Governing Enterprises in China: Corporate Boards, Ownership and Markets (Dynamics of Asian Development) 9811631158, 9789811631153

Citation preview

Dynamics of Asian Development

Zhang Cheng Rajah Rasiah Kee Cheok Cheong

Governing Enterprises in China Corporate Boards, Ownership and Markets

Dynamics of Asian Development Series Editors Anthony P. D’Costa, College of Business, The University of Alabama, Huntsville, AL, USA Rajah Rasiah, University of Malaya, Kuala Lumpur, Malaysia Editorial Board Tony Addison, UNU-WIDER, Helsinki, Finland Amiya Bagchi, Institute of Development Studies Kolkata, Kolkata, India Amrita Chhachhi, Institute of Social Studies (ISS), Erasmus University Rotterdam, Rotterdam, The Netherlands Akira Goto, National Graduate Institute for Policy Studies (GRIPS), Tokyo, Japan Barbara Harriss-White, Oxford University, Oxford, UK Keun Lee, Department of Economics, Seoul National University, Seoul, Korea (Republic of) R. Nagaraj, Indira Gandhi Institute of Development Research, Mumbai, India Rene E. Ofreneo, Center for Labor Justice, University of Philippines, Quezon, Philippines Ma Rong, Peking University, Beijing, China Ashwani Saith, Institute of Social Studies (ISS), Erasmus University Rotterdam, Rotterdam, The Netherlands Gita Sen, Indian Institute of Management, Bangalore, India Andrew Walter, University of Melbourne, Melbourne, Australia Christine Wong, University of Melbourne, Melbourne, Australia Achin Chakraborty, Institute of Development Studies Kolkata, Kolkata, West Bengal, India

The series situates contemporary development processes and outcomes in Asia in a global context. State intervention as well as neoliberal policies have created unusual economic and social development opportunities. There are also serious setbacks for marginalized communities, workers, the environment, and social justice. The rise of China, India, and new dynamism of South Korea, Indonesia, and Vietnam in East and South East Asia have given a new meaning to Asian development dynamics. Japan’s energetic ties with India and Vietnam, Korea joining the OECD’s Development Assistance Committee, and China and India’s investments and foreign aid in Africa and Latin America are some of the new processes of development whose impact transcends the vast Asian region. Globalization compounds uneven development, affecting macroeconomic stability, internal and international migration, class and caste dynamics, gender relations, regional parity, education and health, agriculture and rural employment, informal sector, innovation possibilities, and equity. Thus the series views development studies as an unfinished agenda of economic, social, political, cultural interactions, and possible transformations in a fluid policy and global contexts. The editor, with the assistance of a distinguished group of development scholars from Asia and elsewhere specializing in a variety of disciplinary and thematic areas, welcomes proposals that critically assess the above-mentioned wide-ranging developing issues facing Asian societies. With Asia’s contemporary transformation, the series promotes the understanding of Asia’s influence on the prospects of development elsewhere. The editor encourages interdisciplinary, heterodox approaches within the social sciences, and comparative work with solid theoretically informed empirical research. Critical development policy debates in Asia and regional governance issues that have a bearing on development outcomes are also sought.

More information about this series at http://www.springer.com/series/13342

Zhang Cheng · Rajah Rasiah · Kee Cheok Cheong

Governing Enterprises in China Corporate Boards, Ownership and Markets

Zhang Cheng School of Accounting Nanjing University of Finance and Economy Nanjing, China

Rajah Rasiah Asia-Europe Institute University of Malaya Kuala Lumpur, Malaysia

Kee Cheok Cheong Asia-Europe Institute University of Malaya Kuala Lumpur, Malaysia

ISSN 2198-9923 ISSN 2198-9931 (electronic) Dynamics of Asian Development ISBN 978-981-16-3115-3 ISBN 978-981-16-3116-0 (eBook) https://doi.org/10.1007/978-981-16-3116-0 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Although corporate governance theories are strongly characterized by a focus on the scope of operations (resource dependent theory), monitoring (agency theory), bargaining (power theory), other governance mechanisms (stakeholder theory), and regulations (institution theory), Western theories claim that autonomous boards with private ownership and market institutions provide the best corporate governance framework for firm performance.1 Hence, despite increasing marketization, China’s introduction of company boards have often been criticized for the lack autonomy from government control, especially those involving state-owned enterprises. However, such arguments have often been questioned whenever particular economies crash to send contractionary waves to countries linked through investment and trade. Hence, increased focus on corporate governance have often been among the earliest responses by scholars to major regional and global financial crises. Consequently, the Organization for Economic Co-operation and Development (OECD) enacted a series of corporate governance principles following the Asian Financial crisis of 1997–1998 to guide its member and non-member countries to evaluate and improve the governance of their legal, economic, and political system. Similarly, the contagion from the implosion of the United States economy, (which arguably is the world’s most market-oriented economy), not only derailed several other economies, it also brought back concerns over market-oriented corporate governance. Thus, it is clear that good corporate governance, regardless of market system, has a critical role to play to protect investors and regulate the smooth development of capital markets across the world. While corporate governance continues to play an important role in economic development, corporate governance issues are more complex in transition economies, such as China than in the developed market economies. Consequently, policy makers in transition economies have been busy attempting to establish a sound corporate governance system appropriate to their specific country situations. This book uses China as a laboratory to examine corporate governance issues in transition economies, as China is not only the most populous economy in the world, it is also 1 See

Veblen (1919), Nelson and Winter (1982), Coase (1992), North (1990) and Rasiah (2011) for a cogent account of institutions and institutional change. v

vi

Preface

undergoing transition from a socialist to a market economy since 1978. The need for strengthening corporate governance in the country has become all the more pressing as rapid economic growth has transformed the country into the world’s second largest economy. Of no less significance is the fact that some of its enterprises have grown to be among the largest in the world. How they are governed therefore have global consequences. As will be elaborated later, China’s corporate governance model (see Liao, 2017) is distinct from the predominant Anglo-Saxon models (see Veblen, 1919; North, 1990; Williamson, 1985; Mueller, 2006; Ahmad and Omar, 2016) ones in several ways. First, it has a two-tier board system, considting of a supervisory board and a board of directors. Second is the prevalence of large shareholders, regardless of whether enterprises are state or non-state. And third is the important role played by state enterprises that are being transformed by successive reforms. Of the many reforms, China experienced a watershed in corporate governance when the split-share structure reform was introduced in 2005 with far reaching consequences. Hence, we analyse in this book the impact of this reform on board composition, firm performance, and firm risk over the period before and after the introduction of the split-share reform. Using the classification of firms by ownership into central government, local government, state owned enterprise (SOE), and privately controlled firms, the book deploys static and dynamic panel data estimation methods to examine the determinants of corporate board composition, relationships between corporate governance mechanisms and firm performance, and between corporate governance mechanisms and firm risk. The results show that Chinese corporate board composition is jointly determined by the scope of operations, monitoring, bargaining, other governance mechanisms, and regulations. While the government has been the critical player in constituting board composition before the split-share structure reform was introduced, independent directors became more important after that. Whereas private firms and SOEs are more concerned about cost than other factors when appointing independent directors, corporate board exerts a positive influence on firm performance once the share of independent directors reached a certain threshold. Central government controlled firms show outstanding accounting and market performance. While increasing corporate board size reduced firm risk, its independence, state ownership and ownership concentration increased firm risk after the reforms. Nevertheless, although the evidence calls for a shift from government control to market-orientation, it also shows that a mix between the two may be a good option for transition economies. We take this opportunity to acknowledge several members without whose support the preparation of this book would not have been possible. The book is essentially drawn from Zhang Cheng’s doctoral thesis, especially the empirical data and the econometric regressions. Formost, we would also like to thank Zhang Cheng’s parents, Zhang Qi and Li Aiju for the encouragement they have given her to persevere and ultimately finish her doctoral thesis. We would like to extend our gratitude to the late Dr. Che Hashim Hassan, who served as her initial supervisor until his demise, to fellow travellers, some more senior, on her academic journey Aslam Mia, Zhang Chen (Grace), Wang Nan (Zoe), Wang Qianyi, Ibrahim Mohammed Adamu, Qian

Preface

vii

Long Kweh, Li Ran and Zhang Miao for their valuable comments and suggestions while preparing the book. Academic staff at the Faculty of Economics and Administration, notably Prof. Goh Kim Leng and Dr Lim Kian Peng have to be thanked for their constructive comments and guidance at critical junctures. Nanjing, China Kuala Lumpur, Malaysia Kuala Lumpur, Malaysia

Zhang Cheng Rajah Rasiah Kee Cheok Cheong

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Fundamentals of Corporate Governance . . . . . . . . . . . . . . . . . . . . . . . 1.2 Background of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 China’s State Enterprise Reform by Stages . . . . . . . . . . . . . . . 1.2.2 China’s Capital Market Reform . . . . . . . . . . . . . . . . . . . . . . . . 1.2.3 Split-Share Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.4 Legal Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.5 Corporate Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Motivation for Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Problematizing Corporate Governance . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Outline of Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 3 3 5 6 7 8 9 11 12 13

2 Theoretical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Critical Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Agency Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Institutional Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Stewardship Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.4 Stakeholder Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.5 Power Circulation Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Key Concepts and Past Empirical Evidence . . . . . . . . . . . . . . . . . . . . 2.3.1 Board Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Board Independence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Supervisory Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Board Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.5 CEO’s Influence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.6 Ownership Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.7 Split-Share Structure Reform . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.8 Controlling Shareholders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Hypotheses Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Scope of Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15 15 15 16 18 19 20 20 21 21 21 23 25 26 30 34 36 37 38 ix

x

Contents

2.4.2 Monitoring Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.3 Bargaining Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

38 39 40

3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Conceptual Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Analytical Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Research Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Hypotheses Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.4 Sample and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.5 Statistical Robustness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43 43 43 45 47 47 48 53 60 63 65

4 Determinants of Board Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Description of Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Variations in Corporate Governance Mechanisms During Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Board Size and Board Independence During the Reform . . . . . . . . . 4.3.1 Determinants of Board Size and Board Independence . . . . . 4.3.2 Determinants of Board Size . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Determinants of Board Independence . . . . . . . . . . . . . . . . . . . 4.3.4 Determinants of Board Size and Board Independence During Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.5 Determinants of Board Size and Board Independence Before and After Split-Share Structure Reform . . . . . . . . . . . 4.3.6 Determinants of Board Size: A Comparison of Controlling Shareholders . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.7 Determinants of Board Independence: Comparing Controlling Shareholders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

67 67 68 70 72 72 75 77 80 80 82 84 86

5 Corporate Governance Mechanisms and Firm Performance . . . . . . . . 89 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 5.2 Influence of Corporate Governance Mechanisms on Firm Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 5.2.1 Linear Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 5.2.2 Nonlinear Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 5.3 Joint Effect of Board Structure, Ownership Structure and CEO Characteristics on Firm Performance . . . . . . . . . . . . . . . . . . 98 5.4 Influence of Controlling Shareholders on Firm Performance . . . . . . 100 5.5 Controlling Shareholders and Corporate Governance’s Impact on Firm Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Contents

xi

5.6 Corporate Governance Mechanisms and Firm Performance: Before and After Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 5.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 6 Corporate Governance Mechanisms and Firm Risk-Taking . . . . . . . . 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Influence of Corporate Governance Mechanisms on Firm Risk-Taking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.1 Linear Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.2 Non-linear Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Joint Effects of Board Structure and Ownership Structure on Firm Risk-Taking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Influence of Controlling Shareholder Types on Firm Risk-Taking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Influence of Corporate Governance Mechanism on Firm Risk-Taking: Comparison of Controlling Shareholders . . . . . . . . . . . 6.6 Influence of Corporate Governance Mechanisms on Firm Risk-Taking Before and After Reform . . . . . . . . . . . . . . . . . . . . . . . . . 6.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Synthesis of Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.1 Determinants of Board Composition . . . . . . . . . . . . . . . . . . . . 7.1.2 Corporate Governance and Firm Performance . . . . . . . . . . . . 7.1.3 Corporate Governance and Firm’s Risk-Taking Conduct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Implications for Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Implications for Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Directions for Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

113 113 113 113 115 116 118 122 124 125 127 128 128 129 130 130 133 135

Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Appendix C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Appendix D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

About the Authors

Zhang Cheng, Ph.D. is Professor at School of Accounting, Nanjing University of Finance and Economics. Rajah Rasiah, Ph.D. is Distinguished Professor of Economics at the Asia-Europe Institute, University of Malaya. His research interests include technology and international development, foreign investment, human capital, public health, and environment. He holds external positions at UNU-MERIT, Oxford University, Cambridge University, and Universiti Tenaga Nasional. He is the 2014 recipient of the Celso Furtado Prize awarded by the World Academy of Sciences for advancing the frontiers of social science thought. Kee Cheok Cheong is Senior Advisor at the Asia-Europe Institute, University of Malaya. His research interests include economic development, transition economies, employment and poverty, and international economic relations. Some of his selected publications include Revisiting Malaysia’s Population-Development Nexus: The Past In Its Future (UMPress) and Government-Linked Companies And Sustainable, Equitable Development (Routledge).

xiii

Symbols and Abbreviations

After Before Central CSRC FE GMM Local Max MDB Mean Min MOSOE N OECD OLS P25 P50 P75 Private RandD SASAC Sd SOE

After the split-share structure reform Before the split-share structure reform Central government controlled firm China Securities Regulatory Commission Fixed Effect Generalized Method of Moments Local government controlled firm Maximum value Malaysia Development Berhad Mean Value Minimum Value Market-oriented SOE/SOE entity controlled firm Number Organization for Economic Co-operation and Development Ordinary Least Squares 25 percentile 50 percentile 75 percentile Private firm Research and Development State-owned Assets Supervision and Administration Commission of the State Council Standard Deviation State Owned Enterprise

xv

List of Figures

Fig. 3.1 Fig. 3.2 Fig. 3.3 Fig. 3.4 Fig. 3.5 Fig. 4.1

Fig. 4.2

Fig. 5.1 Fig. 5.2 Fig. 5.3 Fig. 6.1 Fig. 6.2

Conceptual framework. Source prepared by authors . . . . . . . . . . . Flow of analysis. Source Prepared by authors . . . . . . . . . . . . . . . . . Analytic framework 1. Source Prepared by authors . . . . . . . . . . . . Analytic framework 2. Source Prepared by authors . . . . . . . . . . . . Analytic framework 3. Source Prepared by authors . . . . . . . . . . . . Board size, 2000–2012. Source Plotted by the authors, it also appeared on “Determinants of Board Composition and Corporate Governance in Chinese Enterprises Since Reforms Began: A Comparison of Controlling Shareholders” . . . Changes in Board independence. Source Plotted by the authors, it also appeared on “Determinants of Board Composition and Corporate Governance in Chinese Enterprises Since Reforms Began: A Comparison of Controlling Shareholders” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Board independence and ROA. Source Plotted by the authors . . . . Board independence and ROE. Source Plotted by the authors . . . . Board independence and Tobin Q. Source Plotted by the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Board independence and firm risk. Source Plotted by the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . State ownership and firm risk. Source Plotted by the authors . . . .

44 45 46 47 48

74

74 97 97 98 117 118

xvii

List of Tables

Table 3.1 Table 4.1 Table 4.2 Table 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 4.9 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 6.1 Table 6.2

Summary of variables, definitions and measurement . . . . . . . . . . Descriptive statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Correlation matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . t-test of mean differences, key variables, 2000–2004 and 2008–2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . t-test of mean differences, key variables, state and private enterprises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determinants of board size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determinants of board independence . . . . . . . . . . . . . . . . . . . . . . Determinants of board size and board independence, before and after split-share structure reform . . . . . . . . . . . . . . . . . Determinants of board size: a comparison of controlling shareholders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determinants of board independence, comparison of controlling shareholders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corporate governance mechanisms and firm performance, Linear estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corporate governance and firm performance, nonlinear estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . U-test, board independence and firm performance . . . . . . . . . . . . Joint effect of ownership structure, board structure and CEO characteristics on firm performance . . . . . . . . . . . . . . . Effects of different controlling shareholders on firm performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corporate governance and firm performance, a comparison of different controlling shareholders . . . . . . . . . . . . . . . . . . . . . . . Corporate governance and firm performance, before and after reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corporate governance and firm risk-taking, linear estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corporate governance and firm risk-taking, nonlinear estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

54 68 69 71 73 76 78 81 83 85 91 94 96 99 101 105 109 114 116 xix

xx

Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 6.7

List of Tables

U-test, board independence, state ownership and firm risk . . . . . Joint effect of board size and ownership concentration on firm risk-taking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of controlling shareholder types on firm risk-taking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of corporate governance mechanism on firm risk-taking, a comparison of controlling shareholders . . . . . . . . . Influence of corporate governance mechanisms on firm risk-taking, a comparison of before and after the reform . . . . . . .

117 119 120 123 124

Chapter 1

Introduction

1.1 Fundamentals of Corporate Governance Corporate governance is a key to corporate success since good corporate governance is associated with better firm performance and firm value (Brown & Caylor, 2004; Conheady et al., 2014; Sami et al., 2011). Weak corporate governance systems may lead to corrupt practices, currency depreciation and stock market crashes. The Asian financial crisis that struck in July 1997 is a case in point whereby inefficient corporate governance frameworks derailed rapidly the East and Southeast Asian economies of South Korea, Malaysia, Thailand, Indonesia and the Philippines (Johnson et al., 2000; Rasiah, 1998). Campos et al. (2002) produced empirical evidence to show that 80% of investors were willing to invest in companies with good corporate governance at a premium price. The Organization for Economic Co-operation and Development (OECD) enacted a series of corporate governance principles in 1998 to guide its member and non-member countries to evaluate and improve their legal, economic and political system to improve corporate governance. Thus, it is clear that good corporate governance has a critical role to play to protect investors and regulate the smooth development of capital markets across the world. The concept of corporate governance can be explained from multiple perspectives. From a narrow perspective, corporate governance can be viewed as the way investors assure themselves of getting a return for their investments (Shleifer & Vishny, 1997). From a broader perspective, corporate governance can be viewed as a structure of rights and responsibilities facing parties that have a legitimate interest in firms (Aoki, 2000), including shareholders, creditors, managers, employees, government and society as a whole. However, most conceptualized corporate governance systems that we know of come from the advanced capitalist economies. Corporate governance systems in transitional economies, such as China, Russia and Vietnam, remain a blurred, complex and controversial issue. Specifically, the Anglo-American corporate governance model, which is known as the market-oriented corporate governance model, is one

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0_1

1

2

1 Introduction

in which corporate ownership is highly dispersed among numerous small shareholders. Monitoring and financing tasks are mostly undertaken by highly developed capital markets. Such a model is considered by neoclassical economists as ideal as no one shareholder can exert monopolistic pressure on the governance of such corporations. However, the German-Japanese corporate governance model or the bank-based model is one in which most of firms’ ownership is highly concentrated in the hands of large shareholders. Despite the highly developed structure of capital markets, banks still undertook most of the financing and monitoring activities in such large corporations (Goergen et al., 2008). While there obviously has been serious problems faced by state-owned enterprises, simply exposing enterprises to the discipline of markets can be debilitating as can be seen from the Russian experience (Chang & Nolan, 1995). Also, corporate scandals that occurred in market economies, such as Enron, WorldCom and Goldman Sachs in the United States and Toshiba in Japan, suggest that there is no specific structural model that is insulated from financial disasters. There is also evidence that that firms with more independent boards and greater institutional ownership experienced worse stock returns during the global financial crisis period of 2007– 08 (Erkens et al., 2012). Indeed, the claim that Asian management styles are rife with bad corporate governance has been proven wrong with the scandals that rocked Enron and WorldCom. Being a transitional economy since economic reforms began in 1978, markets are still evolving in China, which, using the neoclassical dictum, will be inefficient as it has not evolved yet the capacity to check corporate scandals (Zhang et al., 2019). This issue can be seen in Chinese enterprise Kelon’s rise and fall during the split-share structure reform. At the same time, when markets have not fully evolved do not necessarily mean that corporate innovation does not exist. Indeed, Cao (2019) argued in favour of the employee-shareholding model that is worthy of emulation. This point of view is shared by Mitchell (2018: 77) who, in elaborating on what he referred to as “reciprocal convergences”, introduced “indigenous Chinese ideas about corporate social responsibility as the vehicle that might embody these values”. Another empirical finding is that Chinese enterprises’ market valuations had not adequately reflected their corporate governance performance, at least as of 2007–08 (Cheung et al., 2008). Hence, to explore and make better-known how to improve Chinese corporate governance is worthy of scientific research, especially so when is little consensus on the topic. Wu (1994) had emphasized on the check-and-balance relations between three main corporate participants: shareholders, managers and board directors. He suggested that a perfect corporate governance system should clearly explicate the responsibilities, power and the interests of shareholders, managers and board directors. In this context, Zhang (1999) focused on the role of corporate ownership arrangement on corporate governance by arguing from a narrow perspective that corporate governance issues can be viewed as a series of institutional arrangements regarding the corporate board’s functions, structures and shareholder’s rights. However, from a broader perspective, it can be viewed as a system of law, culture and institutional arrangements to allocate corporate control rights and the residual claim rights.

1.1 Fundamentals of Corporate Governance

3

Whereas Li (2000) focused on stakeholders’ interests by arguing that corporate governance is a governance mechanism for corporate owners to monitor corporate managers, from a broader perspective, it can be considered as a series of formal and informal relationships involving both internal and external mechanisms to balance corporate interests, including shareholders, creditors, employees, government and society. Lin (1997) emphasized the role played by markets on corporate governance and stated that corporate governance is an institutional arrangement for corporate owners to monitor managers and corporate performance. The fundamental structure of corporate governance requires both indirect control and external governance through market competition. While markets have a major role to play, business law and management schools emphasize the important function the legal system plays in corporate governance and in the process declare that markets alone cannot ensure the smooth functioning of good corporate governance. They propose that the effectiveness of corporate governance cannot be guaranteed without a strong legal system. Therefore, the government must set up and enforce the legal rules to protect the interests of shareholders, creditors and contractual performance, as well as stakeholders they serve (La Porta et al., 2000; Porta et al., 1996, 1997; Zhang & Rasiah, 2015).

1.2 Background of the Study While the interest on corporate reforms in general, and state-owned enterprises in particular, has been on the rise, the evidence advanced so far has been mixed. Studies that portray China as a success tend to generalize from isolated cases and anecdotal evidence (see Zhang & Rasiah, 2015) or from reductionist regressions that offer little rigour (Zhang et al., 2019). The development of capital markets in China has already evolved extensively, and hence, the focus of this book is not only whether it has been viable but on its impact on the Chinese economy. We provide some background discussion to locate the study in perspective.

1.2.1 China’s State Enterprise Reform by Stages Since President Deng Xiao Peng launched his open door policy in 1978, China has gradually transformed its economy from a centrally planned one to that which is characterized by coordination between central, provincial and municipal (county) governments and increased operations of the market (see Zhang & Rasiah, 2015). The market and ownership approaches are two competing approaches that have been used to carry out China’s split-share structure reform. The former approach was based on the belief that the performance of state enterprises can be improved without ownership transfer so long as the product, labour and corporate takeover markets are established and well-functioning. The latter emphasizes that private ownership is

4

1 Introduction

necessary for the improvement of state enterprises’ efficiency, and hence, it is necessary to privatize state-owned enterprises to improve efficiency (Qiang, 2003). While there are extensive theoretical and empirical works that suggest that privatization improves firm performance (Frydman, 1997; Loc et al., 2006; Megginson & Netter, 2001; Saul Estrin & Evžen Koˇcenda, 2009), there are also considerable works that suggest otherwise, including with evidence coming from Russia and China (Nolan & Chang, 1995; Popov, 2007). However, instead of complete privatization like Russia and Eastern Europe (Sun et al., 2002), China’s state enterprise reform took on a gradualist approach with control of critical SOEs still held by the state (Zhang & Rasiah, 2015). Moreover, as a transition economy, China’s enterprise system is quite different from other countries since it started from being wholly state-owned and controlled before undergoing varying degrees of privatization, which has led to the diversification of enterprises’ control.1 Prior to 1978, the majority of Chinese enterprises were state-owned or held as collective entities. However, due to the problems of state control and the objectives of setting up a socialist market economy, China’s state-owned enterprises (SOEs) have experienced several stages of reforms. The first stage (1978–1984) of reforms was characterized by an expansion in enterprise autonomy. At this stage of reform, most Chinese economists believed that market reforms are more suitable for China because it was expected to attract scientific management from the participation of professionals. The Chinese State Council started to reform the SOEs by expanding enterprise autonomy in pilot enterprises in Sichuan province in October 1978, marking the commencement of Chinese enterprises reform and regime reform. In May 1984, the State Council issued the “Regulations for Further Expanding Autonomy of State-owned Enterprises,” to offer SOEs more autonomy in production and profit retention. The government then started promoting the Management Responsibility Contract System (Cheng Bao Ze Ren Zhi) in large SOEs since January 1987, with management authority transferred to the enterprises and allowed them to retain some of their profits (Su, 2005). By the end of 1987, about 80% of the large- and medium-sized SOEs had adopted the Management Responsibility Contract system with almost all of the SOEs adopting it by 1989. The Leasing Management Responsibility Contract system was introduced for small SOEs. A few large SOEs adopted the joint-stock system, together with the “dualtrack price system”, which allowed them to sell their products at market prices (Qian, 2000). The dual-track price system took on an intermediate price system between the existing planned pricing system and the free market price system. Although reforms at this stage brought some positive outcomes, state enterprises still did not perform as well as the joint venture enterprises,2 township and village 1 According to the China Securities Regulatory Commission (CSRC), the ultimate controlling share-

holders are the investors who (i) hold directly or indirectly 50% of the total outstanding shares, (ii) control directly or indirectly 30% of the total voting rights, (iii) can use the voting rights to select more than 50% of board directors, (iv) have a significant influence over the decision-making in shareholder’s meeting and (v) other situations recognized by CSRC. 2 Joint venture is a business enterprise undertaken by two or more persons or organizations to share the expense and profit of a particular business project.

1.2 Background of the Study

5

enterprises3 owing to low efficiency levels, resulting in most SOEs being unable to survive local market competition following the loss in government’s preferential policy and financial support. In the early 1990s, a lot of state enterprises experienced rollover risks, which threatened the stability of China’s banking system. Consequently, the Chinese government and economists started to realize that ownership structure reform ought to be made an integral part of future reform. The second stage (1993–2003) was characterized by the establishment of the Modern Enterprises System, which was initiated by the Third Plenum of the Fourteenth Party Congress in November 1993 that passed “the Decision on Issues concerning the Establishment of a Socialist Market Economic Structure”. Since then the direction of China’s SOE reform pursued the Modern Enterprises System with clearly defined property rights, specified responsibilities and authority, separation of government from enterprise management and scientific management, which resulted in the privatization of small SOEs on a large scale from 1995. By the end of 1996, over half of the small SOEs were privatized. This was followed by the privatization of SOEs that followed the slogan, “grasping the large, letting go the small” (Zhua Da Fang Xiao), which was promoted by the Fifteenth Party Congress in 1997 (Qian & Wu, 2003). In order to manage the existing SOEs, the Tenth National People’s Congress affirmed in 2003 the establishment of the State-owned Assets Supervision and Administration Commission (SASAC) to supervise SOEs and set the future direction of state enterprises reform. The SASAC was set up to to manage state assets, personnel and operations of SOEs on behalf of the state. The third stage (2004-present) is characterized by further development of the Modern Enterprises System following the launching of the SASAC in 2004. SASAC at central government level is responsible for managing important SOEs that are crucial to national security and the lifeline of the national economy, while others are managed by SASAC at the provincial and local government levels. These developments were then accompanied by capital market reforms.

1.2.2 China’s Capital Market Reform China’s economic reforms experience suggests that SOEs have sought to diversify their funding channels, which has paved the way for the emergence of the Chinese capital market. The establishment of the Shanghai and Shenzhen Stock Markets in 1990 and 1991 represents a major milestone of China’s capital market development. The capital market has grown rapidly since the promulgation of the Securities Law in 1998. Over the past two decades, the number of Chinese-listed firms increased

3 Township,

village enterprises are collectively owned enterprises located in townships or villages.

6

1 Introduction

from 10 in 1990 to 2489 in 2013, whereas the total market capitalization of the shares reached 23,907.7 billion yuan in 2013.4 Following reforms, Chinese domestic shares can be divided into tradable shares and non-tradable shares (split-share structure). Tradable shares can be traded freely in Chinese stock markets, including common A shares, B shares, H shares and other shares. The rest of the shares are non-tradable shares that cannot be traded without the approval of the relevant regulatory authority, including state shares and legal person shares. Each kind of shares has the same voting and cash flow rights, but differs in the nature of shareholders involved. State shares are held by central and local governments and their agencies, including state asset management bureaus, financial ministries and SOE entities (Qiang, 2003). Legal person shares are held by domestic institutions, such as securities companies, trust and investment companies, funds and foundations, which are ultimately owned by the state. A shares are issued by Chinese companies listed on the Shanghai and Shenzhen stock exchanges and are traded in the Chinese currency, the Renminbi (RMB). B shares are exchanged in either US dollars on the Shanghai Stock Exchange or in Hong Kong dollars on the Shenzhen Stock Exchange. They are only traded among foreigners and people from Hong Kong, Macao and Taiwan. Meanwhile, H shares are listed in Hong Kong, while N shares are listed in New York. Other foreign shares are listed in Singapore and London (Firth et al., 2007). In order to regulate the Chinese stock market, the State Council set up the China Securities and Regulatory Commission (CSRC) in October 1992 as the main regulatory commission of China’s stock markets.

1.2.3 Split-Share Reform Despite massive economic boom in the period 2001–2005, Chinese stock markets performed dismally over those years. The benchmark Shanghai Stock Exchange Composite Index dropped more than half (Jiang et al., 2008). Consequently, the split-share structure caused several problems that discouraged non-tradable shareholders from obtaining financing through stock markets, which seriously depressed the development of Chinese capital markets. Three key problems have been identified to have caused the problems. Firstly, a serious conflict of interests arose between tradable shareholders and non-tradable shareholders due to the pricing mechanism as the non-tradable shareholders could not benefit from an increase in the market price of the shares. Secondly, non-tradable shareholders tend to expropriate tradable shareholders’ interests through related party transactions, such as asset sales and product purchase. Thirdly, most of the tradable shareholders emerged as speculative investors, who have little knowledge of the stock markets and the companies they invested in with their focus limited to short-term returns rather than long-term capital 4 In

2020, China’s Alibaba and Tencent were ranked seventh and eighth by market capitalization (https://www.statista.com/statistics/263264/top-companies-in-the-world-by-market-capita lization/).

1.2 Background of the Study

7

gains. These problems explain the high turnover ratio in Shanghai and Shengzhen stock markets, which was around eight times more than that in the United States, United Kingdom and Japan. The stock return volatility in Chinese stock markets then was among the highest in the world (Liao et al., 2014). The government launched the split-share structure reform in 2005 to address the above problems, which has become an important milestone in China’s privatization and transformation towards a market-oriented economy. The reform required nontradable shareholders to negotiate with tradable shareholders to arrive at a mutually acceptable compensation to gain liquidity in the stock market. The China Securities and Regulatory Commission (CSRC) issued the Circular on Relevant Issues Regarding Pilot Programs for Non-tradable Share Reform of Listed Companies to start the split-share structure reforms in April 2005. The first batch of four pilot firms undertook non-tradable share reform in May 2005, which was initiated by negotiations between corporate board of directors and non-tradable shareholders. It was then followed by a second batch of 42 pilot firms. The reform became mainstream when the CSRC, SASAC, Ministry of Finance, People’s Bank of China and the Ministry of Commerce jointly issued the Guideline on the Non-tradable Share Reform of Listed Companies, and CSRC issued the Measures on the Non-tradable Share Reform of Listed Companies in 23 August 2005. By the end of 2007, 98 percent of all listed firms had completed the split-share structure reform that represented 98 percent of total market capitalization in China (CSRC, 2008).

1.2.4 Legal Infrastructure The importance of the legal system in corporate governance was studied by La Porta et al. (1998) who ranked 49 countries on the basis of quality. Under this framework, Allen et al. (2005) found that the quality of China’s legal system is significantly below average among the developing countries due to poor legal enforcement and severe corruption. Also, Kato and Long (2006) found that China lacks a comprehensive set of legal rules and regulations to protect the interests of shareholders. In addition, the Chinese government caused uncertainity as it has retained the power to intervene in the enforcement of the law, which demonstrates that China does not have an independent judicial system. Jiang and Kim (2015b) argued that one of the main reasons for China’s weak legal environment is the light punishment imposed on offenders. The security law stipulated various fines and penalties for the violation of securities law in Articles 188–235, but most of the fines are light, being only between 30,000–300,000 RMB and 60,000–600,000 RMB. For example, Danhua Chemical Technology paid 1.5 billion RMB to its related party without holding a board meeting or shareholder meeting when it was listed during 2003–2006. The firm also reported several false bank deposits of 205 million and 479 million RMB. However, the firm only received a warning and a small fine of 300,000 RMB for these wrongdoings. Moreover, the executives behind those fraudulent activities were only fined 30,000–300,000 RMB (Jiang & Kim, 2015b).

8

1 Introduction

Despite these criticisms, China has taken a number of little steps to improve its legal system. For example, a company law was first promulgated in December 1993, which was amended several times in 1999, 2004 and 2005. Meanwhile, the security law was first promulgated in 1999 and revised in 2006. Whereas company laws deal with the Chinese corporate structure, the responsibility, the liability of the listed company, shareholders, directors and managers, security laws address the rules regarding share issues, exchanges and corporate acquisition, as well as penalties for corporate wrongdoings. In 2001, the Code of Corporate Governance for Listed Firms in China was set up in China based on the principles enshrined in the Organization for Economic Corporation and Development (OECD). In 2001, the Guidelines for Establishing an Independent Directors System for Listed Firms stipulated that corporate boards must have one third of independent directors by the end of June 2003.

1.2.5 Corporate Markets In an effective corporate control market, strong firms can take over weak firms. Its effective administration requires that a powerful corporate governance mechanism is in place since managers have to work hard to avoid losing their jobs and reputation during takeovers. However, such well-organized takeover markets are largely absent in China due to several factors. The first is that state enterprises cannot sell freely without government permission. Furthermore, the concentrated ownership structures often make takeovers difficult. Besides, both independent and non-independent directors do not have incentives to allow corporate takeovers. Consequently, many studies attribute the dominance of inactive corporate markets to concentrated ownership structures (Allen et al., 2005; Liu, 2006). In an effective labour market, managers have incentives to work hard to improve their reputation, so that they will get the opportunity to be employed by other companies. However, such markets hardly exist in China as managers of SOEs are appointed by the government, and therefore, they are naturally more likely to uphold government interests rather than the market. This is somewhat paradoxical as it contradicts the purported aim of the government to marketize these enterprises. Apart from that, most private enterprises are family firms, which prefer to choose managers from within their relatives or families rather than from the market (Jiang & Kim, 2015b). In product markets in China, managers in both SOEs and private enterprises have incentives to work hard to make sure that the firm succeeds in such markets as otherwise, they will lose their jobs and political career (Jiang & Kim, 2015b).

1.3 Motivation for Study

9

1.3 Motivation for Study This study chooses China, the world’s largest transition economy, as a research laboratory to study corporate governance issues. As a transition economy, it faces more complex and controversial ownership and performance issues than capitalist economies. Corporate governance systems developed by capitalist economies, such as the Anglo-American model and the German-Japanese model are not perfect models. But they have evolved from and function within a market-based framework. Being a transition economy with weak legal infrastructure, inactive labour and product markets and characterized by frequent political interventions, China’s corporate governance system is much more complex than its capitalist economy counterparts and is still very much a work in progress. The corporate scandals that occurred during the period of China’s share structure reforms and privatization efforts suggest that there is a long way to go for China to make its corporate governance reforms effective. Unlike Russia, which jumped to a market economy immediately when the Soviet Union collapsed (Bramall, 1995), China has undergone gradual institutional change as reforms have created a corporate environment that is unique. China has followed a gradual or a partial path to transform its economy from a centrally planned to a market economy since 1978. The trial and error approach adopted has given China the space to correct any failures that have resulted from the policies introduced (Zhang & Rasiah, 2015). During this process, some state enterprises have been fully privatized, while others have been partially privatized. The split-share structure reform in 2005 was instrumental in introducing elements of privatization in Chinese state enterprises, resulting in diversified institution arrangements, including state and private control (Zhang et al., 2019). Nevertheless, state firms continue to enjoy financial privileges not available to private firms (Wei et al., 2014). Since government has provided political support, banks are willing to lend to state firms (Liang et al., 2012). Furthermore, state enterprises select CEOs largely based on social and political objectives, whereas private enterprises choose CEOs largely based on their ability to maximize shareholder’s wealth (Berkman et al., 2012). Too much should not be made of the state enterprise and private enterprise distinction however. When the People’s Republic of China was established as a command economy, all enterprises belonged to the state. All decisions related to enterprise development were by the central government. Enterprise managers had little discretion beyond meeting production quotas. Employee compensation was also determined centrally. As a consequence, enterprise managers had no management incentives to increase production (Guo et al., 2013). It was not until Deng Xiaoping liberalized the economy in 1978 that non-state enterprises began to emerge. These were collective enterprises, the so-called township and village enterprises (TVEs) set up by local governments that were more competitive than state enterprises (Fu & Balasubramanyam, 2003). Even then, critics steeped in neoliberal thinking had been vocal in planning China’s state sector. Indeed, Chang (2010) concluded that: “China cannot make much progress toward (the rule of law), at least as long as the

10

1 Introduction

Communist Party is around.” and by extension, China’s numerous state enterprises needed to be reformed through privatization or liquidation.5 China did not collapse nor did the significance of its state enterprises diminish but a series of state enterprise reforms has produced private/non-state enterprises and a much transformed state enterprise sector. Of the former, Kanamori and Zhao (2004) traced the emergence of China’s private sector today to the move from political struggle to socialist modernization and reconstruction announced by the Third Plenum of the Eleventh Congress of the Communist Party of China, together with a series of announcements directed at agriculture and the establishment of individual enterprises (getihu 个体户) but with the public economy as its base. Further policy announcements saw these enterprises grow and enjoy greater freedom. In the 1990s, large-scale privatizations of state enterprises occurred under the premiership of Zhu Rongji as part of state enterprise reform, while President Jiang Zemin’s announcement of the Principle of Three Representations (San ge Daibiao, 三个代 表) incorporated private enterprise as an important social class. Thus, far from being distinct from public enterprises, China’s private enterprise sector is a construct of, and implicitly subordinate to, the state sector. On the latter, Curtis Milhaupt, Professor at Stanford Law School, observed that “Until fairly recently, most scholarship on Chinese corporate governance, including the governance of its huge, publicly listed state-owned enterprises, tended to stay close to the law on the books and was framed almost exclusively in comparison to Western companies, which … missed the intricacies of how the system actually operates.” (Lee, 2019). He added “Outsiders often think of China’s so-called state-owned enterprises as completely owned by the government and insulated from standard Western corporate practices… But those firms, whether China Mobile or Sinopec, are actually “mixed-ownership” enterprises with both government and private shareholders. They’re a hybrid entity, neither completely state-owned nor completely private.” Despite this, there exist characteristics among major state enterprises that set their corporate governance apart from private enterprises. First is the significance of social and non-economic objectives that may conflict with the interests of non-state owners. These leave the owners vulnerable to the overriding dictates of the state. Second, since state owners do not oversee day-to-day operations of state enterprises, their managers may pursue self-serving behaviour at the expense of the state and minority owners (Jiang & Kim, 2020: 735). Thus, state enterprises face a vertical agency problem just like private sector companies in developed countries. This is an area that successive state enterprise reform had tried to remedy. China’s capital market development has been slow since it emerged from 1990 to 1991 when the Shanghai and Shenzhen stock markets were established. Until the enactment of security laws in 1999, the legal status of China’s stock market was not formalized. Also, before the split-share structure reform of 2005, the effective

5 In

Gordon Chang’s book The Coming Collapse of China (London: Arrow 2001), Chapter 3 was titled “State Enterprises are Dying” and Chapter 7 “The State Attacks the Private Sector”.

1.3 Motivation for Study

11

functioning of the Chinese stock market was hindered by the dominance of nontradable shares that experienced serious impasse. China’s domestic-listed firms have a unique ownership structure that has left them vulnerable to serious agency problems. Chinese firms were characterized by concentrated ownership, split-share structure and state ownership. Chen, Firth and Xu (2009) reported that on average the largest shareholder owns 43.8% of enterprise’s equity in China. In contrast, Fan and Wong (2002) reported that the average largest shareholder ownership in the East Asian countries is only 25.8% of total equity. At the same time, non-tradable shares tightly controlled by the government make up more than two thirds of equity issued, leaving less than a third of the total shares to be freely traded. These non-tradable shares have impeded China’s capital market development until the split-share structure reform was launched in 2005. Even so, state ownership is still dominant in most of China’s listed firms, though it has been reduced significantly by reforms since. For example, Yang et al. (2011) found that by the end of 2009, more than 50% of the listed firm’s shares were ultimately owned by the state.

1.4 Problematizing Corporate Governance The corporate board, as the link between shareholders and managers, serves the most important role in corporate governance. How to manage the corporate board composition is a critical question that has received considerable debate in both developed and developing countries. The debate is far more intriguing in transition economies where like in China the state is still largely in control of a market that has been allowed to function to stimulate and appropriate positive synergies from capitalist growth. However, in attempting to discuss China, it will be a mistake to classify China as a single extractive polity as claimed by Acemoglu and Robinson (2012).6 Not only is China diverse geographically with its Western and Central states much less developed and topographically more rugged than the East, the former is also less characterized by Confucian culture than the latter (Zhang & Rasiah, 2015). Also, almost all key institutional instruments are located in the East. Planning in China is characterized by central initiation, provincial intermediation and municipal (and county) implementation (Zhang & Rasiah, 2015), which has been a major driver of differential development in China. Being relatively new to modern economics, transition economies, including China, studies on corporate governance, and the functioning and effectiveness of boards are enduring topics but have produced little consensus. Hence, explaining what determines the corporate board composition is crucial to understand the roles the corporate board can play in a firm’s decision-making process. Especially when

6 See

Zhang & Rasiah (2015) for an incisive critique of this argument.

12

1 Introduction

an economy is experiencing the introduction of regulations, including when undergoing a transition, the corporate board plays an important role in the enforcement of government regulations. In addition, China’s corporate governance has its own unique characteristics, which obviously calls into question its (firm-level corporate government mechanisms and corporate board structure) impact on firm performance and value creation during the transition period. This topic has been a great concern to investors, creditors, government and society. For example, a two-tier board structure with a main board and a supervisory board has evolved, which is similar to the structure of GermanJapanese corporate governance model. Paradoxically, however, the main duty of the Chinese supervisory board is to supervise the board of directors, but they do not have the right to dismiss the board of directors (Shan & Round, 2012). In addition, despite lacking the authority to hire and fire directors, the corporate board functions as a decision-making maker and is the highest authority in determining corporate operations and investment plans. The extent to which the corporate board and other participants would like to shoulder risks when they make a decision is critical to corporate success and long-term development. Last but not least, the way China is transforming its corporate governance system during economic transition remains a controversial issue and thus is worth studying.

1.5 Objectives This book seeks to examine the impact of economic reforms on the introduction, as well as corporate governance in China with specific focus on the corporate board structure, ownership and markets, and how institutional change has transformed governance and performance of enterprises. In doing so, it seeks to achieve the following objectives: 1.

2.

3.

4.

Identify the determinants of corporate board size and independence against a myriad of controlling shareholder types and during different periods of Chinese split-share structure reform. Examine the relationship between corporate governance mechanisms and firm performance among different controlling shareholder types and during different periods of Chinese split-share structure reform. Investigate the relationship between corporate governance mechanisms and corporate risk-taking among different controlling shareholder types and during different periods of Chinese split-share structure reform. To draw implications of the above findings for state enterprise reform.

In doing so, this book seeks to deepen our understanding of corporate governance mechanisms as most of previous studies have largely limited their findings using mainstream perspectives grounded on neoclassical theory. The empirical base is China, a large transition economy that has undergone a unique set of reforms,

1.5 Objectives

13

which is radically different from the market economies to which most extant theories have been applied. For example, the agency theory assumes that there is a conflict of interests between shareholders and managers. In contrast, China is characterized by a collectivistic culture that is more likely to have a principle–steward relationship, in which a trustworthy management team is important. Especially in private-controlled firms, an interpersonal relationship is a major factor that influences business transactions. Furthermore, a thorough analysis of corporate governance can benefit government policy-makers to further Chinese SOE reforms and corporate governance reforms. Since China’s SOE reform approach includes privatization, this analysis can shed light on the effectiveness of this approach by comparing corporate governance efficiencies between state- and private-controlled firms and between the periods before and after the reform. Lastly, this book seeks to fill a serious lacuna in existing works on China on the determinants of board composition by introducing the influence of controlling shareholders and state enterprise reform. Furthermore, it seeks to extend existing works on corporate board governance and risk-taking in China by introducing the independent director’s risk preference in different controlling shareholder categories.

1.6 Outline of Book This book is structured as follows. Following the introductory chapter, Chap. 2 reviews the extant literature on corporate governance with a special emphasis on China. Chapter 3 discusses the methodology and data used for analysis in the book. Chapter 4 is the first analytical chapter, which investigates the determinants of board composition in China. Chapter 5 then assesses the role of corporate governance by examining the link between corporate governance and firm performance. Chapter 6 looks further at the relationship between corporate governance and firm-level risktaking. Chapter 7 presents the synthesis of the empirical findings in the book to draw implications for theory and policy. The chapter also offers directions for future studies on corporate governance in general and on transition economies in particular.

Chapter 2

Theoretical Considerations

2.1 Introduction In reviewing the related literature, this chapter seeks to identify first corporate governance theories that are potentially applicable to China, viz. agency theory, resource dependency theory, institutional theory, stewardship theory, stakeholder theory and power circulation theory. Second, we review empirical works on corporate governance. The rest of the chapter is organized as follows. Section 2.2 reviews corporate governance theories, which is then followed by a profound assessment of empirical works on the topic in Sect. 2.3. The first part of Sect. 2.3 reviews works on the determinants of board composition and in doing so formulates three hypotheses, i.e. operations, monitoring and bargaining. The second assesses the effects of Chinese corporate governance characteristics, including board structure, ownership structure and CEO’s character on firm performance and firm risk taking conduct. Section 2.4 surveys the literature on the role of controlling shareholders in China. Finally, Sect. 2.5 presents the summary.

2.2 Critical Theories Corporate governance has been dominated by Western theories that were formulated from the experience of the developed economies. Although there are examples of work that deal with Japan, the nature of governance in the previously dominant Zaibatsus (prior to World War 2) and the currently dominant Kereitsus has hardly been researched extensively in relation to financial performance. Much of the Japanese works have tended to focus on the nature of control and self-expansion of these business organizations. Owing to the dominance of small firms in Italy, a significant literature has evolved on social corporatist formations that is largely governed systematically at the district rather than firm level (Brusco, 1982; Becattini, 1990; © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0_2

15

16

2 Theoretical Considerations

Piore & Sabel, 1984). Hence, we examine in this chapter theories that have largely evolved from the West.

2.2.1 Agency Theory The agency theory originated from Berle and Means (1932)’s seminal work on the separation of ownership and control and deals with the most common relationships within corporations. Shareholders own the company as the principals, whereas managers control the company’s operations as the agents. Agency problems are generated when shareholders and managers have conflicting interests. Managers and shareholders may have different goals for the corporation. Managers may be primarily concerned with power, wealth and job security, while investors may be seeking the highest possible returns. Manager and shareholders may have different attitudes towards risk. Shareholders may be attracted to high-risk and high-return projects since they can diversify their investment portfolios. However, managers may be reluctant to take high-risk projects due to the potential for failure, which creates employment risk for these executives (Jensen & Meckling, 1976). In order to minimize agency problems, agency theory deals with the most efficient contracts that can govern relationships between shareholders and managers to overcome conflicts of interests (Eisenhardt, 1989). Another type of agency problem exists between large and small shareholders. Since large shareholders undertake most of the corporate governance responsibilities, they may expropriate small shareholders’ interests leaving the latter to face free rider problems (Shleifer & Vishny, 1986). Agency theory argues that as managers control the firm, they may be able to pursue self-benefit rather than that of the owners. Agency theorists focus on identifying governance mechanisms that limit the managers’ self-serving behaviour, and they argue that the corporate board is a key internal governance mechanism of the firm to protect owners’ interests. Therefore, the primary role of the corporate board is to exercise the governance function and reduce agency cost. In doing so, they have to rely on outside directors, who are considered less likely than insiders to collude with managers to expropriate owners’ benefits. Thus, the corporate board acts as an effective monitor that will ensure the management performs in the best interest of the company and contribute to the reduction of agency costs. Agency theory proposes that the quality of monitoring depends on board independence. Independent directors as outsiders have a lower potential for a conflict of interest, and they are viewed as a tool that links the external stakeholders with firms. Unlike other board members, independent directors are presumed to have better motivation to monitor managers and executive directors to avoid their possible selfserving behaviour, since they have no financial interests in the firm. Their personal interests lie in enhancing their reputation as independent experts in the market for directors (Fama & Jensen, 1983). In addition, competition among independent directors encourages them to make every effort to monitor the actions of management. Agency theory advocates that independent directors should hold a majority of board

2.2 Critical Theories

17

seats because they can monitor the agents’ decisions effectively and reduce agency cost. The agency theory is applied under the assumption that individuals are selfinterested, adverse to risk and are subjected to abuse when their interests conflict. Meanwhile, the organization is assumed to have conflicts of interests among corporate participants, and information asymmetry between shareholders and managers (Eisenhardt, 1989). Agency problems can be mitigated through incentive compensation (Tosi et al., 1997; Zajac & Westphal, 1994). However, solutions for agency problems in developing economies may not be effective in economies undergoing economic transition, since new agency relationships are created during the process of transferring state ownership into other forms of ownership (Dharwadkar et al., 2000). According to agency theory, the main function of the corporate board is to monitor managers on behalf of shareholders. Effective monitoring can reduce agency costs incurred when self-serving managers expropriate shareholders’ interests (Hillman & Dalziel, 2003). Hence, the board of directors connecting shareholders and managers has the responsibility to guarantee that managers are acting in the interest of shareholders and ultimately to minimize the potential agency cost inherent in the separation of ownership and control. Besides, effective monitoring is expected to prevent managers from making decisions that harm shareholders’ interests. Tosi et al. (1997) define monitoring as “observation of an agent’s effort or outcomes that is accomplished through supervision, accounting controls, and other devices”. It is assumed that monitoring is free of cost, while any kind of monitoring of an agent will result in gains for the principal, except when an agent’s action may have no negative effects on the principal’s outcomes. Since Salancik and Pfeffer (1978)’s publication of “The External Control of Organizations: A Resource Dependence Perspective”, which was rooted in Thompson (1967)’s open system view of the organization, the resource dependency theory has become a popular organizational theory. The resource dependency theory recognizes the interdependence between the organization and the external environment. Although organizational behaviour is influenced by external factors and constrained by context, organizational controllers can use their power to reduce environmental uncertainty and dependence (Hillman et al., 2009). Galbreath (2005) posits that resources can be categorized as tangible resources and intangible resources firms use to conceive of and implement their strategies. Resources can create advantages for firms when they are unique or valuable, which are generally thought to be intangibles such as reputation, skills capabilities, expertise and know-how. The corporate board is the vehicle through which corporations absorb important external resources. Corporate board size and composition are not randomly constituted and depend instead on the conditions imposed by the external environment (Pfeffer, 1972). Hillman et al. (2000) found that the corporate board is an important linkage between a corporation and its external environment, and firms respond to significant changes in the external environment by altering the composition of the board. Hillman and Dalziel (2003) summarized several specific activities of the corporate board, including (1) providing expertise, advice and counsel; (2) linking

18

2 Theoretical Considerations

the firm to key stakeholders and other important entities; (3) building external relations and diffusing innovation and (4) aiding in the formulation of strategy and other important firm decisions. The second key role of a corporate board is that of resource provider. The extent to which directors benefit the firm depends on whether their inclusion provides access to valued resources and information, reduces environmental dependency or aids in establishing legitimacy. The resource provider function comes in the form of advice, counsel, skills and access to key information that can impact on a firm’s strategies and legitimacy (Pfeffer & Salancik, 1978). Here, the board acts not so much as a monitor but rather is actively involved in influencing strategy and programs through the board members’ skills and expertise (Hillman & Dalziel, 2003). According to Hillman and Dalziel (2003), the monitoring and resource providing roles are most effective when the necessary human resources are in place at the board level. Wright et al. (2002) demonstrate that boards that contain outsiders are more effective in monitoring because they have the requisite human resources to objectively evaluate management strategies and the ability of management.

2.2.2 Institutional Theory The key idea behind institutionalization is that much organizational action reflects a pattern of doing evolves over time and becomes legitimated within an organization and an environment. Therefore, it is possible to predict practices within organizations from perceptions of legitimate behaviour derived from cultural values, industry tradition and firm history. North (1990) defined institutions as “the rules of the game in a society, or more formally, the humanly devised constraints that shape human interactions”. Firms and organizations are players that respond to a blend of socioeconomic and political institutions (Veblen, 1919). The differential performance of economies over time is fundamentally determined by the way institutions evolve. Institutional theory offers a broader range of influences on corporate governance than agency theory by arguing that it is shaped by institutional embeddedness (Nelson & Winter, 1982). According to institutional theory, organizations are the way they are for no other reason other than the fact that they represent legitimately organized bodies to solve collective action problems (Rasiah, 2011). Critics of agency theory and supporters of institutional theory have pointed out its “under-contextualized” nature and hence its inability to accurately compare and explain the diversity of corporate governance arrangements across different institutional contexts (Aguilera & Jackson, 2003). First, agency theory overlooks the diverse identities within the principal–agent relationship, including different types of investors such as state, banks and families. Second, it overlooks the importance of interdependencies among other stakeholders in the firm. Third, it retains a narrow view of the institutional environment influencing corporate governance (Aguilera & Jackson, 2003). Institutional theory suggests that firms may appoint outside directors for regulative, normative and cognitive reasons. China enacts legal requirements for

2.2 Critical Theories

19

listed companies to appoint outsiders on their board since 2001 when CSRC issue the guidelines. The major institutional transition in China focuses on the corporate governance of traditional SOEs, which are directly supervised by the government and do not have boards of directors. However, since officials and bureaucrats are agents of the government they often lack incentives to push SOEs to focus on performance. Consequently, transformation of SOEs to state-owned joint-stock companies was initiated in the 1980s with the explicit aim to enhance performance. Since the opening of the Shanghai and Shenzhen stock exchanges in 1990 and 1991, respectively, only jointstock companies with boards are allowed to list their shares. Uribe–Bohorquez (2018) examined the board independence and firm financial performance relationship based on the impact of the formal and informal institutions of 18 national business systems. They show that while the direct effect of independence is weak, national-level institutions significantly moderate the independence and performance relationship. Their findings suggest that the efficacy of board structures is likely to be contingent on the specific national context.

2.2.3 Stewardship Theory Stewardship theory is different from agency theory in the sense that it assumes managers are not self-interested individuals but have aligned their interests with shareholders (Davis et al., 1997). Managers are viewed as trustworthy stewards interested in achieving high performance and maximizing shareholders’ interests, since they are motivated by intrinsic motivators, such as success, recognition, respect and work ethic (Muth & Donaldson, 1998). The principle–steward relationship is more likely to be observed in collectivist and low-power distance cultures. China is generally considered to have a collectivist and low-power distance culture since its firms and organizations operate in a socialist structure (Tian & Lau, 2001). Stewardship theory assumes that individuals identify with the mission of the organization to pursue organizational goals. If CEOs have a strong sense of responsibility regarding the welfare of the organization, they will act in the interest of the organization, leading to low monitoring costs and fewer conflicts. This, in turn, improves the development of distinctive core capabilities, competitive advantages and leads to higher financial returns over time. When CEOs are also board members, they are more likely to view themselves as stewards and have a strong organizational identity (Donaldson & Davis, 1991). When CEOs’ actual power and self-perception of power are high, the CEO’s organizational identification will also be high. The agency costs will be mitigated since the CEO acts as a steward. In fact, stewardship theory suggests that CEOs act as stewards because they are trustworthy, collectivist and pro-organizational individuals who behave in the best interest of the organization by minimizing their own self-serving behaviour. This implies that powerful CEOs will improve firm operating efficiency, lower financing costs and increase firm productivity.

20

2 Theoretical Considerations

2.2.4 Stakeholder Theory The term “stakeholder” refers to a group of constituents who supply critical resources and have a legitimate claim on the firm. The stakeholder theory views companies consisting of networks of relationships between different stakeholders, which are defined as any group or individual who can affect or be affected by an organization. For example, stockholders, creditors, managers, employees, customers, suppliers, local communities and the public are all stakeholders. The stakeholder theory can be seen as a nexus of contracts between all resource holders, including those in both implicit and explicit contractual relationships (Hill & Jones, 1992). The key task of managers is to create value for stakeholders by aligning the interests of different stakeholders in pursuit of creating mutual interests between these stakeholders, instead of primarily weighting conflicting interests. It is in the broad sense of the management that it also recommends attitudes, structures and practices, that taken together, constitute stakeholder management (Donaldson & Preston, 1995). The basic practical idea of stakeholder theory is that the success of a company is very dependent on smooth cooperation with its stakeholders, so that the company has to follow the needs and wants of these stakeholders. Three aspects of stakeholder theory have been advanced and justified, namely descriptive, instrumental and normative. The descriptive aspect describes the corporation as a constellation of cooperative and competitive interests possessing intrinsic value. The instrumental aspect establishes a framework for examining the connections between the practice of stakeholder management and the achievement of various corporate performance goals. Finally, the most fundamental aspect of stakeholder theory is the normative aspect, which follows the idea that stakeholders are identified by their interests in the corporation and that the interest of all stakeholders have intrinsic value (Donaldson & Preston, 1995). Instead of completely searching for profit and earnings, stakeholders pursue cooperative opportunities and balance conflicting interests. For example, customers prefer lower prices; the employees want higher wages, and the suppliers want higher price.

2.2.5 Power Circulation Theory According to the political theory’s perspective of organization, the firm is seen as a political coalition and the executives as its main political broker. Firm behaviour responds to the interests and beliefs of the dominant coalition. The circulation of power emphasizes the internal contests for control and opposition of CEOs, who exercise their influence through formal authorities and informal power (Ocasio, 1994). Following the power circulation theory, some studies argue that corporate board composition is a result of bargaining between the CEO and the rest of the directors. When the CEO is in a dominant coalition, the corporate board tends to be composed

2.2 Critical Theories

21

of executive directors. In contrast, if shareholders are in the dominant coalition, the corporate board tends to be independent (Combs et al., 2007).

2.3 Key Concepts and Past Empirical Evidence This section reviews the key empirical works on corporate governance. Board structure, ownership structure and CEO’s influence are the most important corporate governance mechanisms and consequently the focus of this book. The following section reviews the influence of these mechanisms on firm performance and firm risk.

2.3.1 Board Structure Anglo-Saxon models have focused on the board structure as the fulcrum of corporate governance. Independence between managers and shareholders, avoidance of ownership dominance, nature of supervision and board size are considered critical in addressing corporate governance.

2.3.2 Board Independence The effectiveness of board monitoring is closely related to board composition. Board monitoring becomes less effective when the board consists primarily of insiders or dependent outsiders, such as current or former managers, employees, directors who have business, family or social relationships with the CEO. Thus, board independence is key to the effectiveness of board monitoring (Lynall et al., 2003). Independent directors, as the outside party, are considered an important corporate governance instrument, serving primarily to monitor managerial behaviour and to stand up for the interests of minority shareholders. Chinese board independence has become an mandatory requirement since the CSRC issued the “Guidelines for introducing independent directors” in 2001, stipulating that by the end of June 2002, Chinese-listed firms must have two independent directors, and by the end of June 2003, Chinese-listed firms must have at least a third as independent directors. According to the CSRC’s regulations, independent directors are those who have no relationship with the company, which may bias its independent judgement. Specifically, (1) immediate family members of senior management are not permitted to work for or own a significant stake in the company or any controlling shareholder stakes, (2) their immediate family members cannot provide consulting services to the company, (3) at least one of the independent directors must have an accounting background, (4) they can only serve a term of three years and a maximum of two

22

2 Theoretical Considerations

consecutive terms, and (5) they cannot serve as an independent director in more than five firms (Zhu et al., 2015). Tricker (2011) suggested that “the more independent a director is, the less he or she will know about the company and its industry. The more a non-executive director knows a company’s business, organization, strategies, markets, competitors, and technologies, the less independent he or she may become.” Non-executive directors can also provide valuable services to boards by offering additional expertise and competencies, broadening their knowledge base for key decisions, contributing actively to the strategic decision-making process and securing access to critical resources. Non-executive directors can help company insiders to lead the firm strategically in the aftermath of the IPO and to deal successfully with the complexities associated with the transition to public company status. Beyond this, non-executive directors can provide access to critical resources. Board independence is viewed as a strong monitoring mechanism preventing opportunistic actions of managers and thus improving the company’s reputation (Fama & Jensen, 1983).Their status, image and reputations depend on the quality of monitoring of management (Fama & Jensen, 1983). Independent directors could contribute their independent views and actively participate in board discussion. They will represent shareholders on the company’s board. As independent persons, they must ensure their presence and performance are free from any influence of insiders or management. The company appoints independent directors to monitor the performance of executive directors and top managements. Therefore, they would pursue the interest of shareholders by maximizing shareholders’ value. Although the Chinese government believes that board independence maximizes shareholders’ interests and improves firm performance, the empirical evidence is mixed. In Wang (2014)’s study of 30 selected studies, only half support the positive effects of board independence on firm performance, whereas the other studies provided contrary results. Wang found that Chinese board independence did not improve firm performance in the studies that produced contrasting results, and the role of independent directors is mainly advisory rather than to monitor. This is because, firstly, independent directors know much less about the companies than the inside directors. Secondly, it is difficult for independent directors to be absolutely independent; in fact, since insiders can select independent directors based on the regulatory requirement, this renders board independence merely as a “quack corporate governance instrument” (Romano, 2004). Ma and Khanna (2015) investigate the role of independent directors and find that independent directors would generally defer to top management as they feel indebted for being offered a director’s position and in exchange independent directors provide support to management. Independent directors are more likely to dissent due to this social exchange relationship breakdown that is when the board chair who appointed the independent director or the director herself is leaving the board. Shan and Xu (2012) suggested that corporate governance reforms relating to corporate board and supervisory board in China are insufficient to ensure that they properly fulfil their roles of oversight. Rather, they are perhaps playing a “window dressing” or “rubber stamp” role within the current two-tier corporate governance system. The independent directors were found to have no impact on firms’ financial performance.

2.3 Key Concepts and Past Empirical Evidence

23

Nevertheless, more recent Chinese board independence studies, such as by Li et al. (2015) and Liu et al. (2015), found consistent positive results between the percentage share of independent directors and firm performance. However, studies conducted in Anglo-Saxon countries have yet to find such a positive relationship. Hsu and Wu (2013) found that in the United Kingdom, the likelihood of corporate failure increases with the proportion of independent directors. Bhagat and Black (2001) also challenge the positive effect of board independence in the United States and found that firms with more independent directors do not perform better than others. Furthermore, Faleye et al. (2011) and Kim et al. (2014) found that board independence did not contribute to firm performance in the United States. However, Black and Kim (2012) provide empirical evidence that the requirement of 50 percent outside directors improved firm performance and market values in Korea. Similarly, Muniandy and Hillier (2014) found that board independence contributes to firm performance in South Africa. Also, Lefort and Urzúa (2008) found that the percentage of outside directors positively affects firm market value and performance. On firm risk, in general, outside directors facilitate firms’ borrowings, information acquisitions and alliance formations. Therefore, board independence is more likely to result in firms with both resources and risks. Besides, independent directors represent the interests of small shareholders, who can be risk neutral since they can diversify their investment, and are willing to accept any project that may gain net present value regardless of the risk (Deutsch et al., 2011). Independent directors can exert influence over corporate risk-taking through monitoring managerial behaviours and advising managers on designing and executing corporate strategies (Li et al., 2015). Brick and Chidambaran (2008) found a negative relationship between board independence and firm risk. Since risky firms are usually associated with greater information asymmetry between managers and shareholders, independent directors find it difficult to monitor risky firms. Minton, Taillard and Williamson (2011) found that larger boards and lower levels of board independence are associated with lower levels of risk-taking. However, Cheng et al. (2010) found that board independence did not affect corporate risk-taking. Besides, Dionne and Triki (2005) suggested that the New York Stock Exchange requiring a majority of independent directors is not necessary for shareholders’ wealth and risk management.

2.3.3 Supervisory Board China’s company law of 1993 stipulates that Chinese-listed firms and non-listed joint-stock firms should adopt a two-tier board structure with a board of directors and a supervisory board. Whereas the board of directors is the decision-making organ, the supervisory board is the monitoring organ. Unlike the supervisory boards in Europe, Chinese supervisory boards and boards of directors are in the same organizational hierarchy; hence, both of them are appointed and are held responsible for shareholders’ meetings, which is the highest power organization in listed firms.

24

2 Theoretical Considerations

The supervisory board should consist of at least three persons, including representatives of employees and shareholders, but excluding managers, directors and financial controllers. The Chinese supervisory board is aimed at making up for the deficiencies of the existing corporate board system, which enjoys independence from the CEO and in some cases from the chairman of the board (Dahya et al., 2002). The main function of the supervisory board is to monitor the board of directors and executives, as well as provide advice for the firm’s operations. China’s 1993 company law prescribed the responsibilities of the supervisory board, but failed to define its legal power boundaries. Unlike German and Japanese supervisory boards, the Chinese supervisory board does not have the right to appoint and dismiss board directors, which weakens its effectiveness as a corporate governance instrument. Xiao et al. (2004) found that a Chinese supervisory board’s function is shaped by various factors, including the firm’s ownership structure, political influence, relationships with independent directors and its own characteristics. Most Chinese supervisory boards perform as honoured guests, friendly advisors, independent watchdogs or censored watchdogs. Dahya et al. (2003)’s interviews of Chinese supervisory boards show that most supervisory boards are not truly independent, which calls for a need to improve the functioning of the supervisory board. Hu et al. (2010) argue that the Chinese supervisory board is hindered by ownership concentration and weak external governance environmental conditions that inhibit its positive governance role to improve firm performance. Chinese company laws established in 1999, 2004, 2005 and 2006 have made several amendments, which have greatly improved the functioning of the supervisory board. Supervisory boards now have the right to propose the dismissal of board directors or top managers, as well as to curb top managers’ compensation. Also, they are allowed to attend board of directors’ meetings and raise questions (Ding et al., 2010). Ding et al. (2010) found that before the enactment of the 2006 company law, supervisory boards did not have a say over managers’ compensation; after the enforcement of company law 2006, however, both supervisory boards’ size and their meetings frequency had a significant impact on managers’ compensation. Now, the supervisory board as approval and monitoring body has duties to examine firm’s financial reports; ensure managers and directors obey the laws and procedures of the firm; correct the behaviour of managers and directors when their action is detrimental to the firm; arrange shareholder meetings; take on other duties as specified in the firm charter. Ran et al. (2015) studied the personal characteristics of supervisory board members and suggest that supervisors’ accounting or academic background, supervisors’ compensation and female supervisors improve accounting information quality in China. Chen and Hamilton (2020) argue that supervisory board could play a role in advancing the interests of corporate stakeholders and improve the transparency of corporate information disclosure, resulting in a positive relationship between overall corporate social responsibility performance and supervisory board size. As the Chinese governance structure combines features of both the German-Japanese and the Anglo-American models, there is a conflict between the main board and supervisory board with respect to monitoring. Wang (2008) claims that independent

2.3 Key Concepts and Past Empirical Evidence

25

directors have made a limited contribution to Chinese corporate governance and that there should be a fundamental regulatory reform to enhance both the overall board effectiveness and the relationship between the board of directors and the supervisory board. Farag and Mallin (2019) find that the independent directors substitute the supervisors’ monitoring role.

2.3.4 Board Size Literature on board size has emphasized that both too large and too small corporate boards have their advantages and weaknesses. Large boards’ diffused responsibility and lack of cohesiveness suffer from increased problems of communication, coordination and decision-making. Small boards suffer from decreased ability to fulfil their function. However, according to the resource dependence theory, larger boards improve the ability to access critical resources through environmental links and therefore increase firm performance. A large board size enables the pooling of a diverse group of experts to strengthen the board’s task performance. Small boards are expected to have higher performance because of their lower agency costs. Yeung (2018) reveals a curvilinear relationship between board size and firm performance for firms listed in major Hong Kong exchanges, and this relationship was moderated by the cultural backgrounds of directors. The proportion of Anglo-American directors on a board negatively moderates this relationship when the board is small and positively moderates the effect when the board is large. The 2005 Chinese company law stipulates that China-listed firms must have about 5–19 board directors. Lipton and Lorsch (1992) argue that limiting board size is beneficial to firm performance; when board size exceeds ten persons, it becomes difficult for directors to hold candid discussions on firm performance. Thus, the optimal board size should be around 8–9 persons. Supporting this, Yermack (1996) found empirical evidence that small board size is more effective in American industrial enterprises by showing the negative relationship between board size and Tobin Q. The incentives for CEOs are found to be stronger with small boards; this is because board size increases incremental cost such as coordination, communication and free riding. Research on Finnish firms by Eisenberg, Sundgren and Wells (1998) supported the negative relationship between board size and firm performance. Chen et al. (2008) also suggested that a small board improved firms’ accounting and market performance when takeover intensity is stronger before the United States launched its anti-takeover law. Guest (2009) supports the argument that large boards are ineffective due to poor communication and decision-making in the United Kingdom. Also, Mak and Kusnadi (2005) found that the negative relationship between board size and firm performance exists in Singapore and Malaysia. Khan (2019) found that board size has a significant negative effect on firm performance in Malaysia firms. For firm risk, the negative relationship between board size and firm risk-taking was initially supported by social psychology studies (Moscovici & Zavalloni, 1969), which show that a large group is less likely to make extreme decisions since it takes

26

2 Theoretical Considerations

considerable compromise to reach consensus (Sah & Stiglitz, 1986, 1991). Members within large board groups have different judgement and information processing capabilities that can be costly for communication. Group decisions reflect compromises between different members. With regard to risk-taking, a large group is unlikely to accept risky projects because it is difficult to secure majority acceptance in such a group. Chen (2008) found that in the United States, larger boards are associated with lower variability in firm performance than smaller boards, including stock return, ROA and Tobin Q, as well as other investment and financial activities, such as acquisition and RandD spending. Wang (2012) found that the size of a corporate board is more important than its composition in determining high-risk investments like RandD, though it was not the case with low-risk investments, such as property, plant and equipment. Whereas Nakano and Nguyen (2012) found that a large board reduces performance volatility and bankruptcy risk in Japan, in New Zealand and China, and a large board is associated with greater risk-taking (Huang & Wang, 2015; Koerniadi et al., 2013). Furthermore, Haider and Fang (2016) investigated the relationship between board size and corporate risk-taking in China and show that despite the uniqueness of Chinese corporate governance features, board size is negatively associated with corporate risk-taking. However, the economic impact of change in board size is much smaller than other developed market context. Upadhyay (2015) suggested that large boards pursue conservative investment policies and that their decision outcomes are moderate, which promote an environment of risk aversion.

2.3.5 CEO’s Influence 2.3.5.1

CEO Duality

CEOs in China are not necessarily the ultimate corporate controllers of firms like in the Anglo-Saxon countries. It is common to see general managers functioning as CEOs who lead their respective firms. In most cases, the board chairman has more power than CEOs to control and run firms, since the board chairman is the firm’s legal representative who is selected by the shareholders. CEO duality refers to a practice where the CEO and board chairman positions are held by the same person. CEO duality is a controversial corporate governance phenomenon, which has been hotly debated over the past two decades because of its potential to function as a “doubleedged sword” (Finkelstein & D’Aveni, 1994). On the one hand is the positive “edge”, which can be supported by stewardship theory that rests on the assumption that it brings about a stronger leadership through faster decision-making and improved managerial efficiency. On the other hand, there is the negative “edge”, which can be explained by agency theory as CEO duality gives extra power to CEOs, which can weaken monitoring of managers’ self-serving behaviour (Krause et al., 2013). Also, CEO duality is not a preferable leadership structure because the extra power enables the CEO to hinder board members involved in their monitoring role, for instance by controlling information flows, by setting a rigid board meeting agenda or by creating

2.3 Key Concepts and Past Empirical Evidence

27

norms in which it is inappropriate to question management’s effectiveness. Tuggle et al. (2010) suggest that board members do not maintain constant levels of attention towards the monitoring function. CEO duality reduces the boards’ allocation of attention to monitoring by the control of the meeting’s agenda and location. CEO duality is negatively associated with board monitoring in a privately held firm context. Deman et al. (2018) argue that CEO duality is the result of a large shareholder agency problem. The effect of CEO duality on board members’ involvement in monitoring is contingent on the type of monitoring task and moderated by the firm’s ownership structure. They show that CEO duality is only negatively related to the board’s behavioural control task and this negative effect is contingent on whether ownership is concentrated in the hands of a controlling shareholder as well as the type of controlling shareholder. Lam and Lee (2008) suggested that neither agency theory nor stewardship theory alone can adequately explain the duality and performance relationship. They suggests that the relationship between CEO duality and accounting performance is contingent on the presence of the family control factor. CEO duality is good for non-family firms, while non-duality is good for family-controlled firms. CEO duality becomes a contentious issue when it comes to firm performance. Some studies found that CEO duality has no effect on firm performance (e.g. Baliga et al., 1996; Daily & Dalton, 1992, 1993). However, other studies found contrasting evidence regarding its effects on firm performance. Although some studies found positive effects on firm performance (e.g. Boyd, 1995; Brickley et al., 1997), others found negative effects on firm performance (e.g. Chen et al., 2005; Daily & Dalton, 1994; Rechner & Dalton, 1991). Boyd (1995) argues that CEO duality is beneficial when there is environmental uncertainty since an united command and execution could facilitate firms to speed up the decision-making processes necessary for dealing with environmental uncertainty. On the one hand, when the environment is stable, separate CEO and chairman positions could reduce opportunistic behaviour by the CEO. On the other hand, CEO non-duality generally raises the costs associated with information delivery between the CEO and board chairman (Brickley et al., 1997). He and Wang (2009) suggested that costs are magnified especially when firms need large amounts of knowledge assets; CEO duality enhances the already positive relationship between knowledge assets and firm performance. Chang et al. (2019) report evidence that CEO duality benefits a firm when economic policy uncertainty is high. They propose that CEO duality is an advantageous governance mechanism for coping with economic policy uncertainty. Han et al. (2016) suggested that CEOs are more effective in rapid decision-making but the quality of decision-making may be compromised. A powerful CEO is less likely to receive independent advice or to have her decisions scrutinized. Hsu et al. (2019) investigate the effects of CEO duality on firm performance and the moderating effect of information costs on the relationship between CEO duality and firm performance in Taiwan. They find that CEO duality has statistically significant negative impacts on firm performance when information costs are high. Tang (2017) found that the effect of CEO duality was negative when the CEO had dominant power relative to other executives and when the board had a block holding outside director, but was non-significant otherwise.

28

2 Theoretical Considerations

Yang and Zhao (2014) underscore the benefits of CEO duality in saving information costs and making speedy decisions and find that duality firms outperform non-duality firms by 3–4% when their competitive environments change. Further, the performance difference is larger for firms with higher information costs and better corporate governance. Duru et al. (2016) provide evidence that duality has a statistically significant negative impact on firm performance that is positively and significantly moderated by board independence. They suggests that outside board members serve as effective monitors, limiting managerial opportunism and playing a disciplinary role while exploiting the benefits of decisive leadership associated with a joint board leadership structure. Goergen et al. (2020) show that, first, the most frequent reason for combining the CEO and chairman roles is unified leadership. Second, duality allows them to make the best possible use of the CEO’s in-depth knowledge of the firm and its day-to-day operations. Third, duality enables the CEO to act as a bridge between the management and the board of directors, which promotes the flow of information between the management and board. Besides, CEOs’ risk preference is fundamental for corporate development. Specifically, managers and CEOs’ risk-taking attitude depends on multiple factors, including job security, personal reputation, stock incentives and compensation. Most managers and CEOs prefer projects that possess low risks (March & Shapira, 1987), since they have to be responsible for corporate insolvency and financial distress if projects fail. Kim and Buchanan (2008) found a negative relationship between CEO duality and firms’ risk-taking behaviour. In contrast, Li and Tang (2010) found a positive relationship between CEO duality and firms’ risk-taking conduct. In this case, CEO duality strengthened CEO hubris, which increases risk-taking. There is an increasing trend to separate the positions of CEO and board chairman in China since it could provide more effective monitoring against CEOs’ self-serving behaviour (Yang et al., 2011). According to the power circulation theory, top management tends to use power to dominate the board by choosing loyal directors from within rather than from among outside independent directors; ultimately, this results in avoiding power contests between rivals (Combs et al., 2007). Lew et al. (2018) show that separating the posts of CEO and chair promotes better performance of Chinese manufacturing firms.

2.3.5.2

CEO Tenure

Recent research tends to align CEOs’ tenure with firm performance, which is characterized by life cycles, while these evidences show that CEOs’ contribution towards firm performance tends to take an inverse U-shaped relationship (e.g. Hambrick & Fukutomi, 1991; Henderson et al., 2006; Miller & Shamsie, 2001). Hambrick and Fukutomi (1991) proposed that new CEOs face knowledge shortage at the very beginning, but as they become familiarized with their jobs and organizations to gain sufficient experience, they gradually start to contribute to firm performance. However, after reaching a certain threshold, the CEO becomes increasingly insular with regard to the outside environment so as to overly commit to his or her paradigm, thereby

2.3 Key Concepts and Past Empirical Evidence

29

causing a downtrend in firm performance. Notheless, Henderson et al. (2006) argue that CEOs’ within-firm learning is affected by the stability of the environment as they found that firm performance increases steadily with CEO tenure in the stable food industry, with the downturn occurring only after 10–15 years. In contrast, the CEO starts with the best performance but experiences a steady decline over time in the computer industry (Henderson et al., 2006). Chen and Zheng (2014) found four influences from CEO tenure, namely power, human capital investment, experiences and career concerns. Firstly, CEO tenure is an important instrument for building CEO power, which is defined as the capacity of individual actors to achieve their ends (Finkelstein, 1992). CEOs initially rely on other executives for corporate knowledge and insight, but they develop their leadership skills gradually over time and eventually exert profound influence on the firm (Shen, 2003). Secondly, CEOs’ expand their experience over time. Thirdly, CEOs’ human capital investment in specific firms increases with CEO tenure (Buchholtz et al., 2003). A CEO develops firm-specific knowledge, which is only valued within a particular firm; a long CEO tenure thus constrains the diversity of the CEO’s human capital investment (Chen & Zheng, 2014). Fourthly, CEOs’ concerns over their future career development decrease with their tenure (Gibbons & Murphy, 1991). There are two competing arguments regarding the effect of CEO tenure on corporate risk-taking. On one hand, CEO tenure increases managerial power and experience as his or her social capital and knowledge accumulates over time. It is likely that CEOs with long tenure may have the tacitness and experience to reduce the risks associated with projects (Simsek, 2007). Furthermore, Chen and Zheng (2014) found that career concerns fell as long tenures combined with deepened links with the firm helped increase CEOs’ confidence to undertake risky projects. Also, CEOs with long tenure are more likely to acquire more task knowledge, know better how to run their organizations, are more familiar with decision-making process and have stronger capabilities to lead their firms to pursue risky strategies (Xie, 2014).

2.3.5.3

CEO Age

CEOs’ age typically will have a significant impact on firm performance because their career concerns fall as they approach the age of retirement (Antia et al., 2010). CEOs approaching retirement tend to be more “myopic” than other CEOs; since income from the firms will no longer be there after their retirement, associated CEOs may reject or strategize less good investments because of “myopia” (Campbell & Marino, 1994).1 Besides, the ageing process changes an individual’s physical and psychological condition, which in turn affects their decision-making, strategic choice, risk preference, motivations and leadership. Consequently, Zhang (2010) found that CEOs’ age is negatively correlated with firm growth and market value. CEOs’ age also 1 Unless

the firms involved provide ex-post retirement extensions, which some firms do for exemplary performers.

30

2 Theoretical Considerations

influences risk-taking decisions as Bertrand and Schoar (2002) offered empirical evidence to argue that older CEOs tend to be more conservative than younger CEOs in decision-making.

2.3.6 Ownership Structure 2.3.6.1

Managerial Ownership

Agency theorists suggest that managerial ownership aligns the interests of managers and shareholders, as well as motivates managers to work towards value maximization, leading to improved firm performance (Jensen & Meckling, 1976), which implies that increased managerial ownership can improve firm performance. From a sample of Fortune 500 manufacturing firms, Cho (1998) found that managerial ownership influences firm values through corporate investment. Meanwhile, Chen (2001) found that managerial shareholdings improve firm performance in Chineselisted firms, which was supported by the findings of Li et al. (2007) in privatized state-owned enterprises and firms with high managerial ownership exhibit a smaller performance decline after privatization. Gao and Kling (2008) suggested that giving senior managers stock incentives can help reduce tunnelling and improve firm performance. However, Morck et al. (1988) suggested that when managerial ownership is high, these managers tend to be entrenched in their mindsets, and their actions may be out of the shareholders’ control. Managers may believe that the benefits brought by the perquisites (such as the use of company cars, subsidies in travel and reimbursements of entertainment for business purposes) may outweigh losses in firm value. In addition, high ownership gives managers sufficient power to pursue their own interests without discipline from other owners (Short & Keasey, 1999). Since managers are concerned about their job security, compensation and future career development, most of them are adverse to risk. Coles et al. (2006) found that managerial risk-averse behaviour may reduce firm risk-taking, but equity-based compensation encourages managers to take more risk. Kim and Lu (2011) found that the relationship between CEO ownership and risk-taking is an inverse U-shaped curve, risk-taking increase at low levels of ownership but reduced at high levels of ownership. Florackis et al. (2020) argue that, on the one hand, managerial ownership may have a positive impact on firm value due to its interest alignment effect, but, on the other hand, it may also have a negative impact due to its risk-substitution effect. Managers can choose to pass up innovative projects with high firm-specific risk in favour of the standard projects, which lead to suboptimal corporate strategy and lower firm value. Vijayakumaran (2021) suggested that managerial ownership exerts a positive direct effect on corporate investment decisions by aligning management’s incentives with the interests of shareholders. Managerial ownership acts as a form of credible guarantee to lenders and helps to reduce the degree of financial constraints faced by firms. When managers entrenched through ownership, they do not fear the detrimental effects of risky innovative projects, have an incentive to overinvest in

2.3 Key Concepts and Past Empirical Evidence

31

innovation for reasons of growth, implying higher managers remuneration, power and prestige. Beyer et al. (2012) find that managers holding no company shares underinvest in innovative projects compared with 100% owners. The existent literature suggested that the relationship between firm value and managerial ownership is nonlinear. A higher degree of managerial ownership benefits shareholders because it increases managers’ incentives to increase firm value. However, when managerial ownership exceeds a certain point, it encourages managers to entrench themselves, so that firm value falls when managerial ownership increases beyond a certain point. McConnell and Servaes (1990) found a curvilinear relationship between managerial ownership and firm performance, and the negative effects start from about 40–50% of managerial ownership. Hu and Zhou (2008) argue that the relation between firm performance and managerial ownership is nonlinear, and the turning point at which the relation turns negative occurs at ownership above 50%, and firms with significant managerial ownership outperform firms whose managers do not have ownership. Florackis et al. (2009) suggested that the alignment effect of managerial ownership exists at levels lower than 15%. Kamardin (2014) also found an U-shaped relationship between managerial ownership and firm performance with the turning point at 31.38%. Jelinek and Stuerke (2009) find a nonlinear relationship between managerial ownership and agency cost and suggest that the ability of managerial ownership to reduce agency costs decreases as levels of ownership increase. Benson et al. (2020) argue that firm location may influence the effectiveness of managerial ownership through information asymmetry. Manager and shareholder’s agency conflicts in remotely located firms could be exacerbated due to long distances that result in higher levels of information asymmetry. They found that managerial ownership is more effective in mitigating agency conflicts in rural areas with higher levels of information asymmetry and lower degrees of local trustworthy constituents.

2.3.6.2

Ownership Concentration

The corporate board has the responsibility to support shareholders’ interests, especially when ownership is dispersed. However, in most Asian companies, ownership is usually concentrated in the hands of a few large shareholders (Jiang et al., 2010). Wruck (1989) proposed that ownership concentration strengthens firm value if block holders use their voting power to facilitate value-increasing takeovers or efficiently manage corporate resources. In contrast, ownership concentration is harmful to firm value when block holders insulate managers from market discipline and takeover threats. On the positive side, ownership concentration is an effective corporate governance mechanism that contributes to firm performance, especially when the corporate context lacks market discipline and a sound legal system (Claessens & Djankov, 1999). When ownership is highly concentrated,2 minority shareholders 2 Ownership

concentration refers to the concentration of ownership rights that arejointly owned by several large shareholders.

32

2 Theoretical Considerations

find it costly to monitor management, since minority shareholders share relatively little of the firm’s revenue and they tend to have little knowledge about corporate governance. Most of these investors are “free riders”. Hence, large shareholders who have more interest in the firm assume more responsibilities in monitoring management (Shleifer & Vishny, 1997). If they cannot monitor the management themselves, they can expedite third-party takeovers by sharing their gains with bidders. Therefore, ownership concentration minimizes agency problems between shareholders and managers (Li, 1994). The main agency problem under the concentrated ownership structure is among large shareholders and minority shareholders, due to possible conflict of interests (Yang et al., 2011). Minority shareholders are more flexible in diversifying their investments with a short-term horizon, while large shareholders will be more concerned about the firm’s long-term and sustainable development. Large shareholders usually utilize additional corporate governance mechanisms to monitor the firm, such as improving incentive compensation and conducting direct supervision. In China, large shareholders always involve themselves in major corporate decision-making and managerial processes (Gul et al., 2010). Hence, board independence and ownership concentration tend to act as monitoring substitutes (Lefort & Urzúa, 2008). On the negative side, large shareholders may use their power to expropriate minority shareholders’ interest through collusion with top managers, which is known as “tunneling” that harm firm performance. Heugens et al. (2009) summarized three types of tunnelling practices. Firstly, self-dealing and related-party transactions enable large controlling shareholders to transfer firm resources to other entities. Secondly, large shareholders can increase their share in the firm through minoritydisadvantaging transactions, such as dilutive share issues and insider trading. Thirdly, they can expropriate minority shareholders’ interests by setting up their own compensation at a rate beyond market levels (Johnson et al., 2000). There are also studies which have found a weak relationship between ownership concentration and firm performance (e.g. Omran et al., 2008; Chen et al., 2005). There are also studies suggesting that ownership concentration effectiveness varies across ownership types. Cabeza-García and Gómez-Ansón (2011) argue that ownership concentrated in the hands of private investors has a positive and significant effect on postprivatization efficiency. Janang et al. (2015) found that ownership concentration is not effective corporate governance mechanisms in improving efficiency in Malaysian government-linked companies. Farooq and Zarouali (2016) show that ownership concentration is value relevant only for firms headquartered in nonfinancial centres. The location of a firm’s headquarter in the financial centre can reduce information asymmetries and lower the incentives for concentrated ownership in the financial centres. Hegde et al. (2020) suggested that family firms appear to be weakly associated with positive abnormal returns at high ownership concentration levels. The positive interest alignment effects offset family entrenchment effects on firm performance at high levels of ownership concentration. Theoretically, there are mixed possibilities between ownership concentration and firm risk-taking. On one hand, large shareholders can maximize firm value by promoting high-risk and high-return investments. For example, Koerniadi et al.

2.3 Key Concepts and Past Empirical Evidence

33

(2013) found that in New Zealand, multiple large shareholders took this step to induce higher levels of risk-taking, while Demsetz and Villalonga (2001) found that there is a positive relationship between ownership concentration and firm-specific risk. On the other hand, large shareholders have the authority and incentive to monitor managerial behaviour; managers therefore tend to implement conservative financial strategies in the interest of job security, which often happens when the stock market conditions are poor as large shareholders are less likely to diversify their investment, leading to lower risk-taking to protect their property (John et al., 2008).

2.3.6.3

State Ownership

Due to partial privatization, state ownership remains an important issue in China despite the introduction of economic reforms in 1978. State-owned enterprises have remained important custodians of the government to protect the interests of the masses (Zhang & Rasiah, 2015). However, state ownership leads to more serious agency problems, as state shareholders have the incentive to abuse state assets for their own interest (Wang, 2003). The dominance of state ownership may also result in a divergence in the allocation of capital resources for non-profit uses, such as maintaining higher employment rates, though it could be argued that such firms tend to share the social obligations of the nation. In addition, due to the lack of effective external monitoring over management, the controlling power would ultimately fall into the hands of directors and managers who bear minimal risk for their decisions, as they are largely insulated from the discipline of the market (Oliver et al., 2014). Wei and Varela (2003) found that state ownership is detrimental to firm performance in newly privatized Chinese firms. However, Sun et al. (2002), Yu (2013) and Zhang et al.’s (2019) findings show that the relationship between government ownership and firm performance is an inverted U-shaped curve, which suggests that a certain amount of state ownership is optimal. Consequently, despite criticisms levelled at state ownership for its inefficiency and with that its negative effects on firm performance (Gunasekarage et al., 2007; Hovey et al., 2003; Xu & Wang, 1999), the Chinese government is cautious with its privatization process. Hence, some state-owned enterprises (SOEs) were fully privatized, while others are still controlled by the Chinese government or its agencies, through either retaining ownership before they went public or purchasing tradable shares from the open market. For firm risk, Boubakri et al. (2013) contend that state ownership reduces firm risk-taking, because government shareholders have the objectives of maintaining a lower unemployment rate and social stability. They would endeavour to ensure long tenure of power rather than focus on complete profit maximization; therefore, firms with high state ownership are less likely to engage in risky investments. The government has focused on stabilizing big business groups, which are the key providers of middle-income jobs (Fogel et al., 2008). One could argue that the state’s focus on social benefits through keeping unemployment rates low and stabilizing risks presented by the market may have produced a more stable China compared to the

34

2 Theoretical Considerations

Anglo-Saxon countries, which are often confronted by volatile fluctuations in financial performance, as well as job hirings and firings. State ownership is usually associated with worse financial transparency, greater information asymmetry and poor governance structures. Thus, when cumulated bad news are released suddenly, the stock price may crash. However, state ownership implies that the government will intervene by providing preferential access to credit, tax discounts, implicit guarantees and other forms of financial support (Boubakri et al., 2019). Ding et al. (2020) suggest that the presence of state ownership can reduce crash risk through implicit government guarantees, while altering firms’ information hoarding behaviour and impacts investors’ perception. Xie et al. (2019) find that state ownership is negatively related to stock return volatility, and firms with greater government influence tend to have lower stock volatility. They suggest that government can send credible signals by retaining state ownership, which reduces investor uncertainty. However, investors must weigh the positive signalling effect of residual state ownership in reducing uncertainty, surrounding sudden policy changes, against the inefficiencies of state control.

2.3.7 Split-Share Structure Reform Before the split-share structure reform, the total shares in Chinese domestic A shares stock markets are divided into tradable and non-tradable shares. The existence of nontradable shares is for Chinese government to retain control over important SOEs and for SOEs to raise capital from the stock markets. It has been regarded as the biggest impediment to the development of China’s equity market. Non-tradable shares account for more than two thirds of the total shares that are held by state and legal persons and are not allowed to be freely traded and cannot benefit from any stock price increase. Non-tradable shareholders find it difficult to diversify their investments, and thus, they often engage in corporate governance decisions to tunnel corporate resources and expropriate minority shareholders’ interests, while tradable shareholders could not exert influence over corporate decisions. As a result, the conflict between tradable shareholders and non-tradable shareholders was intense before the reform due to the increasing corporate governance problems and interest conflicts between tradable and no-tradable shareholders. The Chinese government attempted to release these non-tradable shares into the market in 2001, but it ended in failure as the stock market reacted negatively to the huge shares supply. From lessons of past failures, the Chinese Securities Regulatory Commission (CSRC) initiated the split-share structure reform in April 2005. The CSRC required that publicly listed firms take steps to convert non-tradable shares into tradable shares. Due to the large amount of the share supply in the stock markets, the stock price dropped significantly. Non-tradable shareholders were required to negotiate with tradable shareholders for a compensation plan in order to get liquidity. As a compensation for tradable shareholders, non-tradable shareholders offer stocks, stock dividends, warrants and cash

2.3 Key Concepts and Past Empirical Evidence

35

for tradable shareholders. After the reform, non-tradable shares became fully tradable. Prior studies suggest that the split-share structure reform aligned non-tradable shareholders’ interests with those of tradable shareholders, reduced the agency problems and improved firm performance. The split-share structure reform provides state shareholder the opportunity to rapidly privatize by selling their shares on the market. Investors face heightened uncertainty about the non-tradable shareholders’ ownership intentions after the split-share structure reform and concerns of a stock supply shock on the market. By the end of year 2007, firms representing 97% of the Chinese A-share market capitalization had completed the reform. The impact of split-share structure reform has been widely studied. Hou and Lee (2012)’s evaluation of the impact of foreign share discount puzzle in China shows that before reforms were introduced, the state shareholder has little incentive to cooperate with private shareholders to ensure managers work towards firm’s market value maximization because the state holds the majority of non-tradable shares that have no wealth effects from stock price movements. After reforms were introduced, interest alignment between state and private shareholders encouraged their cooperation in monitoring managerial conduct implying that the institutional reform has also benefited minority shareholders. Hou et al. (2012) found that the split-share structure reform strengthened corporate governance incentives of state shareholders. The reform increased stock price informativeness of Chinese-listed firms, which imply that the split-share structure reform has improved the information environment and benefited minority shareholders. Supporting this, Liao et al. (2014) suggested that the market mechanism unleashed by this reform has mitigated the information asymmetry between state and private shareholders. Beltratti et al. (2012) examined the impact of reforms on the Chinese stock market, concluding that it is particularly beneficial for small stocks, stocks with poor historical stock returns and stocks issued by firms with low transparency and poor corporate governance. Liu et al. (2013)’s study of the impact of reforms on cash dividend payment shows that Chinese controlling shareholder’s cash dividend preference is attributed to the illiquidity of their shares rather than their non-tradability. Firth et al. (2010) found evidence that during reforms non-tradable shareholders (majority owned by the state) have incentives to exert political pressure on tradable shareholders (majority are owned by institutional investors) to accept the terms of reforms. Consequently, tradable shareholders end up bowing to political pressures to help state enterprises perform the reform quickly with low cost. Tan et al. (2020) suggest that the split-share structure reform improved the stock price informativeness and interest alignment between state and private shareholders to mitigate the agency problems between large state shareholders and minority private shareholders. Also, the wealth of state shareholders became more strongly related to share price movements, which enhanced the incentives of the state shareholders to monitor and ensure that the managers maximize firm value. Liang et al. (2020) suggests that the firm’s crash risk decreases significantly after the split-share structure reform, especially for private-controlled firms. Government-controlled firms with stronger political incentives tend to have higher crash risk. Sun et al. (2017) suggest that the split-share structure reform induces a significant decrease in crash risk, especially

36

2 Theoretical Considerations

when ownership is concentrated in hands of controlling shareholders. Moreover, the reform mitigates controlling shareholder’s tunnelling behaviors. Liu et al. (2020) suggests that the split-share structure reform mitigated credit misallocation, and due to the implicit government guarantee, banks preferred SOEs for safety of loans. They suggest that the financing privilege of SOEs declined significantly after the split-share structure reform.

2.3.8 Controlling Shareholders Ownership structure is a crucial issue for corporate governance. The highly dispersed ownership structure is mainly observed in USA and Japan, where the central corporate governance issue is to solve the agency problem between managers and shareholders, whereas a highly concentrated ownership structure is often found in European and Asian companies, where protecting the minority shareholders from expropriation by controlling shareholders is a key task. Chinese firms are characterized by a concentrated ownership structure with one or more controlling shareholders holding a large percentage of the shares, giving the controlling shareholders substantial discretionary power to use firm resources for their private gains at the expense of minority shareholders. Su et al. (2017) suggest that the higher control rights the controlling shareholder have, the more capable they can in expropriating minority shareholders’ interest. However, the degree of cash flow rights may act as an incentive to align the interests of the controlling shareholders with the firm. The higher the cash flow rights the controlling shareholder have, the more stimulated they are to secure the firm’s normal operation and thus to enhance their own wealth. They show that the presence of the ultimate controlling shareholder and the divergence between its control right and cash flow right lowers firm value. Wang et al. (2020) confirm that ultimate control rights significantly restrain corporate fraud; the greater the ultimate control rights, the lower the probability of fraud and the fewer the number of frauds. Based on the identity of controlling shareholders, four types of enterprises can be distinguished in China, namely (1) market-oriented SOEs (MOSOE), (such as China National Petroleum Corporation, China Power Investment Corporation and Sinosteel Corporation), (2) SOEs affiliated to central government agencies (central), central state assets management bureaus and state councils, (such as China Great Wall Computer Shenzhen Company Limited, China Merchants Property Development Company Limited), (3) SOEs affiliated to local government agencies (local), local state assets management bureaus and local government, (such as Guangzhou Automobile Group Company Limited, Shanghai Pharmaceuticals Holding Company Limited) and (4) firms controlled by private individuals (private), (such as Suning Commerce Group Company Limited and Zhejiang Busen Garments Company Limited). Central government-controlled firms are the focus of China’s economic reforms and are strictly monitored by the central government (Xu, 2004). The CEOs and chairmen of these firms are usually carefully selected by the central government on

2.3 Key Concepts and Past Empirical Evidence

37

the basis of their leadership qualities since most of them are in line for promoting to the rank of Minister (Chen et al., 2009). Local government-controlled firms are less closely monitored by the central government since they are located in regions far from the Chinese Communist Party’s power centre in Beijing. The distance, both geographical and administrative make central government enforcement of regulations difficult (Chen et al., 2009), but local governments have the right to set up their own regulations to manage local state assets. Both central and local government-controlled firms are less likely to be driven totally by the profit motive since officials of China’s government agencies are public servants. Their promotion depends on how well they implement government instructions and are paid fixed salaries. Even though they have the right to select managers and directors, as well as approve investment plans proposed by management, they are nevertheless prohibited from direct management as the ultimate corporate controllers. They also have no residual cash flow rights, as all dividends are submitted to the Minister of Finance. Therefore, the main agency problems of government-controlled firms arise from controlling shareholders and minority shareholders (Berkman et al., 2012). In contrast, market-oriented state enterprises are likely to pursue profits and enjoy a degree of autonomy from the state as they can retain after-tax profits. Since managers typically receive monetary rewards based on firm performance, the incentive compensation mechanism helps mitigate agency problems between controlling and minority shareholders (Berkman et al., 2012). Private enterprises are firms controlled directly by individuals and have mushroomed since China began opening up its economy in 1978. As they function like their counterparts in the developed countries, they are likely to show market-oriented conduct with strong commercial motivation (Wei et al., 2014). The government has limited supervisory power over private enterprises, and hence, these firms are likely to maximize shareholder wealth as they can select a management team based purely on performance criteria (Berkman et al., 2012). Central, local and MOSOEs are state-controlled firms that enjoy financial privileges not available to private firms, while private firms are more likely to be influenced by market forces and resemble their counterparts in developed countries (Wei et al., 2014). Banks are willing to lend to state-controlled firms since the government would share the losses with them if the firms default on loans (Liang et al., 2012). The Chinese government also provides political and financial support, such as a reduced tax burden and debt-for-equity swap to reduce firms’ debt burden (Sun & Tong, 2003).

2.4 Hypotheses Formulation The empirical literature on the determinants of board structure is generally framed from three basic theories, namely the resource-dependent theory, agency theory and

38

2 Theoretical Considerations

power circulation theory. The resource-dependent theory argues that the main function of the corporate board is to give advice and information needed to facilitate firms’ decision-making and strategic choices (Hillman & Dalziel, 2003; Pugliese et al., 2014). According to agency theory, however, corporate board functions to monitor interactions between managers as the agents acting on behalf of the shareholders as the principal (Fama & Jensen, 1983; Hillman & Dalziel, 2003; Shleifer & Vishny, 1997). Power circulation theory in corporate governance argues that CEOs can gain power from a coalition dominated by themselves, which can only be rivalled by a coalition formed by rival directors and executives (Henderson & Fredrickson, 2001; Ocasio, 1994; Shen & Cannella, 2002). Based on the three main theories above and the empirical evidence reviewed, three basic hypotheses have been formulated.

2.4.1 Scope of Operation The scope of operation hypothesis is proposed with the argument that board size and independence depend on the amount of advice and resources needed for a firm’s daily operations. Larger firms are usually engaged in larger amount and more diversified activities than smaller firms, such as operating different product lines, seeking frequent mergers and acquisitions, promoting products in markets of different regions and possessing more sophisticated financing and governance systems. Hence, larger firms have higher demand for information resources, including the product, labour, financial market information, as well as social connections with their potential cooperative partners and government officers. Therefore, as firms grow and diversify over time and space, they need a larger specialized board to conduct planning, decisionmaking and advising tasks (Coles et al., 2008; Lehn et al., 2009). In addition, Pfeffer (1972) proposed that the corporate board is a vehicle through which a firm can obtain external resources. Hermalin and Weisbach (1988) suggested that firms with diversified operations need more resources. Booth and Deli (1996) argued that larger firms tend to use their board seats to create connections with external parties due to the higher demand for external contracting relationships. Furthermore, scope of operation also influences board composition since firms with a larger scope of operation are usually faced with more serious agency problems than are smaller firms. Hence, more outside directors need to be nominated to monitor managerial behaviour in large firms (Coles et al., 2008). Anderson et al. (1998) and Mayers, Shivdasani and Smith (1997) provided evidence for the important role of outside directors in monitoring managerial behaviour.

2.4.2 Monitoring Hypothesis The main argument of the monitoring hypothesis is that board size and independence increase with the benefit of the board’s monitoring function, but decrease with the

2.4 Hypotheses Formulation

39

cost of the board’s monitoring function (Boone et al., 2007). Stuart Gillan et al. (2004) argue that firms choose corporate governance mechanisms according to their respective costs and benefits. Lehn et al. (2009) link group decision-making benefits and costs with group size; a larger board brings the benefit of collective information, (which is necessary for a board’s monitoring and advising activities), but it also brings the costs of coordination and problems of free riders. Hence, optimal board size should be the trade-off between costs and benefits. The cost of monitoring in the monitoring hypothesis is linked to firms’ growth opportunity. The benefit of monitoring can be attributed to the potential opportunity that managers can enjoy to extract private benefit at the expense of shareholders. Firms with higher growth opportunities usually operate in an unstable and competitive business environment. Demsetz and Lehn (1985) argued that the cost of monitoring increases when firms operate in a “noisy” environment. Hence, firms need a shrewd and cohesive board of directors with rapid decision-making abilities to improve efficiency. The costs of coordinating and processing increase the larger the board size, while free rider problems render it less effective in firms with high growth opportunities. For example, Lipton and Lorsch (1992) illustrated that in the process of group decision-making, every member is expected to elaborate his/her views clearly and explicitly in order to reach consensus. A larger group would definitely need more time to arrive at a consensus. Also, Jensen (1993) suggested that keeping the board smaller can improve efficiency, hence the optimal board size should be constituted of seven or eight members. Growth opportunities also influence board composition. Firms with higher growth opportunities usually face information asymmetry problems, which affect outside directors’ advisory function. Outside directors give advice either with limited information or acquire and process the necessary information at a higher cost; therefore, it is not optimal for a fast-growing firm to have too many outside directors (Adams & Ferreira, 2007; Linck et al., 2008; Maug, 1997).

2.4.3 Bargaining Hypothesis Hermalin and Weisbach (1998) first proposed the bargaining hypothesis, which postulates that board composition is the result of a CEO’s bargain with the rest of the board directors. A CEO’s bargaining power comes from his or her ability to influence the potential appointment of board directors. Directors appointed through the CEO would be more inclined to stand with the CEO; thus, a powerful CEO faces less monitoring. Hermalin and Weisbach (1998) argue that external determinants of board composition become inactive when corporate insiders constrain the selection of outsiders. Arthur (2001) found supporting evidence that the proportion of inside directors enhances the CEO’s bargaining power, leading to a less independent board. Baker and Gompers (2003) also suggested that the proportion of independent directors decreases with CEO’s power, but increases with outside shareholders.

40

2 Theoretical Considerations

Roosenboom (2005) verified the bargaining hypothesis by showing a negative relationship between board independence and managerial power at the time of initial public offering. Boone et al. (2007) additionally provided evidence that board independence falls with managers’ influence, but grows with constraints on managers’ influence. However, empirical evidence addressing these hypotheses shows a lack of uniformity across countries, suggesting that the effect of these determinants varies across different institutional settings. For example, Guest (2008) found from his study of determinants of board structure in the United Kingdom that board structure is less determined by related monitoring factors. In contrast, Lehn et al. (2009) suggested that board structure in the United States is determined by a trade-off between benefits of information brought by additional directors and coordination costs they engender. Linck et al. (2008) also found similar results in that board composition is consistent with the cost and benefit of the board’s monitoring and advising role. Also, Iwasaki (2008) found that Russia’s board structure is primarily determined by bargainingrelated variables and the country’s particularity as a transition economy. Similarly, Arthur (2001) found that the CEO’s bargaining power increases with the number of inside directors on boards in Australia. Meanwhile, Germain et al. (2014) found that Malaysia’s board structure is determined by the scope of operation, board size and monitoring factors. Corporate board structure in Taiwan is sensitive to changes in firm and CEO characteristics, as well as government regulations (Chen, 2014). Mak and Roush (2000) found that in New Zealand, board independence is positively related to firms’ growth opportunities (cost of monitoring). Although the determinants of corporate board structure in China having been studied by Chen and Al-Najjar (2012), its relevance is debatable as it was based on the period from 1999 to 2003, which is before the Chinese split-share structure reform.

2.5 Summary This brief review of the related theories and empirical findings reveals that even though the determinants of board composition have been widely researched in the existing literature, there are some aspects that still lack in-depth study. Firstly, most of these studies are explained from the perspective of mainstream Anglo-Saxon theories with little empirical research on the developing world: agency theory, resourcedependent theory and power circulation theory. Secondly, the majority of these studies pay little attention to the importance of institutional embeddedness and the influence of other stakeholders, especially when the economy is undergoing transformation. Thirdly, most of the conclusions are drawn from research in capitalist economies, such as in the United States and United Kingdom. Fourthly, the splitshare structure reform in 2005 greatly changed the Chinese institutional environment and is unique to the world, so that conclusions drawn from research prior to 2005 can no longer reflect the current situation. In addition, Chinese corporate governance

2.5 Summary

41

has its own characteristics that can act as a substitute for or complement to corporate boards during transition periods. As for the relationship between corporate governance and firm performance, one strand of studies found “linear” relationships, while another strand found “nonlinear” relationships. Thus, there is a lack of consensus among the findings, and most of them only focus on a single dimension of corporate governance, without paying sufficient attention to the joint effect of different corporate governance mechanisms on firm performance. Secondly, most of the previous studies have focused on the relationships between corporate governance and firm performance (Hu et al., 2010; Kang & Kim, 2012; Lefort & Urzúa, 2008; Yu, 2013), using perspectives adapted from mainstream theories, such as the agency theory. According to institutional theory, corporate governance practices vary in firms with different controlling shareholders. Therefore, the relationship between corporate governance mechanisms and firm performance in different institutional settings during China’s reform on SOEs is still under-examined. Although corporate governance and corporate risk-taking is a hotly debated issue, most previous studies have focused only on the relationship between board size and corporate risk-taking (e.g. Ho et al., 2013; Huang & Wang, 2015; Koerniadi et al., 2013; Nakano & Nguyen, 2012). Independent directors’ risk preferences in enterprises with different controlling shareholder types, and at different stages of state enterprise reform, are still under-examined. Besides, in-depth investigation into corporate governance is also lacking in terms of risk preferences of other stakeholders, such as the government, managers and CEOs.

Chapter 3

Methodology

3.1 Introduction This chapter describes the research methods and data used in this thesis. The generic method used for this book is the quantitative method with statistical analysis, which is applied because of the key focus of the study is on quantitative variables, such as corporate financial performance, financial risk, and cost and benefit. The analytical process is primarily based on firms’ financial and corporate governance data disclosed annually in their financial reports. The chapter starts with the construction of a conceptual framework in Sect. 3.2. Section 3.3 shows the analytical framework adopted in Chaps. 4, 5 and 6. Section 3.4 elaborates the research design of this study, specifically, research approach, model specification, hypotheses, data and estimation methods. It is then followed by Sect. 3.5, which discusses the econometric issues encountered in the analysis, including multicollinearity, heterogeneity and autocorrelation, as well as the endogeneity problem all or some of which may bias estimates.

3.2 Conceptual Framework The concept of corporate governance can be explained from multiple perspectives as discussed in Chap. 1, which largely depends on one’s perception of the world (Gillan, 2006). Gillan (2006) argues that corporate governance can be divided into issues external to firms and issues internal to firms. Figure 3.1 shows the conceptual framework used in the analysis in this book that illustrates the influence of corporate governance mechanisms on firm performance. The corporate governance structure applied here is adapted from the corporate governance models developed by Gillan (2006). The board of directors forms the centre of the corporate governance structure that connects shareholders (principal) and the CEO (agent). Zhang (1999) views corporate governance as a series of institutional © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0_3

43

44

3 Methodology

Law

Culture

Government

External

Creditors

Communities

Internal Supervisory Board

Board of Directors

Firm Performance / Risk

Shareholders

Corporate governance structure

Suppliers

Politics

Market

Consumers

CEO Management Scope

Cost

Benefit

Fig. 3.1 Conceptual framework. Source prepared by authors

arrangements regarding the corporate board’s functions, structures and shareholder’s rights. Shareholders, board of directors and CEO (head of managers) are three main corporate participants that constitute the most important part of corporate governance. Other external governance mechanisms, such as law, culture and politics are not the focus here, since this is a single country study in which all listed firms operate in China. Figure 3.2 shows the flow of analysis that illustrates the general idea, scope, circulation of objectives and rationality of actions. To begin with, the arrow at the top of this chart shows background information about China’s economic transition, through which corporate governance issues are raised. To be specific, due to the economic transition, state enterprises started seeking financial support from nonstate entities, and therefore, the ownership structure gradually became decentralized. During this period, the split-share structure reform of April 2005 remained the most important milestone of ownership reform in China. Consequently, whereas some state enterprises retained central government and local government agencies’ ownership, some state enterprises became partially privatized in the form of MOSOE, while some became wholly privatized.

3.2 Conceptual Framework 2000

5years

45

Split share reform 2005

Economic Transition

5 years

Central

2012- - -

Privatization of State enterprise

Ownership Decentralization

Local MOSOE

Private

Corporate Governance Reform

Objective 4: Implications

Objective 1: Determinants

Objective 2: Performance

Objective 3: Risk

Analytical Framework1

Analytical Framework 2

Analytical Framework 3

Fig. 3.2 Flow of analysis. Source Prepared by authors

We start with raising corporate governance issues during reforms by proposing four research objectives, with suggestions for further reforms. Specifically, this book begins with the analysis of the determinants of corporate board composition, which is the most important part of corporate governance, followed by the examination of the impact of corporate governance on firm performance and firm risk. Objective 1 is crucial for the understanding of corporate governance structure. Objective 2 is important for evaluating corporate governance efficiency. Objective 3 is critical for firms’ long-term development. Finally, from the analysis of Objective 1 to 3, we draw policy implications for China and other transition economies as they further their state enterprise and corporate governance reforms, which is the fourth objective. Chapters 4, 5, 6 and 7 will address each of these four objectives.

3.3 Analytical Framework In order to show how the objectives of this book are to be achieved, three analytical frameworks have been developed (see Figs. 3.3, 3.4 and 3.5). Specifically, Fig. 3.3 illustrates the analytical process to evaluate the determinants of board composition. The left-hand factors are potential determinants examined in this analysis; the right

46

3 Methodology

Determinants

Resource Dependent Theory

Central

Scope

Local Cost

Types

Agency Theory

MOSOE Benefit

Power Circulation Theory Stakeholder and Stewardship Theory

Institutional Theory

CEO’s Influence

Private Board Composition

other Governance Mechanisms Regulation and Reform

Before split-share reform Period After split-share reform

Fig. 3.3 Analytic framework 1. Source Prepared by authors

hand displays the comparative analysis applied to different subsamples including different categories of controlling shareholders and reform periods. Figure 3.4 shows the analytical framework to examine the relationships between corporate governance and firm performance. The left hand shows the key factors constituting board structure, ownership structure, CEO’s influence and other mechanisms important in this relationship. The right hand shows the different institutional environments through which the analysis was conducted. As with analytic frameworks 3.3 and 3.4, Fig. 3.5 links the relationship between corporate governance mechanisms and firm risk. The left hand shows the key factors constituting board structure, ownership structure, CEO’s influence and other mechanisms important to this relationship. The right hand shows the different institutional environments through which the analysis was conducted.

3.4 Research Design

47

Corporate Governance Central Agency Theory

Board Structure

Agency Theory

Ownership Structure

Local Types MOSOE Private

Firm Performance

Power and Steward Theory

CEO’s Influence

Institution Theory

Regulation and Reform

Before split-share reform Period

After split-share reform

Fig. 3.4 Analytic framework 2. Source Prepared by authors

3.4 Research Design Following three analytic frameworks, this section describes the research approach, hypotheses, model specification, data and estimation methods used.

3.4.1 Research Approach The analysis in the book uses quantitative methods for several reasons. Firstly, we need to find a numerical answer to the size of corporate boards, percentage of independent directors and firm financial performance, all of which are key indicators investors, shareholders and managers use to evaluate efficiency. Secondly, the numerical changes in firm financial performance and stock return volatility between the prereform period and the postreform period require a quantitative comparison. Thirdly, the hypotheses extracted from the theories reviewed that are being tested require the use of quantitative models (Muijs, 2010). In order to investigate the impact of the reform on corporate governance, the parameter t test for mean difference of the key variables is appropriate. It is suggested that empirically there are differences between firms before and after the reform, as

48

3 Methodology

Corporate Governance Central Agency Theory

Board Structure

Agency Theory

Ownership Structure

Local Types MOSOE Private

Power and Steward Theory

CEO’s Influence

Institution Theory

Regulation and Reform

Firm Risk Before split-share reform Period After split-share reform

Fig. 3.5 Analytic framework 3. Source Prepared by authors

well as firms with state and private controlling shareholders. Secondly, this study used the panel data regression methods to evaluate the hypotheses.

3.4.2 Hypotheses Development While the general hypotheses development was explained in Chap. 2, in this chapter we focus on the specific hypotheses, which are applied to delineated subperiods, as well as to separate groups of enterprises. (a)

Determinants of Board Composition Hypothesis

• Board Size and Independence Ha1: Board size and board independence are positively correlated with the scope of operation in firms with different controlling shareholder types and during different periods of split-share structure reform. The scope of operation hypothesis suggests that the size of the corporate board and the proportion of independent directors depend on the scope, complexity and

3.4 Research Design

49

diversity of the firm’s operation (Boone et al., 2007), because one of the key functions of corporate boards is to provide business information, expertise, social connections and other resources (Hillman & Dalziel, 2003). Lehn et al. (2009) suggested that as firms grow over time in a geographical territory, more directors are needed to facilitate their daily operations. Therefore, this study predicts that Chinese corporate board composition is dependent on the scope of the firm’s operations as advocated by many western studies.

3.4.2.1

Monitoring Hypothesis

Ha2: Board size and board independence are negatively correlated with the cost of monitoring, but are positively correlated with benefits of monitoring in different controlling shareholders types and different periods of split-share structure reform. The monitoring hypothesis is predicated on the argument that board size and independence increase with the benefits of monitoring, but decrease with the costs of monitoring. The benefits of monitoring refer to the potential opportunities for private benefits managers could extract. The cost of monitoring can be attributed to the firm’s growth opportunities because the monitoring of fast growing firms is usually characterized by information asymmetry problems that increase the difficulties for directors to realize their monitoring and advisory role with insufficient or costly information.

3.4.2.2

Bargaining Hypothesis

Ha3: Board independence is negatively correlated with CEO’s influence among different controlling shareholder types and in different periods of split-share structure reform. The bargaining hypothesis proposes that board composition is the result of CEO’s bargaining relationship with the rest of the directors. A self-interested CEO prefers less monitoring and pressures from the corporate board, and therefore, a corporate board dominated by inside directors who follow the CEO’s leadership can facilitate him in realizing his own objectives. Hence, a powerful CEO tends to face less monitoring from independent directors.

3.4.2.3

Other Corporate Governance Mechanisms

Ha4: Corporate governance mechanisms have significant influence on board size and board independence in firms with different controlling shareholder types and during different periods of split-share structure reform.

50

3 Methodology

Following Chen and Al-Najjar’s (2012) arguments, we predict that the supervisory board size, ownership concentration, managerial ownership, state ownership and CEO duality are important corporate governance mechanisms in China that may significantly affect board size and board independence. Specifically, the main function of a supervisory board is to supervise and monitor the corporate board; therefore, a large corporate board requires a large supervisory board to monitor it. Since the functions of a supervisory board and independent directors are overlapping to some degree, the supervisory board may substitute for independent directors. Managerial ownership aligns managers’ interests with shareholders increasing managers’ concerns about the corporate board composition, as managers’ interests are affected by the corporate board’s decision-making and strategic choice. Therefore, this study predicts that board size increases with managerial ownership, as more directors are needed to delegate managers’ interests in the firms. However, managerial ownership would reduce board independence, since one of the independent directors’ key function is to monitor managerial behaviour. Put simply, managers would not like to be monitored by independent directors. Ownership concentration represents the cumulative interests of several large shareholders in the firm. Whereas high ownership concentration shows large shareholders’ capacity to dominate the configuration of the board, low ownership concentration shows otherwise. Also, ownership concentration calls for expanding the board size, as more directors will be needed to meet large shareholders’ interests. However, large shareholders tend not to like independent directors as they may not stand with large shareholders on critical issues. Moreover, minority shareholders’ interests may bring them into conflict with large shareholders. Since state ownership has been retained in most of China’s listed firms, state shareholders can protect their interests through corporate board decisions. Consequently, they look for large and less independent boards because their interests are different from those of other investors. CEO duality gives the CEO leadership roles in both management teams and corporate boards. It is likely that CEOs could abuse their power by shaping a smaller and less independent board, to easily exert influence so as to pursue their own interests at the expense of shareholders. (b)

Corporate governance and firm performance hypotheses

The second research focus of the book is to link corporate governance to firm performance. Three hypotheses are postulated here to address this relationship. Hb1: Board structurehas significant influence on firm performance in firms with different controlling shareholder types and during different periods of split-share structure reform. Board independence in China can hardly exert positive influence on firm performance, because of political interventions that often constrain real autonomy of the independent directors. Although the government has enforced the rule that the percentage of independent directors should not be fewer than 33 percent by 2003, China’s corporate board is still dominated by insiders, increasing difficulties faced by independent directors to access corporate information. Thus, the hypothesis predicts

3.4 Research Design

51

that board independence has a negative effect on firm performance. As for board size, it argues that a smaller board is more effective than a larger one, since the larger board may raise coordination costs, as well as surface free-rider problems that slows decision-making, eventually harming firm performance. Despite that, large board size can bring firms with more resources and advice to facilitate firms’ decision-making and strategic choice. Lipton and Lorsch (1992) suggested that when board size is beyond 10 persons, they are unable to contribute to firm performance. Therefore, the analysis here predicts that the relationship between board size and firm performance is an inverted U-shaped curve. Although the supervisory board is set up to complement independent directors, its efficiency is influenced by government policies, the external environment, and the ownership structure and its own characteristics (Hu et al., 2010). Since China’s supervisory boards do not have the right to dismiss directors and managers, the prediction is their supervisory boards cannot improve firm performance, and hence, the expected relationship with firm performance should be negative. Hb2: The CEO has significant influence on firm performance in firms with different controlling shareholder types and during different periods of splitshare structure reform. A CEO’s career concerns and leadership reduce gradually as he/she approaches the age of retirement. Since changes in physical and psychological conditions in the aging process reduce their decision-making abilities (Antia et al., 2010), CEO age shall be hypothesized to negatively affect firm performance. Besides, while CEOs will have insufficient information and knowledge in the initial year of tenure, as time goes by, they would gradually familiarize himself/herself with the job and firmspecific knowledge, which in turn contributes to firm performance. However, after a time, the CEOs tend to overly focus on the firm’s internal culture and becomes some closed to the outside environment, thereby eventually harming firm performance. Thus, the CEO tenure and firm performance are likely to show a U-shaped curve relationship. When the positions of CEO and chairman of board are held by the same person, it can create a stronger leadership that can support faster decision-making and with that improve management efficiency. Especially during economic reforms, the corporate environment will generally be unstable as the professional managers market in China is imperfect making it difficult for them find new jobs. Thus, CEO duality is predicted to be more favourable for value maximization. Hb3: Ownership structure has significant influence on firm performance in firms with different controlling shareholder types and during different periods of split-share structure reform. State ownership, managerial ownership and ownership concentration are embodiments of ownership structure, which are the focus of this study. State ownership has been criticized for its inefficiency in improving firm performance because it lacks effective external monitoring and is often plagued by serious agency problems between government and domestic shareholders. Since the government has non-profit objectives for the firms it owns or controls, firm value maximization is

52

3 Methodology

often not the main goal of the state. Thus, state ownership may negatively affect firm performance. The managerial ownership connects managers’ interests with the shareholders’, and it motivates managers to work towards firm value and performance maximization. Hence, firm performance is predicted to increase with managerial ownership. Ownership concentration is an important corporate governance mechanism in China, where market conditions and the legal infrastructure are largely imperfect. Large shareholders undertake most of the corporate governance responsibilities but may expropriate small shareholders’ interests, Therefore, ownership concentration may lead to improved firm performance, but if ownership is overly concentrated, firm performance may suffer.

3.4.2.4

Corporate Governance and Firm Risk Hypotheses

A third analytical focus is to ascertain the extent to which corporate governance affects the level of firm risk. We postulate three hypotheses to address this. Hc1: Board structurehas significant influence on firm risk-taking involving different controlling shareholder types and at different periods of split-share structure reform. Consistent with previous findings, this analysis predicts that board size is negatively correlated with firm risk-taking, because large decision-making groups make it difficult to reach consensus. With every member holding different opinions, it is less likely for a large board to undertake risky projects. The main function of independent directors is to monitor managerial behaviour. The independent directors can constrain managers from taking risky projects, and managers tend to be more conservative with independent directors’ monitoring them. Therefore, board independence may reduce firm risk. The main function of the supervisory board is to complement independent directors to monitor both board director’s and managerial conduct. Therefore, a large supervisory board may constrain corporate risk-taking behaviour and implement conservative corporate strategy. Hc2: The CEO has significant influence on firm risk-taking in firms with different controlling shareholder types and during different periods of splitshare structure reform. A CEO’s risk-taking behaviour can be constrained by factors, such as job security, personal reputation and compensation. Thus, most of CEOs are adversely affected by risk. CEO duality gives the CEO extra power and responsibility that allows him/her to reduce risk-taking since he/she has to be responsible for any potential losses. Hence, CEO duality is predicted to negatively affect firm risk. As CEOs’ knowledge and experience are accumulated over time, it is likely that those with long tenures may undertake risky projects because the possibility of loss associated with risky projects reduces with the cumulative learnings, skills and refinement of risk-taking behaviour. Furthermore, CEOs’ concerns about future career development decreases with the length of tenure. Thus, CEO tenure is expected to increase firm risk-taking.

3.4 Research Design

53

Besides, the aging process changes individuals’ physical and psychological conditions that affect their decision-making, strategic choice, risk preference, motivations and leadership. Thus, the older CEOs are predicted to be more conservative than the younger ones. Hc3. Ownership structurehas significant influence on firm risk-taking in firms with different controlling shareholder types and during different periods of split-share structure reform. Under ownership structure, managers tend to be conservative with risk-taking as they have job security concerns. Managerial ownership gives managers more interests in the firm, and thus, managerial ownership tends to negatively affect firm risk. State ownership is likely to reduce firm risk-taking, since the state shareholder has the objectives of maintaining a lower unemployment rate and social stability, and is largely protected by the state. China’s firms are largely characterized by concentrated ownership structures. Large shareholders who have more interest in the firm undertake most of the corporate governance responsibilities. Due to the poor market conditions in China, large shareholders have often found it difficult to diversify their investments. Thus, they tend to be conservative on risk-taking.

3.4.3 Modelling 3.4.3.1

Determinants of Board Composition

The relationships examined here are expressed quantitatively through a number of causal equations. Equations (3.1) to (3.4) were developed to estimate determinants of board size and board independence (see static and dynamic panel equations below). Equations (3.1) to (3.2) are static panel models that were estimated using pooled OLS, random effects and fixed effects based on the likelihood ratio (LR) test and Hausman test. Models (3) to (4) are dynamic panel models that include the past value of dependent variables using an autoregressive process. The main predictors include scope, cost, benefit, CEO’s character, other firm-level corporate governance mechanisms and government regulations. The CEO’s character is used here to predict board independence only, as independent directors are the main group that CEOs bargains with (Chen, 2014). The corporate governance variables in the models used are drawn from Chen and Al-Najjar’s (2012) study, as these China-specific corporate governance characters may have significant impact on board composition. The models also used two-year dummy variables (govern03, govern08) to capture the effect of government regulations and reforms (see also Table 3.1). Static panel: Board Sizeit = αit +

3  j=1

β1 SCOPE jit +

2  j=1

β2 COST jit +

2  j=1

β3 BENEFIT jit

54

3 Methodology

Table 3.1 Summary of variables, definitions and measurement Variable name

Symbol

Measurement

Board Size

boardsize (person)

The total number of directors on corporate board

lnboardsize Natural logarithm of board size. The transformation of board size is based on the literature review about determinants of board size and board independence studies. Most research did the log transformation of board size (Chen, and Al-Najjar, (2012) Board independence

boardinde (ratio)

The number of independent directors divided by total number of directors

Supervisory board

supersize (person)

The number of supervisory directors on the supervisory board

Firm Size

firmsize

Measured by the total assets (CNY) with natural logarithm

Firm Age

firmage (Year)

Counted since the firm was established

Long debt

longdebt (ratio)

Long-term debt over total debt

Stock return volatility

sdreturn

The standard deviation of monthly stock return within 12 months of one year

Tobin Q

tobinq (ratio)

The equity and debt in market value scaled by the total assets in book value

Free cash flow

fcf (ratio)

Net earnings before interest and tax (EBIT) minus depreciation and amortization divided by total assets

Herfindahl index

hhi (ratio)

Measure of industry concentration or product market competition. The Herfindahl index of an industry is ranging from 0 to 1, and it is calculated by the sum of the squared market shares of each firm within the same industry. The market share of each firm is measured by the firm sales over the industry sales

CEO age

ceoage (Year)

The CEO’s age

CEO tenure

ceotenure (Year)

The number of years since CEO had been appointed

CEO duality

duality

Equals to 1 when a CEO is also the board chairman, 0, otherwise

Managerial ownership

manage

The proportion of shares held by the top managers

Ownership concentration concen10

The cumulative proportion of shares held by the top 10 largest shareholders

State ownership

state

The proportion of shares held by the government shareholders

Tobin Q

tobinq (ratio)

The equity and debt in market value scaled by the total assets in book value (continued)

3.4 Research Design

55

Table 3.1 (continued) Variable name

Symbol

Measurement

ROA

roa (ratio)

The net income scaled by total assets

ROE

roe (ratio)

The net income scaled by total equity

Sales growth

sale_grow (ratio)

The sales growth over the last year

Capital expenditure

fa_expen (ratio)

The expenditure in acquire fixed asset scaled by total assets

Regulation

govern 03

Equals to 1 if the year of observations is equal to or larger than 2003, and it is 0 otherwise

Reform

govern 08

Equals to 1 if the year of observations is equal to or larger than 2008, and it is 0 otherwise

Source Prepared by authors

+

5 

β4 GOVERNANCE jit + β5 govern03t + β6 govern08t + εit

j=1

(3.1) Board Independenceit = αit +

3 

β1 SCOPE jit +

2 

j=1

+

2 

β3 BENEFIT jit +

j=1

+

5 

β2 COST jit

j=1 2 

β4 CEO jit

j=1

β5 GOVERNANCE jit + β6 govern03t

j=1

+ β7 govern08t + εit

(3.2)

Dynamic Panel: Board Sizeit = αit + β1 Board Sizeit − 1 + β2 Board Independenceit +

3 

β3 SCOPE jit +

j=1

+

5 

2  j=1

β4 COST jit +

2 

β5 BENEFIT jit

j=1

β6 GOVERNANCE jit + β7 govern03t

j=1

+ β8 govern08t + εit

(3.3)

56

3 Methodology

Board Independenceit = αit + β1 Board Independenceit − 1 + β2 Board Sizeit +

3 

β3 SCOPE jit +

j=1

+

2 

β4 COST jit

j=1

β5 BENEFIT jit +

j=1

+

2 

5 

2 

β6 CEO jit

j=1

β7 GOVERNANCE jit + β8 govern03t

j=1

+ β9 govern08t + εit

(3.4)

Board size (lnboardsize) and board independence (boardinde) are frequently examined board composition variables. The detailed measurements of these variables are explained in Table 3.1, while the variables selected are explained. • Scope of operation (SCOPE) This study used three variables to capture the scope of firms operations, i.e. firm size (firmsize), firm age (firmage) and firm’s long term debt ratio (longdebt), which have been used in previous studies to measure the scope of operation (Chen, 2014; Linck et al., 2008). • Board monitoring cost (COST) The cost of board monitoring increases with firms’ growth opportunity and information asymmetry as elaborated in Chap. 2. This study used Tobin Q (tobinq) and stock return volatility (sdreturn) within one year as proxies for monitoring cost, which follows Guest’s (2008) study of determinants of board structure. Firms with higher growth opportunities usually face information asymmetry problems, which is costly for corporate boards to monitor. • Board monitoring benefits (BENEFIT) The benefit of monitoring boards refers to the potential opportunity for managers to extract private benefit. Boone et al. (2007) and Chen (2014) used free cash flow ratio (fcf ) and industry concentration (hhi) as proxies for the benefit of monitoring. Free cash flow increases managerial discretionary expenditure (Linck et al., 2008), which managers can curtail for personal use without disturbing firms’ short-term profitability and operations. Industry concentration subjects managers to less market discipline, which in turn heightens the potential private benefits managers can extract. • CEO’s influence (CEO) Following Linck et al. (2008), our analysis uses two factors as proxies of CEO’s influence: one, CEO tenure (ceotenure), which is the number of years since a manager

3.4 Research Design

57

has been appointed as CEO; two, CEO’s age (ceoage), which is the measure of CEO’s physical condition and when his time to retire. • Corporate governance characteristics (GOVERNANCE) The key corporate governance mechanisms used to estimate board composition are state ownership (state), managerial ownership (manage), ownership concentration (concen10), CEO duality (duality) and supervisory board size (supersize). As suggested by Chen and Al-Najjar (2012), these are the most representative Chinese corporate governance characteristics that substitute or complement corporate board. • Regulation and reform One of the most important government regulations is “Guidelines for introducing independent directors” (governs03), which came into force in June 2003. Another is the split-share structure reform, which was introduced in 2005 and ended in December 2007 (govern08). Time effects are assumed to be same across all the firms.

3.4.3.2

Corporate Governance and Firm Performance

Equations (3.5) and (3.6) below explain the influence of corporate governance mechanisms on firm performance. Static models were applied to both “linear” and “nonlinear” estimations with the predictors being board structure, ownership structure, CEO’s characteristics and government regulations. The dynamic model captures the impact of past firm performance on current firm performance (see Table 3.1 for the description and measurement of the variables). Static panel: Performanceit = αit +

3 

β1 Board structure jit +

j=1

+

3  j=1

β3 CEO jit +

3 

β2 Ownership Structure jit

j=1 7 

β4 Controls jit + β5 govern03t

j=1

+ β6 govern08t + εit

(3.5)

Dynamic Panel: Performanceit = αit + β1 Performanceit − 1 +

3  j=1

β2 Board structure jit

58

3 Methodology

+

3  j=1

+

7 

β3 Ownership structure jit +

3 

β4 CEO jit

j=1

β5 Controls jit + β6 govern03t + β7 govern08t + εit

j=1

(3.6) Performance is measured by accounting performance (roa and roe) and market performance (tobinq). The independence variables are grouped into board structure, ownership structure, CEO’s influence and other control variables. • Board structure Board size (lnboardsize), board independence (boardinde) and supervisory board (supersize) are the board structure variables measured in this study. China uses a two-tier board structure that is made up of a main board of director and a supervisory board. The supervisory board is composed of representatives of shareholders and employees who are mainly responsible for monitoring. • Ownership structure Managerial ownership (manage), state ownership (state) and ownership concentration (concen10) are most representative of China’s firm ownership structure characteristics during the assessment period. Although other types of ownership also exist, they are less frequently used. Management shareholdings are suggestive of an effective corporate governance mechanism. However, state ownership remains important, while ownership is highly concentrated in China’s listed firms. • CEO’s influence (CEO) Three factors are used to capture CEO’s influence: one, CEO tenure (ceotenure) is the number of years served since a manager was appointed CEO; two, CEO’s age (ceoage), CEO’s physical condition and when his time to retire. CEO duality (duality), which refers to the CEO also serving as chairman of the board. • Controls This study also includes additional variables that may influence firm performance. Following the practice in previous studies (Li et al., 2015, Liu et al., 2015, Conheady et al., 2014), firm size, firm age and long-term debt were used to represent firm’s scope of operation, stock return volatility and free cash flow ratio respectively, which represent the cost of monitoring. Other variables included were sales growth and fixed assets expenditure.

3.4 Research Design

3.4.3.3

59

Corporate Governance and Firm Risk

As with the models above, Eqs. (3.7) and (3.8) below show static and dynamic panels to estimate the influence of corporate governance mechanisms on firm risktaking. The main variables used are described in Table 3.1. Static panel: Firms riskit = αit +

3 

β1 Board structure jit +

j=1

+

3  j=1

β3 CEO jit +

3 

β2 Ownership structure jit

j=1 3 

β4 Controls jit

j=1

+ β5 govern03t + β6 govern08t + εit

(3.7)

Dynamic Panel: Firm riskit = αit + β1 Firm riskit − 1 +

3 

β2 Board structure jit

j=1

+

3  j=1

+

3 

β3 Ownership structure jit +

3 

β4 CEO jit

j=1

β5 Controls jit + β6 govern03t + β7 govern08t + εit

(3.8)

j=1

• Firm risk As presented in classical decision theory, risk is most commonly conceived to denote variation in the distribution of possible outcomes, their likelihoods and their subjective values (March & Shapira, 1987). Based on Chen and Zheng’s (2014) suggestion, the present analysis used stock return volatility to capture firm risk rather than specific policies, such as debt ratio and RandD expenses in order to avoid the ambiguity of the interpretation. Firm risk is measured by the monthly stock return volatility within one year, which is considered to be the total risk faced (Huang & Wang, 2015). • Controls Firm size, ROA and Tobin Q are used to estimate firm risk-taking. A large firm is expected to face less risk, since it is able to diversify firm risk through selling different products and services. Besides, firms with higher ROA and Tobin Q are usually associated with high growth opportunity and are correspondingly more risky (Nguyen, 2011).

60

3 Methodology

3.4.4 Sample and Data The sample used in this study consists of 444 firms that have been continuously listed in the Shanghai and Shenzhen stock markets over the period 2000 to 2012, which was selected because it covers 5 years before (2000–2004 and 5 years after (2008–2012) the split-share reforms (2005–2007) were introduced allowing us to undertake a balanced time-based analysis Besides, the legal status of China’s stock market was formalized through the enforcement of security laws in July 1999, which means that before the year 2000, the Chinese stock market was not well organized and properly supervised. The sample was selected using a set of robust criteria. Firstly, all of the continuously listed firms during the sample period were selected because one of the main objectives is to evaluate the effect of government regulations and reforms on firmlevel behaviour. The baseline sample consisted of 1088 China’s public listed firms in 2000. Out of this population, 496 firms were active over the period between 2001 and 2012. Secondly, financial firms, such as banks, insurance companies, financial institutions (7 firms) were excluded from this study as they follow different governance procedures and the interest rate is not decided by these financial firms. Thirdly, firms controlled by government universities, research institutions, government media agencies (27 firms) were also excluded because they were non-profit organizations that retained their revenues for research and innovation or cultural promotion, rather than having them distributed to shareholders. Fourthly, firms without the critical data for analysis (18 firms) during the sample period were eliminated, including those that did not have a complete accounting period of 12 months and also those with their accounting year not ending on the 31st December. Of the total, the 444 firms that were selected met all the criteria. In addition, several subsamples that covered the before reform period between 2000 and 2004, the reform period between 2005 and 2007 and the after reform period between 2008 and 2012 were drawn from the overall sample. Based on the ultimate controlling shareholder categories,1 this study further classified state enterprises into those controlled by central government agencies, local government agencies, SOE entities, as well as private entities. All of the financial and corporate governance data was drawn from the companies’ annual reports gathered from the China Stock Market Accounting Research Database and the CCER Database developed by GTA Information Technology Company Limited and SinoFin Financial Information Company Limited.

1 According to the China Securities Regulatory Commission (CSRC), the ultimate controlling share-

holders are the investors who (i) hold directly or indirectly 50% of the total outstanding shares, (ii) control directly or indirectly 30% of total voting rights, (iii) can use the voting rights to select more than 50% of board directors, (iv) have significant influence over the decision making in shareholder’s meeting and (v) other situations recognized by CSRC.

3.4 Research Design

3.4.4.1

61

Panel Data Estimation

Panel data was used in this study because they offer more robust econometric estimation than cross-sectional and time series data estimations undertaken separately (see Engle and Granger, 1988). Also, panel data offers more degrees of freedom and variability than pure cross-sectional data, which does not support time effects, or pure time series data, which disregard the individuality of each entity. Hence, panel data give greater capacity for capturing the complexity of economic relationships than a single cross-sectional or time series data series. In addition, panel data allow control of unobserved individual heterogeneity, such as culture, policies, regulations and others (Baltagi, 2008). However, panel data is not bereft of challenges, which include data collection issues, such as non-responses, as well as methodological issues such as endogeneity and heterogeneity (Hsiao, 2007). Based on the characteristics of the panel data set, several estimation methods were applied in this study.

3.4.4.2

Pooled OLS Regression

The pooled ordinary least squares regression or constant coefficient model simply pools all observations together and assumes that all the explanatory variables are strictly exogenous, and the error term is independently and identically distributed. It can be considered as an ideal model based on idealized assumptions of the data set. Even though in reality these assumptions are hard to satisfy, the OLS model provides a benchmark for other models to develop more realistic assumptions. The pooled OLS model equation can be illustrated as follows: BSit = β1 + β2 Sit + β3 Oit + u it . whereby β 1 represents each individual firm and each time point, which have the same intercept value. The subscripts i and t represent the ith firm and tth time point respectively, while uit is the error term. The variables BS, S and O represent board size as the dependent variable, supervisory board size and ownership concentration respectively.

3.4.4.3

Fixed Effects Regression Model

The term “fixed effects” (FE) implies that each individual firm has its own characteristics, and these characteristics do not vary across time. There are three methods to estimate the fixed effect model. The first method is the least squares dummy variable model that captures the individuality of the entities by including the entity dummy variables and allows each entity to have its own intercept value. This kind of dummy variable technique becomes impossible when there are too many entities in the sample that result in a situation called perfect collinearity (Gujarati, 2009). The model can be illustrated as below:

62

3 Methodology

BSit = β1i + β2 Sit + β3 Oit + u it . (Note that β 1 is subscribed to i, which means that each entity has its own intercept). The second method is the fixed effects within group estimator that is designed to eliminate the characteristics of each individual entity so that the pooled OLS can be applied to estimate the model. To eliminate individuality, the mean values of variables within the entity are first calculated and subtracted from the mean values of the original values. The model can be illustrated as follows bsit = β2 sit + β3 oit + u it . (Note that bs, s, o represent the mean value of BS, S, O within each entity. The intercept value of each entity is wiped up at the same time). Lastly, the first difference is another estimator to eliminate individuality by taking into account successive differences. The observation of 1st entity was subtracted from the 2nd entity, then, the 2nd entity from the 3rd entity, and so on. BSit = β1i + β2 Sit + β3 Oit + u it . (Note that:  represents the first difference operator: BSit = BSit − BSit −1 .) However, both the within group method and first difference method were not able to control for the effect of time-invariant variables. The validity of the fixed effect model can be diagnosed by the F-test, where the null hypothesis is that all the differential intercepts are equal to 0. If the null hypothesis cannot be rejected, it means that there is no difference in the intercept values, and therefore, the OLS results can be used. Otherwise, the fixed or random effect model results have to be used.

3.4.4.4

Random Effect Model

The random effect (RE) model is different from the fixed effect model in that it deals with individuality. The term randomness means that the variation between entities is random and is captured in the error term. The random effect model treats the intercept β 1i as random with a mean value of β 1 and a random error of Ei , shown as: β 1i = β 1 + Ei . The model is therefore transformed into: BSit = β1 + β2 Sit + β3 Oit + wit (where wit = εi. + u it ) whereby Ei represents an individual specific error with a zero mean and constant variance. The error term uit represents both cross section and time series combined, which is also named as an idiosyncratic term, so that wit is not correlated with the explanatory variables.

3.4 Research Design

63

The likelihood ratio (LR) test is used to check whether the data can be pooled, and the null hypothesis being that OLS is nested in the fixed effect model. If the hypothesis is rejected, it means that each entity has its own characteristics, and therefore that the data cannot be pooled. The Hausman test compares the applicability of the fixed effects model with the random effects model. The null hypothesis is that there is no systematic difference between the fixed effects model and random effects model. If the test is rejected, it can be concluded that the fixed effects model is more appropriate than the random effects model. The Lagrange multiplier test can be used to assist the Hausman test to decide whether the random effects model is more appropriate than the fixed effects model. The null hypothesis of the Lagrange multiplier test is that there are no random effects, and if it is rejected, it means that the random effect is not better than the fixed effect in estimating the model.

3.4.5 Statistical Robustness 3.4.5.1

Multicollinearity

This analysis uses two (2) widely used methods in detecting the problem of multicollinearity in multivariate regression. The first is to look at the correlation matrix between independent variables. If the correlation between two variables is greater than 0.8, it means that these two variables are overlapping to a large extent, and therefore, one of them needs to be dropped. Another popular way is based on the variance inflation factor (VIF) indicator. The higher the VIF value, the more likely there a multicollinearity problem. If a variable has a VIF value more than 10, it means that the variable has serious collinearity problems, and hence, must be removed (Gujarati, 2009).

3.4.5.2

Heteroscedasticity

One of the important assumptions of OLS regressions is that the error term has constant variances, or is homoscedastic. Otherwise, it faces the heteroscedasticity problem, and thus, the OLS estimate becomes not optimal as it gives equal weight to all observations regardless of whether the error term has larger or smaller variances. The Breusch–Pagan tests can be used to detect the heteroscedasticity problem. The null hypothesis is that the error variances are all equal. Nonetheless, heteroscedasticity problems can be mitigated using White’s robust standard error (Williams, 2015).

64

3.4.5.3

3 Methodology

Autocorrelation

Autocorrelation occurs when the error terms of the regression ut+1 , u t+2 , ut+3 are correlated with each other, which violate the assumptions used in classical OLS regressions. The detection of autocorrelation can be realized in the Wooldridge test. The existence of autocorrelation can be handled by dynamic regression models or using the Newey–West standard error.

3.4.5.4

Endogeneity

To address endogeneity problems, Baltagi (2014) proposed the use instrumental variables and the generalized method of moments (GMM), which has now become standard practice. One of the main assumptions of classical OLS regressions is that all the regressors are strictly exogenous and are not correlated with the error term. Violation of this assumption will result in the endogeneity problem occurring. Therefore, it is necessary to introduce instrumental variables, which are correlated with the endogenous regressor but not affected by the error term. In the first stage of the two-stage least squares (2SLS), the endogenous variable is estimated using instrumental variables after which it is used as the estimated value (not the original value) in the main regression. The dynamic panel data estimation method was applied when the current value of the dependent variable depends on its past value, which was developed with the assumption that the instrumental variable methods cannot fully evaluate the information in dynamic panel data. The generalized method of moments (GMM) in dynamic panel data estimations (developed by Arellano and Bond (1991)) has proved to be more efficient than other methods. The key argument of the Arellano and Bond estimator is that the essential instruments are within the model equation. The endogenous variables can be instrumented by its past values, as well as external instruments. Roodman (2006) further developed the Arellano and Bond estimator by allowing additional features, such as that the instruments can further be classified by “IV-style” and “GMM-style”. There are two important diagnostic tests in the GMM estimator, namely the Sargan or Hansen test for the validity of instruments and the AR(2) test of second-order serial correlation.

3.4.5.5

Heckman’s Sample Selection Model

Heckman’s sample selection model is a significant statistical procedure that addresses the problem of selection bias, which produces statistical error caused by the bias in determining a portion of sample for experiment. He provided a theoretical framework that models the dummy endogenous variable that estimates the probability of a participant being in one out of two conditions indicated by the endogenous dummy variable. A sample selection model always involves two equations, (1) a regression equation that determines the outcome variable and (2) a selection equation that

3.4 Research Design

65

models the selection process of a portion of the sample indicated by the dummy endogenous variables. The treatment effect model has two unique characteristics: one, the dummy endogenous variable (W i ), which enters into the regression equation directly; and two, the outcome variable Y i which is observable in both W i = 1 and W i = 0. Z i to determine W i . The equations are shown below. (1) (2)

Regression equation: Y i = βX i + δW i + εi Selection equation: W i = γ Z i + ui .

The treatment effect model assumes that the correlation between two error terms εi and ui is nonzero, which can be detected through the LR test, where P < 0.05 means that the assumption is met. The overall goodness of fit of the model is examined by the Wald test, where P < 0.05 means that the covariate used in the model should be appropriate (Guo & Fraser, 2014).

3.5 Summary This chapter presented the methodology data used in the book, including the conceptual framework, research design and analytic frameworks. Also discussed are the empirical models derived from the analytic frameworks to test the hypotheses. The important econometric concerns regarding multicollinearity, heteroscedasticity, autocorrelation and endogeneity were also discussed in detail. The sampling strategy (including the periodization used) and data collection methods applied were explained in this chapter. In addition, this study presented different panel data estimation methods to greatly increase the accuracy of the estimations in the study. In doing so, various statistical instruments were discussed to address problems of collinearity, heteroscedasticity, autocorrelation and endogeneity problems. Following Wintoki et al. (2012), the GMM estimator was applied to address endogeneity problems in the regressions.

Chapter 4

Determinants of Board Composition

4.1 Introduction The corporate board is at the centre of corporate governance. Especially during economic transition, the corporate board has played an important role in enforcing government regulations. The China Securities Regulatory Commission issued in 2001, “Guidelines of Introducing Independent Directors”, which stipulated that by the end of June 2003, China’s listed firms must have at least one third of its directors as independent directors in the corporate board. During the split-share structure reform in 2005–2007, the corporate board was an important connection through which non-tradable shareholders were able to negotiate with tradable shareholders for liquidation. Therefore, understanding what determines board composition is crucial for understanding the roles that the corporate board played in corporate governance during economic reforms. This chapter examines six categories of determinants, including the scope, cost, benefit, CEO characteristics, firm-level corporate governance mechanisms and government regulation effects. The analysis begins with a descriptive analysis (as specified in Sect. 4.2), of the nature of variables used in the analysis and the associated correlation matrix. Section 4.3 discusses the t-test results of board size and board independence between before and after the reform, as well as between state and private enterprises. This is followed by Sect. 4.4, which examines the graphical images of board size and board independence. Section 4.5 illustrates the determinants of board size and board independence in China, while Sect. 4.6 compares the determinants of board size before and after the reform, and different controlling shareholders. Finally, Sect. 4.7 summarizes the chapter findings.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0_4

67

68

4 Determinants of Board Composition

4.1.1 Description of Variables This section describes the nature of the key variables investigated in the study, including distribution and correlations between variables. Table 4.1 shows the descriptive statistics of the key variables. The rest of the descriptive statistics of different subsamples are shown in Appendix A. In order to get rid of the influence of extreme values and outliers, all variables are winsorized at 1%. The definition and measurement of these variables were explained in Chap. 3. Table 4.2 shows the correlation matrix of the key variables investigated in this chapter whereby it is evident that the highest correlation coefficient is 0.5098, between firm age and board independence, which suggests there is no serious multicollinearity issues in the analysis, since all correlations are lower than 0.8 as suggested by the rule of thumb. We also checked the value of the variance inflation factor (VIF) in each regression model to make sure that the multicollinearity is not a significant problem. Table 4.1 Descriptive statistics Pool Board Size/ boardinde Independence boardsize Scope

N

Mean

Min

Max

Sd

P25

P50

P75

5772

0.298 0

0.556 0.132 0.286

0.333 0.364

5772

9.283 5

15

9

2.14

8

11

lnboardsize 5772

2.213 1.609

2.890 0.237 2.079

2.197 2.398

firmage

5772

8.318 7

9

0.615 8

8

9

firmsize

5772 21.554 19

25

1.171 21

21

22

longdebt

5772

0.148 0

0.735 0.179 0.004

0.077 0.229

tobinq

5772

1.626 0.772

6.887 0.972 1.066

1.304 1.792

sdreturn

5772

0.125 0.039

0.347 0.058 0.084

0.110 0.151

fcf

5772

0.039 -0.629

0.361 0.148 0

0.054 0.116

hhi

5772

0.115 0.019

0.860 0.138 0.050

0.070 0.123

CEO’s influence

ceotenure

5772

3.737 1

13

3

ceoage

5772 46.385 32

62

6.633 42

46

51

Governance

supersize

5772

4.118 2

9

1.338 3

3

5

concen10

5772

0.561 0.222

0.879 0.146 0.458

Cost Benefit

Others

2.700 2

5

0.572 0.669

state

5772

0.256 0

0.750 0.255 0

0.221 0.493

manage

5772

0.014 0

0.296 0.040 0

0.002 0.011

sale_grow

5772

0.222 −0.710 4.464 0.618 −0.025 0.125 0.311

fa_expen

5772

0.054 0.000

0.261 0.055 0.013

0.036 0.075

roa

5772

0.023 −0.348 0.202 0.074 0.009

0.028 0.054

roe

5772

0.036 −1.366 0.393 0.209 0.017

0.062 0.110

Note All the variables are winsorized at 1% Source Computed by authors

0.0037

−0.0238

−0.0757

0.1138

0.0673

−0.0009

0.0913

−0.2133

−0.0106

−0.2626

−0.021

0.008

−0.0664

concen10

duality

state

manage

sale_grow

fa_expen

Source Estimated by the authors

0.0859

0.1919

−0.0865

0.1119

0.0065

0.0234

0.1313

0.0717

ceoage

0.1373

0.1543

−0.1338 0.1628

−0.0042 0.0573

0.0474

0.0765

fcf

0.3098

0.0403

0.033

0.0479 0.0167

0.0539

0.0234 0.0122

0.0226

0.053

0.2999

−0.0056 −0.1679

0.0373

0.0176

−0.0658 −0.0114 −0.0401

−0.0029 −0.0374 0.1586

−0.0914 −0.1317 0.2215

0.0192

−0.0418 −0.032

0.0001

−0.0245 −0.181

−0.1207 −0.1199 0.0105

0.1182

−0.0419 0.0534

0.1196

−0.0031 0.4508

0.1703

state

manage

−0.0485 0.0947

sale_grow

0.0269

0.0439

−0.0112 1

−0.0949 1 −0.0233 0.0096

0.1022

−0.0995 1

1

concen10 duality

−0.0289 1

1

ceotenure ceoage

−0.0026 1

1

hhi

−0.0301 −0.0576 0.1314

−0.2251 −0.1017 0.0737

0.0157

−0.1821 −0.1348 0.0731

0.0226

0.0069

−0.0972 −0.0374 0.0537

0.0482

0.0501

0.0459

−0.0725 0.0019 0.1853

1

sdreturn

−0.0409 −0.0831 −0.0604 1

−0.4601 −0.0076 0.0365

0.057

tobinq

−0.0296 −0.0516 0.2309

−0.3989 0.0843

0.0881

0.0224

0.2215

−0.0259

1

longdebt

−0.3581 −0.1631 1

0.3091

1

firmsize

−0.0592 0.0856

0.2244

0.1608

0.0326

0.1508

1

−0.0184

ceotenure

0.0221

0.0063

0.2264

hhi

0.0412

0.0568

0.0473

−0.0688

−0.0679

fcf

−0.0138

sdreturn

0.2281

0.0158

tobinq

−0.0806

0.0455

longdebt

0.1288

−0.0726

0.2226

−0.068

−0.1262

0.2167

firmsize

1

0.316

0.0866

0.5098

firmage

0.0687

−0.1192

supersize

1

boardinde lnboardsize supersize firmage

1

lnboardsize −0.0989

boardinde

Table 4.2 Correlation matrix

4.1 Introduction 69

70

4 Determinants of Board Composition

4.2 Variations in Corporate Governance Mechanisms During Reform The t-tests results shown in Tables 4.3 and 4.4 suggest that statistically most of the investigated variables vary significantly before and after the reform, as well as between state and private enterprises. Hence, the comparative analysis undertaken in Sect. 4.3 is meaningful. Table 4.3 shows the t-test results of mean differences of key variables between the before-reform and after-reform periods. It shows that most of the variables investigated vary significantly before and after the reform, except for the Herfindahl index (measure of ownership concentration), duality, managerial ownership, sales growth, ROA and ROE. For board structure, the analysis found that board size fell to around one person after the reform. In contrast, the percentage of independent directors increased about 17%, which means that China’s listed firms have adjusted board composition to fulfil government regulations, but kept the board size unchanged. On the scope of operation, the study found that firm size, firm age and level of long-term debt ratio increased significantly after the reform. The cost of monitoring including Tobin Q and stock return volatility decreased significantly, but the benefit of monitoring, in terms of free cash flow ratio, increased. Besides, the average CEO age and tenure also increased after the reform. As for the investigated corporate governance variables, the average supervisory board size decreased slightly. Ownership concentration of the largest 10 shareholders decreased by 10% from 0.614 to 0.506, while state ownership decreased by 25% from 0.369 to 0.116, which suggests that the split-share structure reform had decentralized ownership from large shareholders to small shareholders, and from state shareholders to non-state shareholders and individuals. Table 4.4 displays t-test results of mean differences of key variables between state and private enterprises, which shows that most of the variables vary significantly between state and private enterprises, except for CEO duality. On average, state and private enterprises show about 28.6% and 33.1% independent directors respectively. On scope of operation, state enterprises tend to have larger firm size and long-term debt than private enterprises, which is because privatization targeted the small SOEs, sparing the large SOEs. On cost of monitoring, private enterprises show higher Tobin Q and stock return volatility than state enterprises. On the benefits of monitoring, state enterprises show higher free cash flow ratio, which means that managers in state enterprises enjoy more discretionary expenditure than among private enterprises. Meanwhile, CEOs in state enterprises tend to be older than those in private enterprises but not in terms of their tenure. This is likely because the owners of private enterprises tend to install themselves as CEOs in private firms, so that private enterprises’ CEOs usually show longer tenure. As for the investigated corporate governance variables, state enterprises show larger supervisory boards, ownership concentration and state ownership than private enterprises, but managerial ownership in private enterprises tends to be higher. This suggests that private enterprises motivate managers better than state enterprises.

4.2 Variations in Corporate Governance Mechanisms During Reform

71

Table 4.3 t-test of mean differences, key variables, 2000–2004 and 2008–2012 Reform

N

Before

2220

After

2220

0.369

Lnboardsize

Before

2220

2.234

After

2220

2.190

Supersize

Before

2220

4.300

After

2220

3.950

Before

2220

7.850

After

2220

8.760

firmsize

Before

2220

21.170

After

2220

21.970

longdebt

Before

2220

0.138

After

2220

0.166

Before

2220

1.445

After

2220

1.869

sdreturn

Before

2220

0.096

After

2220

0.135

Fcf

Before

2220

0.044

After

2220

0.028

Before

2220

0.111

After

2220

0.118

Before

2220

2.810

After

2220

4.580

ceoage

Before

2220

45.250

After

2220

47.710

concen10

Before

2220

0.614

After

2220

0.506

Before

2220

12.39%

After

2220

13.24%

State

Before

2220

0.369

After

2220

0.116

manage

Before

2220

0.010

After

2220

0.020

Before

2220

0.237

After

2220

0.201

Before

2220

0.060

Boardinde

firmage

tobinq

Hhi ceotenure

dualitya

sale_grow fa_expen

Mean

Mean difference

t statistics

−0.173***

−49.817

0.044***

6.145

0.353***

8.834

−0.904***

−64.618

−0.795***

−23.512

−0.028***

−5.170

−0.424***

−14.741

−0.039***

−28.428

0.016***

3.526

0.196

−0.006

−1.528

−1.773***

−22.766

−2.453***

−12.371

0.108***

25.910

−0.85% 0.253***

−0.853 37.390

−0.001

−1.154

0.036

1.952 (continued)

72

4 Determinants of Board Composition

Table 4.3 (continued) Reform

Roa Roe

N

Mean

After

2220

0.048

Before

2220

0.023

After

2220

0.025

Before

2220

0.028

After

2220

0.045

Mean difference 0.012***

t statistics 7.431

−0.002

−2.489

−0.016

−1.024

a Test

Note of difference in proportions Source Computed by the authors

4.3 Board Size and Board Independence During the Reform Figure 4.1 describes changes in board size during the period between 2000 and 2012, which shows that board size reached its highest point between 2002 and 2003. This phenomenon can be attributed to government regulations towards increasing independent board directors, with the stipulation that by the end of June 2003, China’s listed firms must have at least one-third independent directors. Thus, board size has increased with the implementation of independent director regulations. After 2003, board size decreased gradually over the years, which means that China’s firms have adjusted their board size to reach its optimum of about 8 to 9 person, which is also the figure advocated by Lipton and Lorsch (1992). In addition, state enterprises tend to have a slightly larger board size than private enterprises. State enterprises usually have a board size of 9 to 10 persons compared to 8 to 9 persons in private enterprises. Figure 4.2 shows changes in board independence during the period between 2000 and 2012. Since the CSRC issued the guidelines on introducing independent directors in corporate boards, board independence increased dramatically since 2001. After the year 2003, the trend of board independence was slowly upwards over time. It seems that Chinese firms increase board independence only because of the pressure from the government. Even then, private enterprises tend to have greater board independence than state enterprises.

4.3.1 Determinants of Board Size and Board Independence This section discusses the determinants of board size and board independence. Interestingly, some of the hypotheses developed in Chap. 3 are supported, while and some are not. The regression methods shown in this section include the pooled OLS, fixed and random effects models, as well as the dynamic two-step GMM estimation method. All methods were estimated using robust clustered standard errors, which is a widely used method to minimize the impact of heteroscedasticity. Besides, the

4.3 Board Size and Board Independence During the Reform

73

Table 4.4 t-test of mean differences, key variables, state and private enterprises N

Mean

4220

0.286

boardinde

State Private

1552

0.331

lnboardsize

State

4220

2.236

Private

1552

2.151

State

4220

4.290

Private

1552

3.660

firmage

State

4220

8.240

Private

1552

8.520

firmsize

State

4220

21.650

Private

1552

21.300

State

4220

0.154

Private

1552

0.133

tobinq

State

4220

1.550

Private

1552

1.833

sdreturn

State

4220

0.122

Private

1552

0.133

State

4220

0.046

Private

1552

0.019

Hhi

State

4220

0.117

Private

1552

0.111

ceotenure

State

4220

3.670

Private

1552

3.900

State

4220

46.770

Private

1552

45.320

State

4220

0.578

Private

1552

0.513

State

4220

10.36%

Private

1552

18.69%

State

4220

0.334

Private

1552

0.045

manage

State

4220

0.010

Private

1552

0.020

sale_grow

State

4220

0.204

Private

1552

0.271

State

4220

0.056

Private

1552

0.047

supersize

longdebt

Fcf

ceoage concen10 dualitya State

fa_expen

Mean difference

t

−0.045***

−13.055

0.085***

12.468

0.624***

17.966

−0.276***

−16.149

0.344***

9.919

0.021***

3.983

−0.283***

−8.46

−0.011***

−6.201

0.027***

5.625

0.005

1.338

−0.230***

−2.692

1.451***

7.083

0.065***

15.3

−8.3%***

−7.605

0.288***

60.427

−0.008

−5.403

−0.067***

−3.131

0.010***

6.065 (continued)

74

4 Determinants of Board Composition

Table 4.4 (continued) N

Mean

4220

0.023

Private

1552

0.024

State

4220

0.032

1552

0.046

Roa

State

Roe

Private

Mean difference

t

−0.001

0.5731

−0.014

2.1755

a Test

Note of difference in proportions Source Computed by the authors

Fig. 4.1 Board size, 2000–2012. Source Plotted by the authors, it also appeared on “Determinants of Board Composition and Corporate Governance in Chinese Enterprises Since Reforms Began: A Comparison of Controlling Shareholders”

Board size 10.5 10 9.5 9 8.5 8 7.5 7

Year

6.5 2000

2005 Pool

Fig. 4.2 Changes in Board independence. Source Plotted by the authors, it also appeared on “Determinants of Board Composition and Corporate Governance in Chinese Enterprises Since Reforms Began: A Comparison of Controlling Shareholders”

2010 Private

State

Board independence 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05

Year

0 2000

2005 Pool

2010 Private

State

model was also estimated using the dynamic GMM method with robust standard errors to deal with endogeneity issues.

4.3 Board Size and Board Independence During the Reform

75

4.3.2 Determinants of Board Size Table 4.5 illustrates the determinants of board size during the period between 2000 and 2012 using different estimation methods. The LR test (P < 0.05) rejected the null hypothesis of OLS being nested in the fixed effect model, which suggests that each firm has its own characteristics and cannot simply be pooled together. As suggested by the Hausman test (P < 0.05), the fixed effects model is more suitable than the random effects model since the assumption of difference in coefficients of not being systematic is rejected. Therefore, the analysis in this model is mainly based on the fixed effects model. To guarantee endogeneity issues will not bias the estimates, the GMM method was used, which is more efficient in dealing with the endogeneity problem (Wintoki et al., 2012). The results from the GMM method are consistent with the fixed effects model results as the signs of the coefficients are the same. Furthermore, White’s robust standard error procedure helped mitigate heteroscedasticity problems. Consistent with the scope of operation hypothesis, Chinese board size is generally found to be positively correlated with firm size when using OLS, FE and GMM methods, which suggests that a larger firm with greater scope of operation needs a larger board size to monitor managerial behaviour and provide information and expertise to facilitate the firm’s daily operations. It also shows that corporate boards are an important resource provider for China’s listed firms. On the monitoring hypothesis, the cost of monitoring reduces board size significantly with respect to stock return volatility. The benefits of monitoring did not show any significant effects on board size as expected, meaning that China’s firms were more conscious of cost than benefits when constituting their corporate boards. To be specific, Tobin Q as a proxy of a firm’s growth opportunity tends to shrink with board size, though its effect is not significant. It appears to reflect the current situation in China where a fast-growing firm faces the challenge of fast decisionmaking and information acquiring costs as they compete with other firms. Thus, an efficient smaller board is deemed more appropriate. Similarly, stock return volatility represents the total risk, which reduces the corporate board size significantly, and suggests that when the corporate environment is uncertain, information asymmetry can entail a high cost for corporate boards to monitor managerial behaviour. For the investigated corporate governance variables, firstly, it is found that board size is positively correlated with supervisory board size, which is because the main function of supervisory boards is to monitor and advise the corporate board. Thus, a large board needs a large supervisory board. Secondly, ownership concentration is found positively correlated with board size, which is because ownership concentration embodies the interests of large shareholders, who have more interest in the firm and have more concerns regarding corporate governance issues. Thus, ownership concentration results in a large board size to take care of large shareholders’ interests. Thirdly, CEO duality tends to reduce board size, which is likely because CEO duality gives CEOs leadership roles in both the management team and corporate board. It is likely that CEOs could abuse their power by forming a smaller and

76

4 Determinants of Board Composition

Table 4.5 Determinants of board size

OLS

FE

GMM

firmage

−0.0123

0.0185

−0.000584

(0.0263)

(0.0223)

(0.0196)

firmsize

0.0506***

0.0259**

0.0253*

(0.00903)

(0.00840)

(0.0126)

−0.0223

−0.0248

−0.0103

(0.0393)

(0.0305)

(0.0261)

−0.000210

−0.00289

−0.00634

(0.00544)

(0.00369)

(0.00327)

−0.138**

−0.118**

0.0411

(0.0493)

(0.0378)

(0.0345)

fcf

0.0307

0.0190

0.00974

(0.0193)

(0.0137)

(0.0156)

hhi1

−0.0261

−0.0249

−0.0916

(0.0302)

(0.0434)

(0.0687)

0.0543***

0.0560***

0.0319**

(0.00518)

(0.00543)

(0.0123)

0.103*

0.131*

0.0233

(0.0506)

(0.0476)

(0.0392)

duality

−0.0241

−0.0280*

−0.0215

(0.0171)

(0.0118)

(0.0133)

state

−0.0122

−0.0216

−0.0203

(0.0270)

(0.0231)

(0.0180)

0.385*

0.232*

-0.190

(0.162)

(0.106)

(0.104)

govern03

0.0164

0.0127

0.0210

(0.0128)

(0.0113)

(0.0228)

govern08

−0.0462***

−0.0457***

−0.00585

(0.0119)

(0.00907)

(0.00585)

Scope

longdebt Cost tobinq1 sdreturn Benefit

Governance supersize concen10

manage

lnboardsizet-1

0.450***

boardinde

−0.709***

(0.0727) (0.199) _cons

0.967***

1.229*** (continued)

4.3 Board Size and Board Independence During the Reform Table 4.5 (continued)

N adj.

R2

77

OLS

FE

(0.267)

(0.232)

5772

5772

0.191

0.189

GMM 4884

Hansenp

0.206

AR(2) p

0.208

Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Source Computed by the authors

less independent board so that they could exert their influence to pursue their interests. Fourthly, managerial ownership tends to improve board size since managers’ concerns about corporate board composition increase with managerial ownership. Therefore, the results show that board size increases with managerial ownership. Finally, state ownership has no significant effects on board size. On government regulation, China’s government regulations, “Guidelines for Introducing Independent Directors” did not have a significant influence on corporate board size, which suggests that firms included independent directors to replace the inside directors so that board size remained nearly unchanged. Consequently, China’s corporate board has gradually transformed from an insider-dominated one to a more outsider-dominated one. The split-share structure reform that ended in 2007 was found to have significant negative effects on board size, which suggests that marketoriented reforms reduced board size, since a smaller board size helps fast decision making and is more efficient in dealing with market competition.

4.3.3 Determinants of Board Independence Table 4.6 shows the determinants of board independence using different regression methods all of which were estimated with robust standard errors. According to the LR test (P < 0.05), the panel regression methods were more suitable than the pooled OLS (P < 0.05) method. Furthermore, the Hausman test results suggest that the fixed effects model is better than the random effects model. On the scope of operation hypothesis, board independence significantly increases with firm size and firm age, which suggests that as the firms grow and diversify over time and territories, more independent directors are needed to monitor managerial behaviour. This is because firms with greater scope of operation usually face more serious agency problems than firms with smaller scope of operation. Independent directors can monitor managerial behaviour and mitigate agency problems better than inside directors. Hence, board independence increases with the scope of operation. On the monitoring hypothesis, the study found that board independence decreases with the cost of monitoring (tobinq), which is because a fast-growing firm with high

78 Table 4.6 Determinants of board independence

4 Determinants of Board Composition Scope

OLE

FE

GMM

Firmage

0.0155***

0.0272***

0.000651

(0.00251)

(0.00302)

(0.00419)

0.00394***

0.00603*

0.00765*

(0.00103)

(0.00253)

(0.00333)

0.00190

0.0107

0.000865

(0.00586)

(0.00993)

(0.0120)

tobinq

−0.000993

−0.00506***

−0.00324**

(0.00106)

(0.00138)

(0.00117)

sdreturn

0.0172

0.00415

0.0502**

(0.0160)

(0.0162)

(0.0176)

−0.00182

0.00364

0.00103

(0.00787)

(0.00802)

(0.00823)

−0.00608

−0.0350

0.0487

(0.00702)

(0.0229)

(0.0561)

−0.000353*

−0.000737**

0.000114

(0.000168)

(0.000278)

(0.000412)

0.000453

0.000921

0.0000728

(0.000332)

(0.000566)

(0.000747)

supersize

−0.00430***

−0.00613**

0.000279

(0.000831)

(0.00202)

(0.00581)

concen10

−0.00861

−0.0698***

−0.0446

(0.00768)

(0.0164)

(0.0308)

−0.00396

0.00263

−0.00772

(0.00340)

(0.00522)

(0.0142)

state

−0.00637

−0.0109

−0.00601

(0.00537)

(0.00688)

(0.0110)

manage

−0.0572*

−0.0789

0.0750

(0.0228)

(0.0406)

(0.116)

0.227***

0.216***

0.125

(0.00396)

(0.00337)

(0.0675)

0.0126***

0.00185

0.00184

(0.00227)

(0.00282)

(0.00236)

Firmsize Longdebt Cost

Benefit fcf hhi

CEO’s Influence ceoage ceotenure Governance

duality

Govern03 Govern08 lnboardsize

−0.121*** (continued)

4.3 Board Size and Board Independence During the Reform Table 4.6 (continued)

Scope

OLE

79 FE

GMM (0.0319)

boardindet−1

0.141 (0.0893) −0.0548

−0.116*

(0.0286)

(0.0530)

N

5772

5772

adj. R2

0.663

0.706

_cons

4884

hansenp

0.143

ar2p

0.225

Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Source Computed by the authors

Tobin q is costly for independent directors to access information to perform their monitoring activities. Therefore, a fast-growing firm tends to reduce independent directors to save cost. On the bargaining hypothesis, CEO’s age exerts a negative influence on board independence, which means that an older CEO with reputation has more bargaining power, which tends to result in corporate boards dominated by insiders. As for other corporate governance variables, firstly, the supervisory board showed a negative effect on board independence, which is because independent directors and the supervisory boards share similar functions since both are responsible for monitoring managers’ and directors’ behaviour. Thus, the supervisory board is able to substitute for independent directors. Secondly, ownership concentration has a negative effect on board independence, which means that large shareholders tend to substitute for independent directors in monitoring managerial behaviour. Majority shareholders who have more interests in the firm will assume more responsibilities in corporate governance than minority shareholders who have less interest in the firm. Due to conflict of interests with minority shareholders, high ownership concentration unites large shareholders to shoulder more responsibilities in corporate governance, which results in the substitution of independent directors. Finally, China’s government regulation, “Guidelines for introducing independent directors” has promoted board independence significantly in China following which, the split-share structure reform has the tendency to shrink board independence in firms as shown in the fixed effects model.

80

4 Determinants of Board Composition

4.3.4 Determinants of Board Size and Board Independence During Reform This section compares the determinants of board size and board independence before and after the split-share structure reform was introduced, and between private and state enterprises. State enterprises are further subclassified into those controlled by central government, local government and SOE entities. The regression method used for the comparison is the fixed effects estimation model with robust clustered standard errors. Fixed effect methods address endogeneity problems due to the unobserved and time-invariant heterogeneities, which is more effective than pooled OLS. Besides, we follow the method used by Li, Lu, Mittoo and Zhang (2015), which also use the fixed effects model to compare the effectiveness of corporate governance in state and private enterprises in China.

4.3.5 Determinants of Board Size and Board Independence Before and After Split-Share Structure Reform Table 4.7 compares the determinants of board size and board independence before and after the split-share structure reform was introduced. On the determinants of board size, board size increases with firm age, but supervisory board size decreases with CEO duality before the reform, but not after the reform. This suggests that the supervisory board has played an important role in complementing the main board of directors to monitor managerial behaviour before the reform, but not after the reform. CEO duality resulted in a smaller board only before the reform, which means that CEO’s leadership increases his or her bargaining power over a smaller board to facilitate his or her own interests. After the reform, firms have been more concerned with cost of monitoring than before the reform. As for determinants of board independence, the results show that board independence was positively correlated with scope of operation prior to reform, but that relationship was no longer significant after reform. In addition, board independence which was negatively correlated with Tobin Q before the reform became positively correlated with Tobin Q after the reform. Better market performance was associated with higher cost of monitoring before the reform due to information asymmetry and acquisition problems. After reform, China’s listed firms became more market-oriented because the reform encouraged China’s domestic investors to be more concerned with corporate governance issues. Therefore, we can conclude that board independence increased Tobin Q after reform. On the other corporate governance variables, the analysis revealed that supervisory board and managerial ownership have negative effects on board independence before the reform, which means that independent directors can be substituted by other firmlevel corporate governance mechanisms before the reform. In contrast, the role of

4.3 Board Size and Board Independence During the Reform

81

Table 4.7 Determinants of board size and board independence, before and after split-share structure reform Board size

Board independence

Before

After

Before

After

Firmage

0.0396 **

−0.00664

0.139***

0.00452

(0.0141)

(0.0105)

(0.00977)

(0.00327)

Firmsize

0.0209

−0.0107

0.0409***

0.00402

(0.0151)

(0.0105)

(0.0105)

(0.00273)

−0.0428

−0.0245

−0.00776

0.00282

(0.0452)

(0.0364)

(0.0288)

(0.00939)

tobinq

−0.0100

−0.0147 *

−0.140***

0.00349**

(0.0171)

(0.00575)

(0.0163)

(0.00114)

sdreturn

−0.0133

0.0538

−0.0189

−0.0273

(0.137)

(0.0800)

(0.0915)

(0.0217)

Fcf

0.0107

−0.0105

−0.0120

−0.000151

(0.0249)

(0.0270)

(0.0197)

(0.00649)

Hhi

0.124

0.129

−0.135

0.0157

(0.163)

(0.0785)

(0.0987)

(0.0259)

supersize

0.0333**

−0.0324

−0.0171**

−0.00522

(0.0121)

(0.0170)

(0.00592)

(0.00407)

concen10

−0.00196

0.0784

−0.227

−0.0374

(0.206)

(0.0818)

(0.177)

(0.0199)

Scope

Longdebt Cost

Benefit

Governance

−0.0481*

−0.0234

−0.0123

0.00664

(0.0206)

(0.0193)

(0.0173)

(0.00631)

State

0.0479

−0.00376

−0.127 ***

0.0100

(0.0611)

(0.0237)

(0.0377)

(0.00743)

manage

−0.0273

−0.00258

−1.081**

0.0554

(0.433)

(0.101)

(0.389)

(0.0306)

−0.00157

0.000431

(0.000961)

(0.000317)

duality

CEO’s Influence ceoage ceotenure _cons

1.334***

2.584***

0.0198***

−0.000124

(0.00326)

(0.000625)

−1.250***

0.252*** (continued)

82

4 Determinants of Board Composition

Table 4.7 (continued) Board size

Board independence

Before

After

Before

After

(0.382)

(0.253)

(0.297)

(0.0640)

N

2220

2220

2220

2220

R2

0.036

0.031

0.513

0.027

Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Source Computed by the authors

independent directors in corporate governance improved after reform so that other corporate governance mechanisms did not supplant their role.

4.3.6 Determinants of Board Size: A Comparison of Controlling Shareholders Table 4.8 compares the determinants of board size between state enterprises and private enterprises. On scope of operation hypothesis, China’s central governmentcontrolled enterprises and SOE entity-controlled enterprises configured their board size based on firm size, whereas local government-controlled and private enterprises did not. As for the investigated corporate governance variables, supervisory board size was found to be the key determinant of board size in all types of firms, which means that the role of the supervisory board was significant in all types of firms as a complement to the main board of directors. Besides, the analysis revealed that CEO duality tends to entrench a CEO in the firm by leading to a smaller board in private enterprises, central government-controlled and local government-controlled enterprises. In private enterprises, the founders tend to install themselves as the CEO or board chairman to gain power to bargain for a smaller board size. In contrast, in central government and local government-controlled enterprises, CEO and board chairman are government officials. They are most likely promoted through government agencies and derive their power through CEO duality by forming a smaller board. On managerial ownership, the results show that the incentive alignment between managers and shareholders through managerial ownership has significant positive effects on board size in private enterprises, which suggests that managers in private enterprises are more likely to express concerns in corporate governance through large board delegations. On government regulation effects, the results show that government regulations, “Guidelines for introducing independent directors” increased board size in central government-controlled enterprises, which means that central government-controlled enterprises increased the number of independent directors without adjusting board

4.3 Board Size and Board Independence During the Reform

83

Table 4.8 Determinants of board size: a comparison of controlling shareholders Private

Central

Local

MOSOE

0.0399

−0.112

−0.0162

−0.00451

(0.0492)

(0.0626)

(0.0328)

(0.0287)

firmsize

0.0187

0.0435**

0.0208

0.0417 **

(0.0149)

(0.0158)

(0.0156)

(0.0129)

longdebt

−0.0305

0.0341

−0.0235

−0.0492

(0.0490)

(0.0657)

(0.0556)

(0.0451)

−0.00742

0.00730

−0.00266

0.00110

(0.00505)

(0.00758)

(0.00605)

(0.00759)

−0.111

0.00202

−0.0860

−0.139

(0.0655)

(0.0781)

(0.0613)

(0.0878)

0.0330

0.0350

0.00583

0.0243

(0.0208)

(0.0479)

(0.0224)

(0.0246)

−0.0583

−0.0187

−0.0331

−0.0267

(0.0658)

(0.0485)

(0.0444)

(0.0509)

supersize

0.0632***

0.0421**

0.0582***

0.0465***

(0.0111)

(0.0134)

(0.00818)

(0.00778)

concen10

0.121

0.114

0.131

0.0120

(0.0715)

(0.127)

(0.0824)

(0.0807)

−0.0433*

−0.0824*

−0.0456 *

−0.00807

(0.0170)

(0.0330)

(0.0216)

(0.0204)

state

0.0588

−0.0261

−0.0508

0.0104

(0.0629)

(0.0389)

(0.0294)

(0.0307)

manage

0.244*

-0.223

0.393

0.0421

(0.119)

(0.164)

(0.266)

(0.348)

0.0215

0.0639**

0.0313

0.00786

(0.0233)

(0.0244)

(0.0216)

(0.0150)

govern08

−0.0610**

−0.00474

−0.0310**

−0.0389*

(0.0193)

(0.0164)

(0.0120)

(0.0161)

_cons

1.169*

1.923***

1.603***

1.188***

(0.479)

(0.424)

(0.411)

(0.361)

1552

458

1916

1846

Scope firmage

Cost tobinq sdreturn Benefit Fcf Hhi Governance

duality

govern03

N

(continued)

84

4 Determinants of Board Composition

Table 4.8 (continued) adj. R2

Private

Central

Local

MOSOE

0.1587

0.1607

0.1877

0.1742

Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Source Computed by the authors

size. Moreover, the split-share structure reform tended to reduce board size significantly in all types of firms, except in central government-controlled firms, which means that the split-share structure reform did not affect board structure in central government-controlled enterprises.

4.3.7 Determinants of Board Independence: Comparing Controlling Shareholders Table 4.9 compares the determinants of board independence among firms with different controlling shareholders. On the scope of operation hypothesis, the results show that board independence was significantly determined by firm age, which means that older firms with greater scope of operations tend to employ more independent directors in all types of firms. As for the cost of monitoring, the evidence shows that only private enterprises and SOE entity-controlled state enterprises have been concerned about cost when adding independent directors, which is because private enterprises and SOE entitycontrolled state enterprises are more profit-oriented, and hence, focus on reducing costs to improve efficiency. However, government-controlled enterprises have objectives, such as to improve the employment rate, balance state fiscal revenue and achieve social stability. It is easier for government-controlled enterprises to get financial support from the government. Furthermore, China’s domestic investors believe that the government will not expropriate their interests. Thus, government-controlled firms are less concerned about cost issues. As for the benefit of monitoring, it was found that only local government-controlled and market-oriented state enterprises consider the benefits of monitoring as important when adding independent directors. With regards to CEO influence, the results show that CEOs’ age tends to reduce board independence in local government-controlled firms, which means that older CEOs tend to entrench themselves by forming a less independent board to pursue their interests in local government-controlled firms, which are located far from central government’s power centre in Beijing. CEO tenure was positively correlated with board independence in local government-controlled enterprises and market-oriented state enterprises, which suggests that CEOs with long tenure in local governmentcontrolled and market-oriented state enterprises require more independent directors

4.3 Board Size and Board Independence During the Reform

85

Table 4.9 Determinants of board independence, comparison of controlling shareholders Private

Central

Local

MOSOE

0.0981***

0.0627**

0.0977***

0.191***

(0.0232)

(0.0205)

(0.0157)

(0.0214)

firmsize

0.0114

−0.00479

0.0146*

0.0140

(0.00800)

(0.00846)

(0.00589)

(0.00905)

longdebt

0.00968

0.0226

−0.0137

0.000619

(0.0232)

(0.0336)

(0.0175)

(0.0281)

−0.00498*

−0.000790

−0.00112

−0.0191***

(0.00206)

(0.00259)

(0.00215)

(0.00453)

0.00310

0.0126

−0.0261

−0.0309

(0.0283)

(0.0394)

(0.0265)

(0.0424)

-0.00517

-0.00968

0.0254*

-0.0116

(0.0131)

(0.0126)

(0.0129)

(0.0181)

0.00667

-0.0425

-0.00275

-0.104*

(0.0574)

(0.0268)

(0.0364)

(0.0466)

ceoage

−0.000379

−0.000625

−0.000987*

−0.000636

(0.000517)

(0.000576)

(0.000455)

(0.000623)

ceotenure

−0.000715

−0.00112

0.00160*

0.00351**

(0.00118)

(0.00127)

(0.000726)

(0.00133)

−0.00232

−0.00806

−0.0117***

−0.0000850

(0.00359)

(0.00712)

(0.00311)

(0.00455)

−0.0545

−0.00607

−0.0516

−0.0257

(0.0285)

(0.0528)

(0.0300)

(0.0455)

duality

0.00297

0.0229**

0.0104

0.0103

(0.00973)

(0.00800)

(0.00713)

(0.0149)

state

0.0382

0.0304

0.0254**

−0.00209

(0.0285)

(0.0163)

(0.00962)

(0.0158)

−0.0510

0.163**

−0.150

−0.877**

(0.0483)

(0.0573)

(0.0935)

(0.283)

govern03

0.170***

0.134***

0.210***

0.160***

(0.00924)

(0.0109)

(0.00826)

(0.00741)

govern08

−0.00503

0.00988

−0.0173**

−0.0609***

Scope firmage

Cost tobinq sdreturn Benefit Fcf Hhi CEO’s Influence

Governance supersize concen10

manage

(continued)

86

4 Determinants of Board Composition

Table 4.9 (continued)

_cons N adj.

R2

Private

Central

Local

MOSOE

(0.00623)

(0.00686)

(0.00533)

(0.00924)

−0.824***

−0.143

−0.870***

−1.601***

(0.150)

(0.166)

(0.155)

(0.198)

1552

458

1916

1846

0.596

0.207

0.627

0.630

Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Source Computed by the authors

to monitor their behaviour. CEOs are less likely to use their power to reduce board independence in these enterprises. On other corporate governance variables, the results show that in private enterprises, other corporate governance mechanisms do affect board independence. In central government-controlled enterprises, CEO duality and managerial ownership increase board independence, which means that more independent directors are needed to monitor managerial and CEO behaviour. Also, independent directors could be substituted by supervisory boards in local government-controlled firms and by managerial ownership in market-oriented state enterprises. As for regulation effects, the “Guidelines for introducing independent directors” had an impact on board independence in only some type of firms. The splitshare structure reform reduced board independence in local government-controlled and market-oriented state enterprises. However, board independence in the central government-controlled and private enterprises was unaffected by the split-share structure reform.

4.4 Summary This chapter examined the determinants of board size and board independence in China and compared the determinants before and after the split-share structure reform, as well as among different controlling shareholders. The influence of the scope of operation, monitoring cost and benefits, CEO’s influence, and other firmlevel corporate governance mechanisms, as well as the effects of government regulations were found to have had varying relationships by period and by ownership. The findings significantly contribute to the body of China’s corporate governance studies using the most comprehensive analytic frameworks. Both static and dynamic estimation methods were used. Overall, it was found that Chinese board size and independence are jointly determined by scope of operation, cost and benefit of monitoring, CEO’s influence and other governance factors. Among these factors, the government was the most important player in constituting corporate boards before the split-share

4.4 Summary

87

structure reform; but not after the reform as board independence became more important than other governance factors. In addition, when constituting corporate boards, private and market-oriented state enterprises were found to be more concerned about cost than are government-controlled enterprises. As for the other corporate governance variables, board size increases with supervisory board size, ownership concentration and managerial ownership, whereas board independence decreases with supervisory board size, ownership concentration, managerial ownership and state ownership, which suggests that other corporate participants who have more interest in the firm tend to substitute inside directors in place of independent directors to serve their own interests. Besides, China’s enterprise board composition is heavily influenced by government regulations. China’s board independence increased significantly with CSRC regulations that required at least one third independent directors by June 2003, whereas board size was almost unaffected by this regulation, which means that China’s listed firms reduced the number of insider directors while increasing the number of independent directors to adhere to government regulations. In addition, the split-share structure reform (which ended in 2007) reduced both board size and board independence, suggesting that ownership decentralization through releasing shares resulted in smaller and less independent boards. Despite the overall trend of board independence growing slightly after the reform (Fig. 4.2), the speed of change declined dramatically. The comparison of determinants across periods and controlling shareholders shows that the impact of these determinants differed. Before reform, board independence was determined by the cost of monitoring and other corporate governance mechanisms such as supervisory board size and managerial ownership. After reform, board independence was no longer affected by most of the investigated factors. Furthermore, the study found that private enterprises and market-oriented state enterprises are more concerned about cost when adding independent directors than government-controlled firms. Meanwhile, the CEO’s influence by age, tenure and leadership by CEO duality only had significant influence in state enterprises. Other corporate governance mechanisms show little influence on board independence in private enterprises, suggesting that board independence is an important mechanism for private enterprises and the split-share structure reform did not affect board independence in private enterprises. Taken together, it can be argued that China’s government regulations and legal enforcement measures have played an important role in constituting the corporate governance system in the country during the transition period, a role that should not be neglected in future corporate governance studies. Besides, the government policy of decentralizing non-tradable shares to revitalize the capital market changed China’s board composition and in the way China’s listed firms constitute corporate boards. Board composition became stable after the reform, and market performance became a factor favouring board independence after the reform. In short, evidence show that China’s government strategy of decentralizing state enterprises through “grabbing the large and letting go of the small” has produced positive results toward market reforms. While private and market-oriented state enterprises considered cost saving

88

4 Determinants of Board Composition

when constituting their corporate boards, central and local government-controlled firms did not. Finally, the chapter provided empirical evidence for corporate governance theories and of institutional arrangements as important factors to analyze the relationship between shareholders, CEOs, board directors and managers.

Chapter 5

Corporate Governance Mechanisms and Firm Performance

5.1 Introduction Ownership structure, board structure and CEO characteristics are important internal corporate governance mechanisms. However, the empirical evidence over the relationship between them has produced mixed results with significant variation across different countries and institutional environments. Yet, analysing the effectiveness of corporate governance mechanisms and their joint effects is crucial for transition economies, (such as China), as it is part and parcel of initiatives to further corporate governance and state enterprise reforms. This chapter analyses the influence of board structure, ownership structure and CEO characteristics on firm performance since the split-share structure reforms were launched in China in 2005–2007. In Sect. 5.2, we discuss both “linear” and “nonlinear” relationships between the corporate governance variables and firm performance. The “linear” effect of board structure, ownership structure and CEO characteristics on firm accounting performance (ROA, ROE) and market performance (Tobin Q) is analysed in Sect. 5.2.1. Section. 5.2.2 checks the “nonlinear” relationships between them (Al Farooque, Van Zijl, Dunstan, and Karim, 2007). Section 5.3 examines the joint effect of different corporate governance mechanisms, e.g. board independence and ownership concentration, on firm performance. Section 5.4 examines the impact of different controlling shareholders, including the central government, local government, SOE entity and private shareholders on firm performance. The findings indicate that controlling shareholder types vary in their influence on firm performance. Section 5.5 compares the effectiveness of corporate governance mechanisms by different controlling shareholders. Section 5.6 compares the effectiveness of corporate governance mechanisms before and after the split-share structure reform. Section 5.7 concludes with a summary of findings.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0_5

89

90

5 Corporate Governance Mechanisms and Firm Performance

5.2 Influence of Corporate Governance Mechanisms on Firm Performance 5.2.1 Linear Estimation The results of the linear and nonlinear relationship between corporate governance mechanisms and firm performance, (including, ROA, ROE and Tobin Q) are shown in Tables 5.1 and 5.2. The OLS, fixed effects and dynamic system GMM methods with White’s clustered robust standard errors were used to estimate the impact of board structure, ownership structure, CEO characteristics and government regulations on firm performance. The use of fixed effects are based on the Hausman test (P < 0.05) and LR test (P < 0.05), which were discussed in Chapter 3. The results produced highly consistent outcomes. The model has no serious multicollinearity problem since the highest correlation coefficient between board independence and firm age was 0.5 (see Table 4.2). These results also passed the VIF threshold value, and hence, multicollinearity issues do not affect the robustness of the results. According to the “linear” model results shown in Table 5.1, board independence as a symbol of good corporate governance negatively affected firm performance in terms of ROA, ROE and Tobin Q, which suggests that China’s independent directors are not effective in improving firm performance. Four reasons are adduced to explain this. Firstly, China’s independent directors lack incentives to serve their companies. The CSRC requires that independent directors cannot hold a significant number of shares. Meanwhile, there are no explicit rules specifying independent directors’ compensation, which are primarily monetary-based, such as annual salaries, allowances for transportation and for attending conferences, while they rarely receive stock incentives. Secondly, independent directors are not accountable for future corporate development since they have no right to participate in corporate operations directly and cannot provide consultancy advice. The assessment of their performance is not based on how much they contribute to firm performance. Thirdly, independent directors hardly act autonomously as the selection of independent directors is affected by the CEO’s power. Powerful CEOs find it easier to handpick the “independent” directors who then will not challenge their leadership (Shivdasani and Yermack, 1999). Besides, in China, the government still leads corporate governance, and therefore, there is a strong possibility that independent directors are politically influenced to represent government interests rather than shareholders’ interests (Wang, 2015). Fourthly, key agency problems exist between the controlling large shareholders and minority shareholders in China since ownership is highly concentrated rather than dispersed as in Western countries. The monitoring role of independent directors is minimized when there is no serious conflict between managers and shareholders. Although the supervisory board can make up for the deficiency shown by independent directors, its positive impact largely manifests itself only in accounting performance. As suggested by Hu et al. (2010), China’s supervisory boards are comprised by large shareholders, which consequently renders external governance mechanisms too weak to exert a positive influence on firm performance.

Duality

Ceotenure

Ceoage

0.001

(0.001)

0.000

(0.000)

−0.007

(0.000)

(0.000)

0.00218***

0.000

(0.015)

(0.012)

0.007

(0.001)

0.00136*

(0.000)

0.000

(0.015)

0.0466**

(0.234)

(0.300)

(0.230)

0.0984***

−0.034

−0.007

0.0451***

(0.008)

(0.007)

−0.013

(0.002)

0.012

0.536*

(0.006)

(0.002)

(0.001)

−0.014

0.001

−0.001

0.001

−0.008

(0.011)

0.005

(0.008)

−0.003

(0.006)

(0.002)

-0.002

(0.126)

0.495***

(1.542)

−0.913

(0.051)

0.015

(0.013)

0.011

(0.074)

0.109

(0.108)

−0.316**

FE

−0.039

(0.003)

−0.008

(0.004)

0.0185*** 0.0112**

(0.001)

−0.002

(0.085)

0.092

(1.220)

3.548**

(0.042)

0.053

(0.007)

0.007

(0.045)

0.003

−0.346** (0.108)

(0.013)

(0.015)

−0.0360** −0.0364** −0.004

(0.013)

OLS

0.000

CEO’s character

Concen10

Manage

State

Ownership structure

Supersize

Inboardsize

Boardinde

Board structure

ROE GMM

OLS

FE

ROA

Table 5.1 Corporate governance mechanisms and firm performance, Linear estimation

0.016

(0.004)

GMM

(0.021)

0.001

(0.087)

−0.113

(0.138)

(0.027)

0.040

(0.119)

−0.140

(0.190)

−0.592*** 0.338

FE

(0.003)

0.00748**

−0.051

(0.007)

(0.096)

0.024

(continued)

−0.049

(0.008)

−0.002 (0.007)

(0.004) −0.003

0.0216***

(0.199)

−0.408*

(3.838)

−3.524

(0.003)

0.004

(0.202)

−0.514* (0.158)

(3.849) −0.046

8.520*

(0.101) (3.364)

0.533

(0.070)

−0.399*** −0.656*** −0.651***

(0.015)

0.019

(0.090)

0.039

(0.186)

−0.084

OLS

Tobin Q

0.0153*** 0.003

(0.003)

−0.003

(0.138)

0.125

(1.624)

−0.604

(0.052)

−0.016

(0.015)

0.0304*

(0.091)

0.048

(0.114)

−0.229*

GMM

5.2 Influence of Corporate Governance Mechanisms on Firm Performance 91

ssROAt−1

Govern08

Govern03

Sale grow

Fa expen

FCF

Sdreturn

Longdebt

Firmsize

Firmage

Controls

0.0127***

(0.004)

0.006

(0.004)

0.002

(0.004)

0.000

(0.002)

(0.002)

(0.004)

0.0246***

(0.029)

(0.029)

0.0245***

0.120***

(0.011)

0.149***

(0.011)

(0.020)

(0.020)

0.0388* **

−0.002

−0.0688***

0.0475***

−0.006

(0.009)

−0.0240**

(0.003)

(0.002)

(0.008)

0.003

(0.007)

(0.004)

0.0113***

−0.001

−0.048 (0.030)

0.260***

(0.003)

0.000

(0.004)

0.003

(0.002)

0.0214***

(0.029)

0.0566*

(0.012)

0.0303**

(0.019)

0.007

(0.008)

−0.011

(0.002)

(0.027)

0.030

(0.036)

0.046

(0.014)

0.121***

(0.229)

0.485*

(0.100)

0.187

(0.166)

−0.538**

(0.044)

0.012

(0.015)

0.00781*** 0.0751***

(0.005)

0.000

(0.006)

(0.034)

OLS

(0.004)

ROE GMM

OLS

FE

ROA

−0.004

(0.004)

Table 5.1 (continued)

(0.032)

0.052

(0.038)

0.031

(0.017)

0.126***

(0.193)

0.570**

(0.103)

0.135

(0.167)

−0.242

(0.057)

0.076

(0.026)

0.0562*

(0.050)

0.001

(0.040)

FE

(0.049)

0.091

(0.052)

OLS

Tobin Q

(0.093)

0.398***

(0.052)

FE

(0.055)

−0.028

(0.089)

GMM

(0.268)

1.269***

(0.093)

(0.024)

0.021

(0.035)

0.049

(0.015)

(0.024)

(0.035)

−0.003

(0.270)

0.312

(0.109)

−0.081

(0.249)

3.236***

(0.083)

−0.066

(0.031)

(0.046)

0.542***

(0.051)

(0.050)

0.405***

(0.039)

(continued)

(0.047)

0.034

(0.031)

−0.241*** −0.246*** −0.028

(0.027)

0.0862***

(0.287)

0.869**

(0.091)

(0.242) −0.115

(0.269)

3.007***

(0.110)

−0.092

(0.048)

−0.174

3.071***

(0.099)

−0.319**

(0.033)

0.0987*** 0.0920***

(0.218)

0.332

(0.121)

0.073

(0.166)

−0.093

(0.054)

0.017

(0.015)

0.0629*** −0.369*** −0.506*** −0.226***

(0.035)

−0.059

(0.055)

GMM

92 5 Corporate Governance Mechanisms and Firm Performance

*p

−1.276***

*** p

(0.360) 0.067

5772

(0.044)

0.141**

GMM

0.041

5772

(0.487)

0.764

0.159

5328

(0.415)

−1.767*** −1.097**

FE

Note Standard errors in parentheses, < 0.05, < 0.01, < 0.001 Source Computer by the authors—Dana-revise all figures other than N in the table to 3 decimal pointss

** p

0.084 0.070

AR(2) P

0.092

5328

(0.054)

−0.210***

Hansen P

0.137

Adj. R2

(0.063)

5772

(0.050)

5772

−0.088

OLS

(0.030)

ROE GMM

OLS

FE

ROA

−0.211***

N

_Cons

Tobinqt−1

ROEt−1

Table 5.1 (continued)

0.292

5772

(0.708)

8.018***

OLS

Tobin Q

0.258

5772

(1.088)

9.507***

FE

0.897

0.089

5328

(0.728)

5.026***

(0.039)

0.449***

GMM

5.2 Influence of Corporate Governance Mechanisms on Firm Performance 93

94

5 Corporate Governance Mechanisms and Firm Performance

Table 5.2 Corporate governance and firm performance, nonlinear estimation ROA

TobinQ

Board structure Boardinde

−0.122***

(0.0297)

−2.118***

Boardinde2

0.193**

(0.0664)

3.547***

(0.686)

Lnboardsize

0.0145

(0.0886)

−0.625

(0.832)

(0.265)

lnboardsize2

−0.000719

(0.0197)

0.146

(0.184)

Supersize

−0.000936

(0.00159)

0.00122

(0.0208)

State

−0.0321

(0.0213)

−1.343***

(0.266)

State2

0.0324

(0.0352)

1.201**

(0.422)

Manage

−2.001

(1.030)

13.53

(13.22)

Ownership structure

Manage2

75.19*

(35.76)

−199.1

(443.0)

Concen10

0.0804

(0.0650)

0.182

(0.938)

Concen102

0.0180

(0.0615)

−0.620

(0.856)

Ceoage

−0.00425

(0.00259)

0.0125

(0.0285)

Ceoage2

0.0000425

(0.0000278)

−0.0000921

(0.000308)

Ceotenure

0.00240

(0.00132)

−0.000740

(0.0154)

Ceotenure2

−0.000134

(0.000115)

−0.000181

(0.00150)

Duality

−0.000944

(0.00439)

0.0156

(0.0527)

Firmage

0.00427

(0.00724)

0.497***

(0.0989)

Firmsize

0.00267

(0.00306)

−0.513***

(0.0480)

Longdebt

−0.00774

(0.00910)

−0.112

(0.108)

Sdreturn

−0.00337

(0.0197)

3.027***

(0.246)

FCF

0.0382***

(0.0108)

−0.122

(0.0900)

Fa expen

0.118***

(0.0291)

0.888**

(0.284)

Sale grow

0.0245***

(0.00211)

0.0832***

(0.0239)

Govern03

0.00218

(0.00407)

−0.245***

(0.0383)

Govern08

0.00990*

(0.00401)

0.354***

(0.0521)

_Cons

−0.0434

(0.129)

8.992***

(1.577)

N

5772

Adj. R2

0.095

CEO’s character

Controls−

*

Standard errors in parentheses, p < 0.05, Source Computed by the authors

5772 0.265 **

p < 0.01,

***

p < 0.001

5.2 Influence of Corporate Governance Mechanisms on Firm Performance

95

Furthermore, board size has no significant influence on firm performance, though CEOs tenure significantly improves firms’ accounting performance, which suggests CEOs learn and familiarize themselves with firms’ operations and jobs gradually over time, leading to improvement in firms’ accounting performance but not in firms’ market performance. This is likely because CEOs with long tenure tend to overly commit to a fixed paradigm even as they lose touch with the outside environment (Hambrick and Fukutomi, 1991). However, market conditions are dynamic and difficult to anticipate and comprehend (Henderson et al., 2006). The results show that CEOs’ age positively affects firms’ market performance suggesting that older CEOs are more able to capture market information than younger CEOs. State ownership showed a negative relationship with firms’ market performance, which is because of the government possessing non-profit objectives that conflict with the objectives of other domestic investors. Therefore, under normal market conditions, state ownership is not an effective corporate governance mechanism to protect non-government shareholders’ interests. Also, it is unfavourable to firms’ market performance due to restrictions imposed on the trading of state-owned shares as they have to be measured using book values. Managerial ownership improved firm performance as expected, which suggests that incentive alignment between shareholders and managers motivate managers to work towards improving firm performance. Ownership concentration as an important internal governance mechanism improves firms’ accounting performance as reflected in ROA and ROE values, but reduces market performance as reflected in Tobin Q, which is because large shareholders have more interest in the firms and, therefore, frequently participate in corporate governance, decision-making, strategic choice and in the management process. Thus, ownership concentration improves firms’ accounting performance (ROA and ROE), which is consistent with Li et al.’s (2015) findings. However, it exacerbates horizontal agency problems between large shareholders and small shareholders. Large shareholders have long-term orientation towards firm performance, whereas small shareholders may pursue short-term gains. Small shareholders with little equity interests tend to participate little in the corporate governance of firms. They are mainly “free riders”, who seek to appropriate short-term gains. They tend to pay less attention to corporate governance and longterm development. Moreover, the ownership power of large shareholders may also lead them to expropriate small shareholders’ interests (Heugens et al., 2009). Hence, ownership concentration is detrimental to firms’ market performance (Tobin Q). Among the effects of government regulations, the split-share structure reform had significantly improved firm accounting and market performance, which suggests that market-oriented share structure reforms have improved firm performance since government intervention was reduced and enabled firms to be more profit oriented and value maximizing. However, the “regulation of guidelines for introducing independent directors” has had no significant effect on a firm’s accounting performance, but negatively affected market performance, providing additional evidence that the independent director system has been detrimental to firm performance.

96

5 Corporate Governance Mechanisms and Firm Performance

Table 5.3 U-test, board independence and firm performance ROA Lower bound

Tobin Q Upper bound

Lower bound

Upper bound

Interval

[0, 0.545]

Slope

−0.121

0.088

−2.148

[0, 0.545] 1.785

t-value

−4.045

1.816

−8.179

3.335

P>t

0.0000

0.0350

0.000

0.000

Extreme point:

0.315

0.298

Source Computed by the authors Note: *p < 0.05, **p < 0.01, ***p < 0.001; The bold numbers simply show that the parameters are statistically significant

5.2.2 Nonlinear Estimation Besides linear estimation, we checked the U-shaped relationships estimated by Al Farooque et al. (2007) by adding the squared terms to the linear models to inspect changes in the signs of the coefficients. The results are shown in Table 5.2. Overall, board independence and state ownership have negative effects on firm performance (consistent with the linear estimation results in Table 5.1), but their squared terms have the opposite effects, which suggests that their relationships with firm performance are likely to be U-shaped. However, the effects of ownership concentration, managerial ownership, CEO age and CEO tenure on firm performance tend to be monotonic as coefficients of the squared terms are insignificant. To confirm the U-shaped relationship between board independence, state ownership and firm performance, the U-test (Kostyshak, 2015) using Stata software was conducted (see Table 5.3). The analysis found that the U-shaped relationship between board independence was statistically significant with ROA and Tobin Q showing turning points at 0.315 and 0.298, respectively, which means that more than 31.5 and 29.8% of independent directors are positively related to firm performance for otherwise it will be negative. Although the ROE and state ownership also show a significant coefficient in the regression, their U-shaped link was not statistically significant. To observe the trend of board independence and firm performance graphically, Figs. 5.1, 5.2 and 5.3 were plotted to show changes in firm performance for the different ranges of board independence. The horizontal axis shows the ranges involving board independence, whereas the bars show the mean values of ROA, ROE and Tobin Q. The values of both ROA and ROE declined to their lowest point between 25 to 30% of independent directors. Both reached their highest point at about 45 to 50% of independent directors, and the second largest value between 5 to 10% of independent directors. The results show that board independence can improve firm performance only after reaching 30% of independent directors. The changes also show that the relationships between board independence and ROA and ROE are U-shaped. Also, Tobin Q showed a U-shaped relationship with share of

5.2 Influence of Corporate Governance Mechanisms on Firm Performance

97

ROA 0.06 0.05 0.04 0.03 0.02 0.01 0 0>0.05

0.1-0.15 ROA

0.2-0.25

0.3-0.35

0.4-0.45

0.5
0.05

0.1-0.15

0.2-0.25

0.3-0.35

0.4-0.45

0.5 chi2 = 0.0000

LR test of pooled regression versus panel regression Regressions for determinants of board independence (Chap. 4) Likelihood-ratio test

LR chi2(443) = 1106.52

Assumption: OLS nested in FE

Prob > chi2 = 0.0000

The Hausman test of fixed effects model versus random effects model Regressions for determinants of board independence (Chap. 4) Hausman test

chi2(14) = 81.66

Assumption:difference in coefficients not systematic

Prob > chi2 = 0.0000

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0

143

144

Appendix B

LR test of pooled regression versus panel regression Regressions for corporate governanceon ROA (Chap. 5) Likelihood-ratio test

LR chi2(443) = 1297.74

Assumption: OLS nested in FE

Prob > chi2 = 0.0000

The Hausman test of fixed effects model versus random effects model Regressions for corporate governance on ROA (Chap. 5) Hausman test

chi2(14) = 291.63

Assumption:difference in coefficients not systematic

Prob > chi2 = 0.0000

LR test of pooled regression versus panel regression Regressions for corporate governance on ROE(Chap. 5) Likelihood-ratio test

LR chi2(443) = 873.39

Assumption: OLS nested in FE

Prob > chi2 = 0.0000

The Hausman test of fixed effectsmodel versus random effectsmodel Regressions for corporate governance on ROE (Chap. 5) Hausman test

chi2(14) = 119.41

Assumption: difference in coefficients not systematic

Prob > chi2 = 0.0000

LR test of pooled regression versus panel regression Regressions for corporate governance on TobinQ (Chap. 5) Likelihood-ratio test

LR chi2(443) = 2336.99

Assumption: OLS nested in FE

Prob > chi2 = 0.0000

The Hausman test of fixed effectsmodel versus random effectsmodel Regressions for corporate governance on TobinQ (Chap. 5) Hausman test

chi2(14) = 90.77

Assumption:difference in coefficients not systematic

Prob > chi2 = 0.0000

LR test of pooled regression versus panel regression Regressions for corporate governance on firms’ risk taking conduct (Chap. 6) Likelihood-ratio test

LR chi2(443) = 838.02

Assumption: OLS nested in FE

Prob > chi2 = 0.0000

The Hausman test of fixed effect model versus random effect Regressions for corporate governance on firms risk taking conduct (Chap. 6) Hausman test

chi2(14) = 369.60

Assumption:difference in coefficients not systematic

Prob > chi2 = 0.0000

Source Computed by authors

Appendix C

See Table A.2.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Z. Cheng et al., Governing Enterprises in China, Dynamics of Asian Development, https://doi.org/10.1007/978-981-16-3116-0

145

12.80

% of observation

Source Computed by authors

Cumulative%

1.7339

739

0.0296

ROE

No.observation

0.0325

ROA

Tobin Q

0–0.05

Board independence

13.25

0.45

26

1.7684

0.0523

0.0386

0.05–0.1

14.40

1.14

66

1.4530

0.0361

0.0220

0.1–0.15

16.86

2.46

142

1.4376

0.0200

0.0202

0.15–0.2

Table A.2 Range of board independence and ROA, ROE and TobinQ

20.17

3.31

191 25.57

5.41

312

1.4388

−0.1265

1.3845

0.0014

−0.0921

0.25–0.3

0.0180

0.2–0.25

68.47

42.90

2476

2.3415

−0.0135

0.0249

0.3–0.35

84.60

16.13

931

1.6198

−0.0356

0.0261

0.35–0.4

96.53

11.94

689

2.8348

−0.0903

0.0153

0.4–0.45

97.19

0.66

38

1.6686

0.0707

0.0558

0.45–0.5

100

2.81

162

2.1761

−0.0616

0.0200

0.5