R 958.11 R 907.68 R 50.43 R 30,259,477.20 * Based upon a processor procuring 600 000 metric tons of maize annually

The results indicate that the benchmark index is outperformed in all of the years under review, as the realized procurement price is lower than the Average Price Index. The extent by which the Indexed strangle strategy beats the market varies substantially, from R6.61/metric ton in 2003 up to R57.52/metric ton in 2004. 7. Evaluation of risk management

strategies versus a benchmark In section 6 three price risk management strategies were discussed and applied on historical market data. The resultant performance of each strategy was compared to the benchmark average July white maize SAFEX price. Even though all three strategies

compared favorable to historical average SAFEX prices, the consistency and extent by which the benchmark average price was outperformed differed significantly.

By comparing the results of the three strategies against one another, the optimum single or combination procurement strategy can be identified. This will be achieved by comparing the consistency of performance as well as the extent by which the benchmark is outperformed.

Figure 7.1 graphically presents the 6-year results of the proposed strategies versus one another and the benchmark average price from 2001 to 2006. This is summarized in table 7.1.

R 700.00

R 750.00

R 800.00

R 850.00

R 900.00

R 950.00

R 1,000.00R 1,050.00

R 1,100.00

R 1,150.00

R 1,200.00

R 1,250.00

R 1,300.00

R 1,350.00

R 1,400.00

2001 2002 2003 2004 2005 2006

Average Price Index

Momentum strategy

Maximum price strategy

Indexed strangle strategy

Figure 1. Comparison of price-risk management strategies vs. benchmark average price


155

Table 5. Comparison between performances of price risk management strategies vs. benchmark

The amount per metric ton by which the average price of the July contract is outperformed/

underperformed over the period 2001 to 2006, is presented in figure 2.

Figure 2. Price-risk management strategies performance vs. benchmark (R/metric ton)

From figure 2 it can be concluded that all three strategies are successful in reaching their objective. This statement is made since 17 out of the 18 strategy applications were able to outperform the average price over the last six years on which the data were tested. It is particularly true of the Maximum price- and Indexed strangle strategies, since both these strategies constantly achieved long futures positions at a price

lower than the average price used as benchmark. These results are directly opposite the efficient market hypothesis and previous research done on this subject (Irwin, Good, Martines-Filho & Hagedorn 2005; Stewart 1934:415-433; Wang 2001:929-952).

The benefits and disadvantages of each individual risk management strategy are summarized in table 6.

Table 6. Benefits and disadvantages of individual price-risk management strategies

Strategy Benefits Disadvantages Momentum strategy

No cost (premium) payable Easy to implement

Inconsistency of results (versus benchmark) Small Rand-value benefit versus benchmark average price No guaranteed maximum price

Maximum price strategy

Maximum price is known at inception of the strategy Consistency in performance versus benchmark High Rand-value benefit versus benchmark average price Relatively easy to implement

Premium payable

Indexed strangle strategy

No cost (premium) payable Easy to implement High Rand-value benefit versus benchmark average price

Price movement outside of short option strike levels results in double hedging/no hedging No guaranteed maximum price

R 65.13

R 6.61 R 0.79 R 1.76R 11.86

R 1.77 R 2.97 R -1.60

R 26.11

R 152.21

R 108.80

R 180.87

R 23.19

R 50.43 R 46.49 R 57.52R 53.39

R 25.46

R -5.00

R 20.00

R 45.00

R 70.00

R 95.00

R 120.00

R 145.00

R 170.00

R 195.00

2001 2002 2003 2004 2005 2006

Momentum strategy

Maximum price strategy Indexed strangle strategy


156

Since the results achieved by these strategies are not consistent with the efficient market hypothesis, the question remains as to why these strategies are able to consistently outperform the market. The most obvious reasons for the superior performance of the individual strategies are as follows: Momentum strategy:

No long positions are taken during bearish market movements, whilst futures are bought once prices move upwards. As such, short-term price trends are captured resulting in a favorable net long position price. Maximum price strategy: Call-options are bought on the day on which the strategy is implemented resulting in a maximum price equal to the sum of the call-option strike and the option premium. Price movements lower than the call-option strike are exploited whilst prices higher than the discussed strike level are nullified by the call-option itself. Since volatility tends to start at low levels and increase over time, the call-options are initially bought at a low premium and the subsequent rise in volatility will be exploited once long call-options are liquidated in favor of long futures contracts. Indexed strangle strategy:

Options are sold during the period over which volatility tends to peak, resulting in a high option premium to be subtracted from the average price realized by going long equal volumes of futures contracts every trading day over the contract-lifetime. Since market movement is limited from the days on which the options are sold to option expiration, the net amount of the premiums will be realized. 8. Conclusion Speculators, and more specifically for the purposes of this paper, hedgers, are currently unable to enter into contracts on the futures exchange in a manner that will minimize the impact of price volatility on their earnings. Indirectly this has a negative impact on consumers of the commodity underlying the futures/options contract.

Even though derivative instruments are available to use as a counter against market fluctuations, the price-risk management success of groups with a concern on SAFEX suggests that this is not achieved as yet, ultimately to the detriment of consumers. The view exists that markets are efficient and the return offered by the futures exchange cannot consistently be outperformed. This paper argues the exact opposite, since the use of futures/options strategies result in returns superior to that of the market. Two of the proposed price risk management strategies outperformed the market in every year under review, which is exactly opposite to popular belief of efficient markets. This is achieved by minimizing price

volatility and gaining from short-term market trends. By applying these strategies to their procurement models, processors will benefit from below-average prices. In turn, this may have a favorable impact on food inflation.

Finally, the following recommendations could assist users of the futures market, particularly processors, in lowering the impact of market movement: 1. Personnel concerned with SAFEX should be

educated on the use of derivative instruments in order to increase their knowledge.

2. Greater emphasis should be placed upon the development of core/satellite risk management strategies which will ultimately result in procurement models based on an indexing strategy.

3. The procurement function should in part be outsourced to companies specializing in exotic options based on the expectation of achieving average prices.

References 1. Berchtold, F. & Norden, L. 2005. Information flows

and option bid/ask spreads. Journal of Futures

Markets, 25(12):1147-1172. 2. Bodie, Z., Kane, A. & Marcus, A.J. 2002. Investments.

5th edition. New York: McGraw-Hill. 3. Botha, E. 2005. The use of derivatives by South

African agricultural co-operatives to hedge financial risks. M.Com-dissertation, University of South Africa, Pretoria.

4. Brigham, E.F., Daves, P.R. & Gapenski, L.C. 1999. Intermediate Financial Management. 6th edition.

Florida: The Dryden Press. 5. Carhart, M.M. 1997. On persistence in mutual fund

performance. Journal of Finance, 57-82. 6. Chabane, N. 2003. Markets, efficiency and public

policy – an evaluation of recent influences on price in the maize market and government responses. Research project, University of the Witwatersrand, Johannesburg.

7. Decision Commodities. 2006. Education-commodity markets. Available from: http://www.decisioncommodities.com/education/index.html (Accessed 14 January 2006).

8. Erb, C.B. & Harvey, C.R. 2005. The tactical and strategic value of commodity futures. Research project, Duke University, Durham.

9. Hagedorn, L.A., Irwin, S.H., Good. D.L., Martines-Filho, J., Sherrick, B.J. & Schnitkey, G.D. 2003. New Generation grain marketing contracts. Research project, University of Illinois, Urbana-Champaign.

10. Hull, J.C. 2002. Fundamentals of futures and options

markets. 4th edition. New Jersey: Prentice Hall. 11. Irwin, S.H., Good, D.L., Martines-Filho, J. & Batts,

R.M. 2006-02. The pricing performance of market advisory services in corn and soybeans over 1995-2004. Research project, University of Illinois, Urbana-Champaign.

12. Irwin, S.H., Good, D.L., Martines-Filho, J. & Batts, R.M. 2006-03. The pricing performance of market advisory services in corn and soybeans over 1995-


157

2004: A non-technical summary. Research project, University of Illinois, Urbana-Champaign.

13. Irwin, S.H., Good, D.L., Martines-Filho, J. & Hagedorn, L.A. 2005. The pricing performance of market advisory services in corn and soybeans over 1995-2003. Research project, University of Illinois, Urbana-Champaign.

14. JSE refer to Johannesburg Securities Exchange 15. Johannesburg Securities Exchange. 2006. Proposal to

trade a 30-day average pricing contract as an alternative hedging tool. JSE Agricultural Products Market Notice number A665.

16. Johnson, T.C. 2004. Rational Momentum effects. Journal of Finance, 57:585-608.

17. Kohls, R.L. & Uhl, J.N. 2002. Marketing of

Agricultural Products. 9th edition. New Jersey: Prentice Hall.

18. Madura, J. 2000. International financial management. 6th edition. Florida: South-Western.

19. Pension Funds plan to sue WJ Morgan. 2003. The Star

Business Report, October 2003.

20. Ryan, S.G. 2002. Financial Instruments and

Institutions: Accounting and Disclosure rules. New Jersey: John Wiley & Sons Inc.

21. Skerrit, P. 2002. Derivatives. Johannesburg: Peter Skerrit & Associates.

22. Spurgin, R. 1999. A benchmark for commodity trading advisor performance. Research project, Clark University, Amherst.

23. Stewart, B. 1934. The profits of professional speculators. Economic Journal, 44(175):415-433.

24. The Star refer to Pension Funds plan to sue WJ Morgan

25. Tiger Brands. 2002. Group results for the year ended 30 September 2002.

26. Tiger Brands. 2003. Tiger Brands Annual report. 27. Tongaat Hulett. 2003. Tongaat Hulett Annual report. 28. Valsamakis, A.C., Vivian R.W. & du Toit G.S. 2003.

Risk management. 2nd edition. Johannesburg: Heinemann.

29. Wang, C. 2001. Investor sentiment and return predictability in agricultural futures markets. Journal

of Futures Markets. 21(10):929-952.


158

SUBSCRIPTION FORM

TO "CORPORATE OWNERSHIP & CONTROL" AND "CORPORATE

BOARD: ROLE, DUTIES & COMPOSITION"

Copy this form and follow guidelines to fill it up. I would like to buy (underline what you choose):

For individual subscribers:

1. Journal of Corporate Ownership & Control:

1.1. Printed version of the journal and version on CD - US$248 / €210.

1.2. Printed version of the journal and its electronic version - US$238 / €190.

1.3. Printed version of the journal - US$220 / €180.

1.4. Version on CD - US$180 / €150.

1.5. Electronic version - US$165 / €140.

2. "Corporate Board: role, duties & composition":



2.3. Version on CD - US$155 / €115.

2.4. Electronic version - US$150 / €105.

For institutional subscribers:

1. Journal of Corporate Ownership & Control:



1.3. Version on CD - US$750 / €480. 1.4. Electronic version - US$730 / €470.

2. "Corporate Board: role, duties & composition":



2.3. Version on CD - US$690 / €460. 2.4. Electronic version - US$650 / €450.

Underline one of the payment methods you prefer, and write amount to pay (if you prefer, you can pay by one cheque/bank transfer to subscribe to both journals): 1. I enclose a cheque for US$ / €_______;

2. Send me an invoice for US$ / €_______ (indicate currency correctly).

Write your contact details here:

Name_________________________________________________________________________

Position________________________________________________________________________

Institution_______________________________________________________________________

Address________________________________________________________________________

E-mail_______________________________Tel_____________________Fax________________

Please, send this form (with a cheque if you prefer to pay by cheque) at:

Dr. Alexander Kostyuk Publishing house "Virtus Interpress" Postal Box 36 Sumy 40014 Ukraine Ask for invoice by fax at +380-542-611025 (if you want to pay by bank transfer). Enter our web-site to download the subscription form at www.virtusinterpress.org. If you prefer, you can subscribe through our global subscription agent – SWETS INFORMATION SERVICES or Globo Libri – our agent in Italy.

Corporate Ownership & Control / Volume 6, Issue 3, Spring ...

Documents

Transcript of Corporate Ownership & Control / Volume 6, Issue 3, Spring ...