Understanding Corporate Risk: A Study of Risk Measurement, Disclosure and Governance [1st ed.] 978-981-13-8140-9;978-981-13-8141-6

Table of contents :
Front Matter ....Pages i-xxx
Introduction to the Study (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 1-8
Research Methodology (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 9-20
Normative Framework for Risk Index and Its Empirical Analysis (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 21-63
Normative Framework for Risk Disclosure Index and Its Empirical Analysis (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 65-167
Normative Framework for Risk Governance Index and Its Empirical Analysis (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 169-203
Examining Risk–Return Relationship (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 205-221
Moderating Role of Governance in Risk-Disclosure Relationship (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 223-249
Practitioners’ Perspective on Risk (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 251-275
Case Studies (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 277-308
Summary and Conclusion (M. V. Shivaani, P. K. Jain, Surendra S. Yadav)....Pages 309-318
Back Matter ....Pages 319-334

Citation preview

India Studies in Business and Economics

M. V. Shivaani P. K. Jain Surendra S. Yadav

Understanding Corporate Risk A Study of Risk Measurement, Disclosure and Governance

India Studies in Business and Economics

The Indian economy is considered to be one of the fastest growing economies of the world with India amongst the most important G-20 economies. Ever since the Indian economy made its presence felt on the global platform, the research community is now even more interested in studying and analyzing what India has to offer. This series aims to bring forth the latest studies and research about India from the areas of economics, business, and management science. The titles featured in this series will present rigorous empirical research, often accompanied by policy recommendations, evoke and evaluate various aspects of the economy and the business and management landscape in India, with a special focus on India’s relationship with the world in terms of business and trade.

More information about this series at http://www.springer.com/series/11234

M. V. Shivaani P. K. Jain Surendra S. Yadav •

•

Understanding Corporate Risk A Study of Risk Measurement, Disclosure and Governance

123

M. V. Shivaani Indian Institute of Management (IIM), VNIT Campus Nagpur, India

P. K. Jain Department of Management Studies Indian Institute of Technology Delhi New Delhi, India

Surendra S. Yadav Department of Management Studies Indian Institute of Technology Delhi New Delhi, India

ISSN 2198-0012 ISSN 2198-0020 (electronic) India Studies in Business and Economics ISBN 978-981-13-8140-9 ISBN 978-981-13-8141-6 (eBook) https://doi.org/10.1007/978-981-13-8141-6 © Springer Nature Singapore Pte Ltd. 2019 This work is subject to copyright. All rights are reserved 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

In the loving memory of Late Dr. Manoj Kumar

Preface

Risk has been an intriguing phenomenon for both managers and academicians. Much of the ambiguity in the domain may be ascribed to different connotations of risk in different contexts. In view of the debacles such as Enron and Satyam (some years back), and global financial crisis, the concept of corporate risk and risk management has been gaining attention not only from investors and corporates but also from policy-makers. The present monograph presents a research work relating to risk measurement, disclosure and governance. For the purpose, well-established research techniques, namely expert opinion, diff-GMM regression, nonparametric tests like Kruskal– Wallis test and Mann–Whitney U test, have been used. Besides these techniques, a survey has been administered among top finance personnel of sample companies. Further, the results have been corroborated by the use of a practice-oriented comparative case study. In operational terms, an attempt has been made to develop a risk index (based on nine major risks, namely market risk, accounting risk, competition risk, contingency risk, credit risk, foreign exchange risk, liquidity risk, operating risk and solvency risk). In addition, the relationship between risk levels and accounting returns has also been examined. Further, a risk disclosure index (based on 69 risk-related items and covering three semantic attributes, namely nature of disclosure (qualitative/quantitative), time orientation (backward-looking/forward-looking/ both), tone of disclosure (neutral/positive/negative/both positive and negative)); and a risk governance index (based on board size, proportion of women directors, proportion of non-executive directors, proportion of independent directors, CEO duality, executive/non-executive status of chairman, chief risk officer (CRO), whistle-blower policy, risk management committee, compulsory committees and voluntary committees) have also been developed. Moreover, the relationships among the three indices have also been examined. For the purpose, difference generalised method of moments (diff-GMM) regression has been used. The notable findings of the secondary data are that the sample Indian firms, on an average, have ‘moderate risk’ levels. These findings are supplemented with the evidence of satisfactory return on assets and return on equity for the sample vii

viii

Preface

companies. Further, as the risk index is a normative framework, recommendatory in nature, the evidence of negative risk–return relationship (i.e., higher the risk, lower the returns) provides support for the methodology used in the construction of the risk index. On the one hand, the sample companies exhibit extremely low scores on risk disclosure index, indicating a lack of transparency regarding the risks and their impacts. On the other hand, the sample companies have sound risk governance structures. But, these could be further strengthened with collaborative efforts of corporates and regulatory bodies. Though no significant relationship could be found between risk levels and quality of governance structure, risk disclosures appear to be positively influenced by both the risk levels and the quality of governance structure. It is noteworthy that risk governance index acts as a moderating variable, influencing the relationship between risk disclosure index and risk index. The survey findings indicate that the sample firms consider risk as something with a positive as well as negative potential. Most firms seem to undertake risk management activities in order to protect firm value. Further, the majority of companies have an internal manual for risk management policies and carry a continuous review of techniques used in risk management. It may be noted that companies that seem to be in favour of voluntary risk reporting (in annual reports) tend to have higher risk levels. In addition, the chief finance officer emerges as the most widely accepted authority on risk. It may further be noted that the results of secondary data analysis and survey analysis have been corroborated by the results of a case study (two cases) analysis. The findings of the cases studied provide credence to the methodology used in the construction of risk index, disclosure index and governance index. Based on the findings of the research, the following recommendations have been made: (i) policy-makers should take initiatives to prescribe a risk index type normative framework for non-financial companies; to help them maintain theoretically tenable risk levels; to enhance transparency and improve communication between companies and stakeholders; and to make risk disclosures mandatory; non-compliance with certain mandatory provisions, such as appointment of woman director, needs to be dealt with swiftly and strictly. (ii) Managers should avoid excessive risk taking as higher risk does not always translate into higher return; they should focus on increasing transparency, improving communication with stakeholders and reducing information asymmetry to garner investor confidence. Nagpur, India New Delhi, India New Delhi, India

M. V. Shivaani P. K. Jain Surendra S. Yadav

Acknowledgements

First of all, we would like to thank the Almighty for His blessings. We are grateful to Prof. V. Ramgopal Rao, Director, IIT Delhi, and Prof. L. S. Murthy, Director, IIM Nagpur, for their kind support and cooperation. We express our sincere thanks to Prof. Kanika T. Bhal, Professor, Department of Management Studies (DMS), IIT Delhi; Prof. M. P. Gupta, Head, DMS, IIT Delhi; Prof. Shveta Singh, DMS, IIT Delhi; and Prof. Smita Kashiramka for their encouragement to pursue this task. We are also thankful to all the colleagues in DMS, IIT Delhi, and in IIM Nagpur for their constant support and good wishes. We have a word of appreciation for the excellent support from Sagarika Ghosh and Nupoor Singh and their team members of Springer for the speedy and excellent publication of the book. Professor P. K. Jain acknowledges his wife Uma for her patience, understanding, cooperation and encouragement. Dr. M. V. Shivaani believes that this book would not have been possible without the unrelenting support, guidance, insights and motivation of her respected co-authors and gurus Prof. P. K. Jain and Prof. Surendra S. Yadav. She would also like to express her sincere gratitude to Prof. C. P. Gupta (Department of Financial Studies, Delhi University) for his constant guidance and motivation. In addition, she expresses her heartfelt gratitude to her mother Ms. Vasudha Gupta and grandmother Ms. Dropati Aggarwal, without whose unwavering support, encouragement, endurance, perseverance and sacrifices, and this academic journey would not have been possible. M. V. Shivaani P. K. Jain Surendra S. Yadav

ix

Contents

1

Introduction to the Study . . . . . . . . . . 1.1 Introduction . . . . . . . . . . . . . . . . 1.2 Regulatory Background . . . . . . . . 1.3 Description of the Problem . . . . . 1.4 Conceptual Framework . . . . . . . . 1.5 Objectives and Significance . . . . . 1.5.1 Objectives of the Study . 1.5.2 Significance of the Study 1.6 Scope of the Study . . . . . . . . . . . 1.7 Methodology of the Study . . . . . . 1.8 Organisation of the Study . . . . . . 1.9 Concluding Observations . . . . . . . References . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

1 1 2 3 4 5 5 6 6 7 7 8 8

2

Research Methodology . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . 2.2 Research Objectives and Hypotheses 2.2.1 Research Objectives . . . . . . 2.2.2 Hypotheses of the Study . . . 2.3 Data . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Secondary Data . . . . . . . . . 2.3.2 Primary Data . . . . . . . . . . . 2.4 Research Methodology and Scope . . 2.4.1 Research Methodology . . . . 2.4.2 Scope of the Study . . . . . . . 2.4.3 Use of Statistical Software . 2.5 Concluding Observations . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

9 9 9 9 12 13 13 14 14 14 19 19 19 20

. . . . . . . . . . . . .

xi

xii

3

4

5

6

Contents

Normative Framework for Risk Index and Its Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Sample Selection . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Empirical Evidence . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Aggregative Analysis . . . . . . . . . . . . . . . . . 3.4.2 Dis-aggregative Analysis . . . . . . . . . . . . . . 3.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Limitations of the Study . . . . . . . . . . . . . . . . . . . . . 3.7 Concluding Observations . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

21 21 22 33 33 33 40 60 61 61 62

....... .......

65 65

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

Normative Framework for Risk Disclosure Index and Its Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Methodology for Construction of Risk Disclosure Index (RDI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Sample Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Empirical Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1 Aggregative Analysis . . . . . . . . . . . . . . . . . . 4.4.2 Dis-aggregative Analysis . . . . . . . . . . . . . . . 4.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Concluding Observations . . . . . . . . . . . . . . . . . . . . . . Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

66 73 73 73 80 142 143 144 166

Normative Framework for Risk Governance Index and Its Empirical Analysis . . . . . . . . . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Empirical Evidence . . . . . . . . . . . . . . . . . . . . . 5.4.1 Aggregative Analysis . . . . . . . . . . . . . 5.4.2 Dis-aggregative Analysis . . . . . . . . . . 5.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Concluding Observations . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

169 169 170 178 178 178 185 199 200 201

Examining Risk–Return Relationship . . . . . . . . 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 6.2 Background and Hypotheses Development . 6.3 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Methodology . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

205 205 205 207 207

. . . . .

. . . . .

. . . . .

Contents

6.5

Empirical Evidence . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1 Descriptive Analysis . . . . . . . . . . . . . . . . . . 6.5.2 Relationship Between Risk Index and ROA (Model 1) . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.3 Relationship Between Risk Index and ROA (Model 2) . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.4 Relationship Between Risk Index and ROE (Model 3) . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.5 Relationship of Risk Index ROE (Model 4) . 6.6 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7 Concluding Observations . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

xiii

. . . . . . . . 210 . . . . . . . . 210 . . . . . . . . 213 . . . . . . . . 215 . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

Moderating Role of Governance in Risk-Disclosure Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Background and Hypothesis Development . . . . . . . . . . . . . . 7.2.1 Risk Index (RI) and Risk Governance Index (RGI) . 7.2.2 Risk Disclosure Index (RDI) and Risk Index (RI) . . 7.2.3 Risk Disclosure Index (RDI) and Risk Governance Index (RGI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5 Empirical Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.1 Descriptive Analysis . . . . . . . . . . . . . . . . . . . . . . . . 7.5.2 Relationship Between Risk Index and Risk Governance Index (Model 1) . . . . . . . . . . . . . . . . . 7.5.3 Relationship Between Risk Disclosure Index and Risk Governance Index (Model 2) . . . . . . . . . . . . . . . . . 7.5.4 Relationship Between Risk Disclosure Index and Risk Index (Model 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.5 Relationship of Risk Disclosure Index with Risk Index and Risk Governance Index (Model 4) . . . . . . 7.5.6 Relationship of Risk Disclosure Index with Risk Index and Risk Governance Index While Factoring for Moderating Effects of Risk Governance Index (Model 5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.7 Further Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.7 Concluding Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

216 217 218 219 220

. . . . .

. . . . .

223 223 225 225 227

. . . . .

. . . . .

228 229 229 231 231

. . 231 . . 232 . . 234 . . 235

. . . . .

. . . . .

238 242 244 245 246

xiv

8

9

Contents

Practitioners’ Perspective on Risk . . . . . . . . . . . . . . . . . . . . . . . 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Survey Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3 Profile of Respondents . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.1 Descriptive Statistics of Respondent Companies . . . 8.3.2 Designation of Respondents in Companies . . . . . . . 8.3.3 Respondents’ Objectivity and Fairness in Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4 Analysis and Empirical Results . . . . . . . . . . . . . . . . . . . . . 8.4.1 Organisations’ Attitude Towards Risk . . . . . . . . . . 8.4.2 Risk Identification and Prioritisation . . . . . . . . . . . 8.4.3 Risk Measurement . . . . . . . . . . . . . . . . . . . . . . . . 8.4.4 Risk Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.5 Risk Communication . . . . . . . . . . . . . . . . . . . . . . 8.4.6 Risk Management Authority and Structure . . . . . . . 8.4.7 Risk and Return . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.8 Evaluation of Risk Management Practices . . . . . . . 8.4.9 Risk Prioritisation, Measurement and Management . 8.4.10 Relationship Between Primary and Secondary Data 8.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6 Concluding Observations . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

251 251 252 253 253 254

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

255 257 257 258 259 259 262 264 266 267 267 269 273 274 275

Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.1 Case Study Methodology . . . . . . . . . . . . . . . . . 9.2.2 Criterion for the Selection of Companies . . . . . . 9.3 Brief Background of Companies Selected for Case Study 9.4 Analysis and Empirical Results . . . . . . . . . . . . . . . . . . . 9.4.1 Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . 9.4.2 Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . 9.5 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.6 Concluding Observations . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10 Summary and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Major Findings of the Research . . . . . . . . . . . . . . . . . 10.2.1 Findings Related to Risk Index . . . . . . . . . . . 10.2.2 Findings Related to Risk Disclosure Index . . . 10.2.3 Findings Related to Risk Governance Index . . 10.2.4 Findings Related to Risk-Return Relationship

. . . . . . .

. . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

277 277 278 278 279 279 280 280 283 306 306 307

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

309 309 310 310 311 313 314

Contents

10.3 10.4 10.5 10.6 10.7

xv

10.2.5 Findings Related to Relationship Among Risk Index, Risk Disclosure Index and Risk Governance Index . . 10.2.6 Findings Related to Risk Management Practices . . . . Recommendations From the Study . . . . . . . . . . . . . . . . . . . . Contribution of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . Scope for Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . Concluding Observations of the Study . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

314 315 316 317 317 318 318

Annexure I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 Annexure II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Annexure III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333

About the Authors

M. V. Shivaani is an Assistant Professor in the area of accounting and finance at the Indian Institute of Management (IIM) Nagpur. She obtained her Ph.D. from the Department of Management Studies, Indian Institute of Technology (IIT) Delhi. She is a Chartered Accountant (ICAI) by profession and has presented her research at various international conferences. She was awarded best paper/presentation award at International Conference on Applied Economics and Management Sciences, Spain and International Conference on Sustainable Growth & Innovation & Revolution in the New Millennium. She has published in journals of national and international repute. She has also been a part of research team to Indian Scientific Expedition to Antarctica. P. K. Jain is an Emeritus Professor of Finance and was the Modi Chair Professor at the Department of Management Studies, Indian Institute of Technology (IIT) Delhi. He was also the Dalmia Chair Professor at the institute. Recently, he received the ‘Best Faculty Award’ at IIT Delhi. He has more than 45 years of teaching experience in subjects related to management accounting, financial management, financial analysis, cost analysis and cost control. He was a member of the visiting faculty at the AIT Bangkok; University of Paris; Howe School of Technology Management at Stevens Institute of Technology, New Jersey; and ICPE, Ljubljana. He has authored three well-known textbooks published by Tata Mcgraw Hill and more than ten research monographs (four of them with Springer), and has published over 200 research papers in respected national and international journals. Surendra S. Yadav is a Professor of Finance at the Department of Management Studies, Indian Institute of Technology (IIT) Delhi, where he was the Head of the Department of Management Studies for six years. He teaches corporate finance, international finance, international business, and security analysis and portfolio management. He has been a Visiting Professor at the University of Paris, Paris School of Management, INSEEC Paris, and the University of Tampa, USA. He has

xvii

xviii

About the Authors

published 12 books (some of them with Springer) and contributed more than 230 papers to research journals and conferences as well as over 30 papers to financial/economic newspapers. He is the editor-in-chief of the Journal of Advances in Management Research (JAMR) published by Emerald Publishing, UK.

Abbreviations

AICPA ANOVA BCBS BOD BOP CAPM CCR CEO CFO CI CICA COO COSO CRO DOL EBIT EOP ERM ESOP EWRM FASB FDI FMCG FSB GARP GIM GMM HR IASB

American Institute of Certified Public Accountants Analysis of variance Basel Committee on Banking Supervision Board of directors Beginning of period Capital asset pricing model Contingency coverage ratio Chief executive officer Chief financial officer Composite indicator Certified Internal Control Auditors Chief operating officer Committee of Sponsoring Organizations of the Treadway Commission Chief risk officer Degree of operating leverage Earnings before interest and taxes End of period Enterprise risk management Employee stock option plan Enterprise-wide risk management Financial Accounting Standards Board Foreign direct investment Fast-moving consumer goods Financial Stability Board Global Association of Risk Professionals Gompers–Ishii–Metrick Generalised method of moments Human resources International Accounting Standard Board

xix

xx

IATR ICAEW ICAI ICSI ICT IFC IFRS IRGC IRRC MD MD&A MDI MMTC NSE OECD OLS PoF PoID PoNED PSU RDI RGI RI RMC ROA ROE SEBI SEC SRM SWOT TD/E UK UNCTAD USA VaR WTD

Abbreviations

Inverse of acid-test ratio Institute of Chartered Accountants in England and Wales Institute of Chartered Accountants of India Institute of Company Secretaries of India Information and communication technology International Finance Corporation International Financial Reporting Standards International Risk Governance Council Investor Responsibility Research Center Managing director Management discussion and analysis Modified defensive interval Metals and Minerals Trading Corporation of India National Stock Exchange Organisation for Economic Co-operation and Development Ordinary least squares Proportion of females Proportion of independent directors Proportion of non-executive directors Public sector undertaking Risk disclosure index Risk governance index Risk index Risk Management Committee Return on assets Return on equity Securities and Exchange Board of India Securities Exchange Commission Spectral risk measures Strengths, weaknesses, opportunities and threats Total debt/equity United Kingdom United Nations Conference on Trade and Development United States of America Value at risk Whole-time director

List of Figures

Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.

1.1 2.1 3.1 3.2 3.3 4.1 4.2 4.3 4.4

Fig. 4.5 Fig. 4.6 Fig. 5.1 Fig. 5.2 Fig. 5.3 Fig. 5.4 Fig. Fig. Fig. Fig.

6.1 6.2 6.3 7.1

Fig. 7.2 Fig. 7.3 Fig. 7.4 Fig. 8.1 Fig. 8.2

Conceptual framework of corporate risk management . . . . . . . . Outline of research methodology . . . . . . . . . . . . . . . . . . . . . . . Scope of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Year-wise descriptive statistics of risk index. . . . . . . . . . . . . . . Pair-wise comparison of risk index industry-level . . . . . . . . . . . Scope of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Year-wise descriptive statistics of RDI . . . . . . . . . . . . . . . . . . . Pairwise comparison of RDI scores of different age group . . . . Age-wise, period-wise mean disclosure scores of sample companies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pair-wise comparison of RDI scores amongst various industry groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Industry-wise, period-wise average disclosure scores of sample companies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scope of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk governance index, 2005–2015 . . . Pair-wise comparison of risk governance index of companies based on age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pair-wise comparison of risk governance index of companies based on industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Year-wise descriptive statistics of risk index. . . . . . . . . . . . . . . Year-wise descriptive statistics of return on assets . . . . . . . . . . Year-wise descriptive statistics of return on equity . . . . . . . . . . Framework of interrelationships among various dimensions of risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exploration of relationship among RDI, RI and RGI . . . . . . . . Moderating effect of RGI on relationship between RDI and RI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Holistic view of relationship among RDI, RGI and RI . . . . . . . Industry-wise classification of respondent companies . . . . . . . . Profile of respondents of survey . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

4 14 32 34 40 72 74 77

..

78

..

80

. . . . . . . .

.. 83 . . 179 . . 180 . . 182 . . . .

. . . .

185 211 212 213

. . 224 . . 237 . . . .

. . . .

240 242 253 255 xxi

xxii

Fig. 8.3 Fig. 8.4 Fig. 8.5

Fig. 8.6

Fig. 8.7

Fig. 8.8 Fig. 8.9 Fig. 9.1 Fig. 9.2 Fig. 9.3 Fig. 9.4 Fig. 9.5 Fig. 9.6 Fig. 9.7

List of Figures

Percentage of respondents using specific techniques for risk identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percentage of companies that attempt to manage risks and reasons why companies do not manage risks . . . . . . . . . . . a Percentage of companies supporting voluntary risk disclosures in annual reports. b Percentage responses for type of voluntary information that companies are willing to share . . . . . . . . . . . . a Percentage of companies that favour voluntary risk disclosures in annual reports. b Percentage responses for various risks that a company would want to disclose . . . . . . . . . . . . . . . . . . . . . . a Percentage responses for whether the company has a top-to-bottom risk reporting system. b Percentage responses for items that are communicated under the top-to-bottom risk reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percentage of respondents preferring different measures of firm performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percentage of respondents using different indicators for the evaluation of success of risk management plans . . . . . . Comparison of risk index of Cyient Ltd., and MMTC Ltd. . . . Comparison of risk disclosure index of Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of risk governance index of Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of return on assets of Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of return on equity of Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of equity returns of Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of profitability of Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . 258 . . 261

. . 263

. . 264

. . 265 . . 267 . . 268 . . 280 . . 281 . . 281 . . 298 . . 299 . . 299 . . 300

List of Tables

Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 3.7 Table 3.8 Table 3.9 Table 3.10

Table 3.11

Table 3.12 Table 3.13 Table 4.1 Table 4.2 Table 4.3

Descriptive statistics of year-wise risk index on full sample, 2005–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk index, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Mann Whitney U test for difference of mean Risk index (pre-recession, post-recession) . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk index, age-wise (2005–2015) . . Kruskal Wallis test for mean difference in risk index among different age groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk index, age-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Mann-Whitney U test for difference of mean risk index (pre-recession, post-recession) for each age category . . . . . . Descriptive statistics of risk index, industry-wise (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk index, industry-wise, phase-wise (pre-recession, (2005–2008); post-recession (2008–2015)) . . Mann Whitney U test for difference of mean risk index, industry-wise, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of various risks, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of various risks, age-wise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of various risks, industry-wise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of items/themes constituting risk disclosure index (RDI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methodology for construction of scale . . . . . . . . . . . . . . . . . Descriptive statistics of RDI, 2005–2015 . . . . . . . . . . . . . . .

..

34

..

35

.. ..

36 36

..

37

..

38

..

38

..

39

..

41

..

42

..

43

..

46

..

50

.. .. ..

67 69 74 xxiii

xxiv

Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 4.9 Table 4.10 Table 4.11 Table 4.12

Table 4.13

Table 4.14

Table 4.15

Table 4.16

Table 4.17

Table 4.18 Table 4.19 Table 4.20 Table 4.21

Table 4.22

List of Tables

Descriptive statistics of RDI, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Mann–Whitney U test for difference of mean RDI (pre-recession, post-recession) . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of RDI, age-wise (2005–2015) . . . . . . Kruskal-Wallis test for difference of mean RDI amongst age categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of RDI, age-wise (pre-recession (2005–2008); post-recession) . . . . . . . . . . . . . . . . . . . . . . . . Mann–Whitney U test for difference of mean RDI (pre-recession, post-recession) for each age category . . . . . . Descriptive statistics of RDI, industry-wise (2005–2015) . . . Descriptive statistics of RDI, industry-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Mann–Whitney U test for difference of mean RDI (pre-recession, post-recession) for each industry classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of top 10% of items, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of bottom10% of items, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of top 10% and bottom 10% of items for young companies, (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Summary of frequency distribution of top 10% and bottom 10% of items for middle-aged companies, (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Summary of frequency distribution of top 10% and bottom 10% of items for old and established companies, (pre-recession (2005–2008); post-recession (2008–2015)) . . Summary of frequency distribution of the top 10% of items, industry-wise, for the period under study (2005–2015) . . . . Summary of the frequency distribution of the top 10% of items, industry-wise, pre-recession (2005–2008) . . . . . . . Summary of the frequency distribution of top 10% of items, industry-wise, post-recession (2008–2015) . . . . . . . . . . . . . . Description of top and bottom 10% of ‘most comprehensively/richly disclosed’ risk items, full sample (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Description of top and bottom 10% of ‘most comprehensively/richly disclosed’ items, pre-recession (2005–2008) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

..

75

.. ..

76 76

..

77

..

78

.. ..

78 79

..

81

..

82

..

84

..

85

..

87

..

89

..

91

..

94

..

98

. . 101

. . 105

. . 106

List of Tables

Table 4.23

Table 4.24

Table 4.25 Table 4.26 Table 4.27

Table 4.28 Table 4.29 Table 4.30

Table 4.31 Table 4.32

Table 4.33

Table 4.34

Table 4.35 Table 4.36

Table 4.37

Table 4.38

Table 4.39

xxv

Description of top and bottom 10% of ‘most comprehensively/richly disclosed’ items, post-recession (2009–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Description of top 10% of ‘most comprehensively/richly disclosed’ items, age-wise, for the period under study (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Description of top 10% of ‘most comprehensively/richly disclosed’ items, age-wise pre-recession (2005–2008) . . . . . Description of top 10% of ‘most comprehensively/richly disclosed’ items, age-wise, post-recession (2008–2015) . . . . Description of bottom 10% of ‘most comprehensively/richly disclosed’ items, age-wise, for the period under study (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Description of bottom 10% of ‘most comprehensively/richly disclosed’ items, age-wise pre-recession (2005–2008) . . . . . Description of bottom 10% of ‘most comprehensively/richly disclosed’ items, age-wise, post-recession (2008–2015) . . . . Description of top 10% of ‘most comprehensively/richly disclosed’ items, industry-wise, for the period under study (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Description of top 10% of ‘most comprehensively/richly disclosed’ items, industry-wise pre-recession (2005–2008) . . Description of top 10% of ‘most comprehensively (rich) disclosed’ items, industry-wise, post-recession (2008–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frequency distribution of items pertaining to risk management (scored on binary scale), full sample (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frequency distribution of items pertaining to risk management (scored on binary scale), age-wise, (pre-recession (2005–2008); post-recession (2008–2015)) . . Examples of coding of risk disclosures in annual reports . . . Summary of frequency distribution of top 10% of ‘least comprehensively (rich) disclosed’ items, industry-wise, for the period under study (2005–2015) . . . . . . . . . . . . . . . . Summary of the frequency distribution of top 10% of ‘least comprehensively (rich) disclosed’ items, industry-wise pre-recession (2005–2008) . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of the frequency distribution of top 10% of ‘least comprehensively (rich) disclosed’ items, industry-wise, post-recession (2008–2015) . . . . . . . . . . . . . . Frequency distribution of items pertaining to risk management (scored on binary scale), industry-wise, (pre-recession (2005–2008); post-recession (2008–2015)) . .

. . 107

. . 109 . . 112 . . 114

. . 117 . . 120 . . 123

. . 127 . . 131

. . 135

. . 139

. . 140 . . 144

. . 145

. . 150

. . 155

. . 159

xxvi

List of Tables

Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9

Table 5.10

Table 5.11

Table 5.12 Table 5.13 Table Table Table Table

6.1 6.2 6.3 6.4

Table 6.5 Table 6.6 Table 6.7 Table 7.1 Table 7.2

Descriptive statistics of year-wise risk governance index on full sample, 2005–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk governance index, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . Mann–Whitney U test for difference of mean risk governance index (pre-recession, post-recession) . . . . . . . . . . . . . . . . . . Descriptive statistics of risk governance index, age-wise (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kruskal–Wallis test for mean difference in risk governance index among different age groups . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk governance index, age-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . Mann–Whitney U test for difference of mean risk governance index (pre-recession, post-recession) for each age category . Descriptive statistics of risk governance index, industry-wise (2005–2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk governance index, industry-wise, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mann Whitney U test for difference of mean risk governance index, industry-wise, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . Summary of frequency distribution of risk governance characteristics, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) . . . . . . . . . . . . . . . . . . . . . . . . . Summary of frequency distribution of risk governance characteristics, age-wise . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of the frequency distribution of risk governance characteristics, industry-wise . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of risk index . . . . . . . . . . . . . . . . . . . . Descriptive statistics of return on assets . . . . . . . . . . . . . . . . Descriptive statistics of return on equity . . . . . . . . . . . . . . . Results of (Arellano-Bond) GMM estimation of ROA on first lag of ROA, RI and control variables . . . . . . . . . . . Results of (Arellano-Bond) GMM estimation of ROA on first lag of ROA, RI, RGI and control variables . . . . . . . Results of (Arellano-Bond) GMM estimation of ROE on first lag of ROE, RI and control variables . . . . . . . . . . . . Results of (Arellano-Bond) GMM estimation of ROE on first lag of ROE, RI, RGI and control variables . . . . . . . Descriptive statistics of relevant variables for the period 2005–2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results of (Arellano–Bond) GMM estimation of RI on the first lag of RI, RGI and control variables . . . . . . . . .

. . 179 . . 180 . . 181 . . 181 . . 182 . . 183 . . 183 . . 184

. . 186

. . 187

. . 188 . . 191 . . . .

. . . .

193 210 212 213

. . 214 . . 216 . . 217 . . 218 . . 231 . . 232

List of Tables

Table 7.3 Table 7.4 Table 7.5 Table 7.6

Table Table Table Table Table Table Table Table Table

7.7 7.8 7.9 8.1 8.2 8.3 8.4 8.5 8.6

Table 8.7 Table 8.8 Table 8.9 Table 8.10 Table 8.11 Table 8.12

Table 8.13 Table 8.14 Table 8.15

Table 8.16 Table 9.1 Table 9.2

xxvii

Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RGI and control variables . . . . . . . . Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RI and control variables . . . . . . . . . Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RI and RGI and control variables . . Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RI, RGI and (RI * RGI) and control variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross tabulation between RI and RGI . . . . . . . . . . . . . . . . . Cross tabulation between RDI and RI . . . . . . . . . . . . . . . . . Cross tabulation between RDI and RGI . . . . . . . . . . . . . . . . Descriptive statistics of respondent companies . . . . . . . . . . . Examination of personal bias in questionnaire responses . . . Examination of personal bias in questionnaire responses . . . Summary of practitioners’ opinion on risk . . . . . . . . . . . . . . Management’s motives for pursuing risk management . . . . . Summary of importance of various factors for risk prioritisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of practitioners’ preferences for various risk measurement techniques and tools . . . . . . . . . . . . . . . . . . . . Summary of risk handling practices prevalent in Indian corporates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of risk communication practices prevalent in Indian corporates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Practitioners’ views on importance of recipients of risk information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of practitioners’ views on risk authority and risk structures in Indian corporates . . . . . . . . . . . . . . . . Ranking of various risks based on their importance, and percentage of companies measuring and managing those risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of relationship between risk disclosure index and practitioners’ responses . . . . . . . . . . . . . . . . . . . . . . . . . Summary of relationship between risk index and organisation’s attitude towards risk . . . . . . . . . . . . . . . . . . . Summary of relationship between risk disclosure index and designation of person drafting the MD&A section of the annual report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of relationship between risk index and practitioners’ responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics of RI, RDI and RGI for Cyient Ltd., and MMTC Ltd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mann–Whitney U test for difference of risk index . . . . . . . .

. . 233 . . 234 . . 236

. . . . . . . . .

. . . . . . . . .

239 243 243 244 254 256 256 257 257

. . 259 . . 260 . . 261 . . 262 . . 266 . . 266

. . 268 . . 270 . . 271

. . 271 . . 272 . . 282 . . 284

xxviii

Table 9.3 Table 9.4 Table 9.5 Table 9.6 Table 9.7 Table 9.8 Table 9.9 Table 9.10 Table 9.11 Table 9.12 Table 9.13 Table 9.14 Table 9.15 Table 9.16 Table 9.17

List of Tables

Descriptive statistics related to various constituents of risk index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary results of Mann–Whitney U test for sub-hypotheses related to risk index . . . . . . . . . . . . . . . . . . . Mann–Whitney U test for difference in risk disclosure index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics related to components of risk disclosure index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary results of Mann–Whitney U test related to various aspects of risk disclosure index . . . . . . . . . . . . . . . . . . . . . . Mann–Whitney U test for difference in risk governance index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive statistics related to various components of risk governance index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary results of Mann–Whitney U test for sub-hypotheses related to risk governance index . . . . . . . . . Descriptive statistics related to various performance indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary results for Mann–Whitney U test related to various performance measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary results of hypothesis testing . . . . . . . . . . . . . . . . . Methodology for categorisation of RDI, RGI, RI from ‘very low’ to ‘very high’ . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross tabulation based on risk index and risk governance index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross tabulation based on risk index and risk disclosure index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross tabulation based on risk governance index and risk disclosure index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . 285 . . 286 . . 290 . . 291 . . 292 . . 295 . . 296 . . 297 . . 301 . . 302 . . 303 . . 304 . . 305 . . 305 . . 305

List of Exhibits

Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit Exhibit

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 5.1 5.2

Exhibit 5.3 Exhibit 5.4 Exhibit 5.5 Exhibit 5.6 Exhibit 5.7 Exhibit 5.8 Exhibit 5.9 Exhibit 5.10 Exhibit 5.11 Exhibit 5.12

Scoring in relation to market risk . . . . . . . . . . . . . . . . . . . Scoring in relation to accounting risk . . . . . . . . . . . . . . . . Scoring in relation to competition risk . . . . . . . . . . . . . . . . Scoring in relation to contingency risk . . . . . . . . . . . . . . . Scoring in relation to credit risk . . . . . . . . . . . . . . . . . . . . Scoring in relation to exchange rate risk . . . . . . . . . . . . . . Scoring in relation to liquidity risk 1 . . . . . . . . . . . . . . . . . Scoring in relation liquidity risk 2 . . . . . . . . . . . . . . . . . . . Scoring in relation to operating risk . . . . . . . . . . . . . . . . . . Scoring in relation to solvency risk 1 . . . . . . . . . . . . . . . . Scoring in relation to solvency risk 2 . . . . . . . . . . . . . . . . Scoring for number of board of directors . . . . . . . . . . . . . . Scoring in relation to the number of women directors on board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scoring in relation to the proportion of non-executive directors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scoring in relation to executive–non-executive status of chairman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scoring in relation to the proportion of independent directors, with executive chairman . . . . . . . . . . . . . . . . . . . Scoring in relation to proportion of independent directors, with non-executive chairman . . . . . . . . . . . . . . . . . . . . . . . Scoring in context of CEO duality . . . . . . . . . . . . . . . . . . . Scoring in the context of the appointment of a CRO . . . . . Scoring in context of the existence of a whistle-blower policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scoring in context of the existence of risk management committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scoring in the context of compulsory committees . . . . . . . Scoring in context of number of voluntary committees constituted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .

. 24 . 25 . 26 . 27 . 27 . 28 . 29 . 30 . 31 . 31 . 32 . 172

. . 172 . . 173 . . 174 . . 174 . . 175 . . 175 . . 175 . . 176 . . 176 . . 177 . . 177 xxix

xxx

Exhibit 7.1 Exhibit 7.2

List of Exhibits

Methodology for computation mean RDI corresponding to various levels of RI and RGI . . . . . . . . . . . . . . . . . . . . . . . . . 237 Methodology for categorisation of RDI, RGI, RI from ‘very low’ to ‘very high’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242

Chapter 1

Introduction to the Study

If you don’t invest in risk management it doesn’t matter what business you are in it’s a risky business. —Gary Cohn

Abstract This chapter aims to provide a brief outline of the research presented in this monograph. It provides a brief background of corporate risk management process and a conceptual framework of risk management. It also highlights the regulatory stance on risk measurement, disclosure and governance. Additionally, the chapter also describes the objectives, scope, need, significance of the study, research methodology (in brief) and the chapter plan of the research.










Keywords Corporate risk Risk management Risk measures Risk disclosure Risk governance

1.1




Introduction

The word risk has different connotations from different perspectives. Some perceive risk in a negative perspective and define it as ‘the potential for unwanted negative consequences of an event or activity’. There are others, who consider risk from a neutral viewpoint and accordingly define it as an ‘unintended or unexpected outcome’. Merna and Al-Thani (2008) define risk management as any set of actions taken by individuals or corporations in an effort to alter the risk arising from their business. Dione (2013) defines corporate risk management as a set of financial or operational activities that maximise the value of the company. He adds that diversification and hedging are the core risk management activities. Firms undertake risk management to minimise the cost of financial distress, risk premium to partners, expected taxes, etc. In other words, risk management is a part of the overall management of an organisation and not a separate activity. It is pervasive across planning, prediction and prevention. In fact, reactive risk management symbolises a lack of vision and mission in an organisation. © Springer Nature Singapore Pte Ltd. 2019 M. V. Shivaani et al., Understanding Corporate Risk, India Studies in Business and Economics, https://doi.org/10.1007/978-981-13-8141-6_1

1

2

1 Introduction to the study

In view of the all-pervasive nature of risk, risk management has been studied from various perspectives including financial economics and strategic management and yet, there is no generally accepted definition of risk. Even the risk–return relationship continues to be a puzzle. Certain studies document a positive risk–return relationship; others provide evidence of negative risk–return relationship. Given the complexity and ambiguity involved in the domain, researchers have been advocating the benefits of increased risk disclosures in annual reports; on the disclosure aspect, there is an ongoing debate as to whether the disclosures should be voluntary or mandatory. Above all, research on the role of the board of directors in the area of risk management has been largely neglected.

1.2

Regulatory Background

Given the importance of risk management as a significant aspect of corporate management, enterprise risk management (ERM) has started to gain currency. In 1992, Sarbanes–Oxley Act (SOX) made Committee of Sponsoring Organizations of the Treadway Commission (COSO) to introduce 17 principles (at a conceptual level) encompassing fields like internal environment, risk assessment, control activities, information and communication, and monitoring. In 2012, COSO has issued a report; the report has underlined the importance of ‘sweet spot’ or optimal risk-taking zone. It emphasises the requirement of a risk assessment process that is practical, sustainable and easy to understand. Accordingly, assessment scales have been developed. These scales on a continuum of 1–5 have been suggested for impact, likelihood, vulnerability and speed of onset (velocity). Further, scenario analysis for various factors has also been attempted; presentation of analysis in the form of risk interaction maps is also suggested in the report. COSO has also provided a framework (primarily designed) to assist management in designing, implementing and maintaining internal control; for the purpose, a five-step transition has been suggested. The development of awareness enterprise and alignment is the first step, followed by conducting a preliminary impact analysis. The next step is to facilitate board awareness, training and comprehensive assessment. The fourth step is to develop and execute the COSO transition plan for SOX compliance. Last but not least is the implementation of a continuous improvement program. In Indian context, with amendments in Clause 49 of Listing agreement1 and enactment of Companies Act 2013, substantial reforms have been brought about in the area of risk management. Clause 49 has made mandatory the constitution of risk

1

An agreement between a company that desires to be listed or is listed on a stock exchange and the concerned stock exchange.

1.2 Regulatory Background

3

management committee. It has also laid down guidelines relating to risk assessment and minimisation. It has also directed companies to formulate a whistle-blower policy, which is supposed to serve as a control on operational risks. Further, as per Section 134(1) of Companies Act 2013, the board of directors are required to present a report (in every annual general meeting), and such a report should contain details of risk management policy implemented by the Board. The report should also contain the details of the risks, if any, that may threaten the existence of the company. Further, in terms of independent directors, their involvement in risk management process has been explicitly highlighted.

1.3

Description of the Problem

In the current turbulent times, the concept of risk, in general, and the process of risk management, in particular, is gaining momentum. Yet, there is no generally acceptable measure of risk or a Basel-like framework for non-financial companies. Evidently, it is imperative to have such a framework in place as it may be a guide to companies in their quest towards effective and efficient risk management. It may help companies in timely detection of signs of distress, areas requiring attention, and formulation of strategic policies. Not only in terms of risk measurement, even in terms of risk disclosures, there appears to be a need for a summary measure that captures not only the quantity of risk items disclosed but also the ‘richness’ of risk items disclosed in terms of semantic properties. With increasing focus on information asymmetry and the call for companies to come across as transparent and accountable organisations, development of such an analytical tool seems pertinent. It may be noted that semantic properties of disclosures have been examined in an international context, but no such study has been observed in the Indian context. It may further be noted, that risk measurement and disclosure need to be viewed in the context of governance structures in vogue in the organisation. The literature is rife with studies on various dimensions of corporate governance but, a relevant dimension, risk governance is still an unexplored domain. This is despite the fact that 3 years have passed since International Finance Corporation (IFC) (2012) first came up with a definition for risk governance and recommended that companies should start focussing on this extremely critical aspect as well. In view of the foregoing, the present study attempts to put forth three normative frameworks (indices): first, pertaining to risk (exposure) measurement, second, dealing with the quality of risk disclosures and third, relating to quality of governance structure. Further, an attempt has been made to examine the interrelationships (if any) that exist amongst these indices. Also, primary data has been used to corroborate the results derived from these indices.

4

1.4

1 Introduction to the study

Conceptual Framework

In the following Fig. 1.1, an attempt has been made to present a conceptual framework for corporate risk management. Every business organisation comes into existence with three progressive objectives: survival, profit and growth. Survival is the will and anxiety to perpetuate into the future as long as it is possible. It is a basic and implicit objective of most organisations. It is generally asserted that private enterprises are primarily motivated by the objective of profit. All other objectives are subservient to the profit motive. This is a promising and widely used objective which is equated with dynamism, vigour, promise and success. Enterprise growth may assume one or more of the forms like increase in assets, manufacturing facilities, increase in sales volume in existing products or through new products, improvement in profits and market share, increase in manpower employment, acquisition of other enterprises and so on. The pursuit of these objectives may take the organisation on relatively unknown and risky paths, full of promises and pitfalls (Institute of Chartered Accountants of India (ICAI) 2016). The risks that may affect the return of a firm can be categorised as market risks, operational risks and credit risk. These risks may arise from within the organisation

Fig. 1.1 Conceptual framework of corporate risk management

1.4 Conceptual Framework

5

due to poor processes and practices or from outside the organisation due to firms’ exposure to macroeconomic variables. Since a company is an artificial person, it cannot function on its own. The direction and control of a company is in the hands of its directors. They are the lifeline of the company. As a consequence, they are responsible for its financial and operational health. The system of rules, laws and factors that control the operations of a company can be defined as corporate governance (Gillan and Starks 1998). At this point, it seems reasonable to state that the responsibility of protecting the company from internal and external risks vests with the Board. In other words, corporate governance and risk management are intimately connected. The process of risk management can be explained through five steps: identifying and prioritising risks, quantifying risks, managing risks, reporting risks and reviewing risks. The first step of the process is of utmost importance as the determination of companies’ strengths, weakness, opportunities and threats depends on it. Once the risks have been identified, it is imperative to assess their probability, impact and severity. Then depending on their severity, steps are needed to mitigate or reduce their impact. In recent times, strategic policymaking and instruments like derivatives have emerged to be effective tools of managing risks. The management of these risks is a prerequisite for the survival and growth of the firm. If not managed, they may have dire consequences for various stakeholders. This necessitates the need to communicate various risk dimensions to different stakeholders (government, employees, investors, etc.). Also, information asymmetry between management and shareholders may have undesirable implications. Of late, regulatory bodies have also turned their attention to ‘disclosure and reporting’ practices followed by firms, in general, and listed firms, in particular. It appears reasonable to believe that transparent and fair disclosures facilitate informed decision making among stakeholders. To keep pace with this ever-changing world, periodic and timely review of the process is now reckoned as a virtual necessity. The first two steps of the process are intended to be captured by constructing a ‘risk index’; the management and reporting part is proposed to be studied by adopting a risk-disclosure index and the whole risk management process is to be analysed in conjunction with the risk management authority, i.e. risk governance structure of the organisation.

1.5 1.5.1

Objectives and Significance Objectives of the Study

The study attempts to develop meaningful and easily comprehensible measures covering various dimensions of the risk management process. The study also seeks to gain insight into risk management practices prevalent in Indian companies. The

6

1 Introduction to the study

study consists of two parts. The objective of the first part is to develop three normative frameworks. The first framework is in the form of a risk index to be used as a comprehensive measure of major corporate risks, the second is in the form of a risk-disclosure index to be used as a measure of quality of risk disclosures in annual reports, and the third is a risk governance index to be used as a measure of quality of governance structure of a company. The objective of the second part is to explore and examine the relationship between and among the three indices. In addition, the relationship between the risk index and accounting return has also been explored. Further, to gain an insight into practitioners’ perceptions of risk management and risk management practices, a managerial survey has been carried out. To corroborate the results of primary and secondary data, a comparative case analysis has been conducted.

1.5.2

Significance of the Study

The study is expected to be of immense utility to academicians as well as practitioners. The study puts forth a Basel-like normative framework for risk measurement of non-financial firms. The risk index based on the framework is expected to be of use in risk measurement to companies and investors, and in risk monitoring to auditors and regulators. Further, the developed risk measure is expected to provide a novel dimension to traditional risk measures used in strategic management literature. The study also presents one of its kind summary measure that captures not only the quantity of risk disclosure but also the quality of disclosures in terms of semantic attributes. The risk-disclosure index is expected to facilitate regulators and academicians in gauging the level of transparency and information asymmetry prevalent in the Indian context. Also, it is perhaps the first study that attempts to examine the risk-disclosure practices of Indian firms. In addition, the study focuses on an important yet neglected aspect of risk management, namely, risk governance. The normative framework of risk governance put forth in the study is expected to help regulators in framing relevant rules and regulations. Moreover, as the results of secondary data have been corroborated by the findings of primary data, the frameworks put forth in the study are expected to make significant contribution to the literature on the subject.

1.6

Scope of the Study

1. The study is limited to non-financial companies (429) that constituted the NIFTY 500 index as on 31.3.2014. 2. The study covers a 10-year period from 1 April 2005 to 31 March 2015. To study the impact of the recession, the period from 1 April 2005 to 31 March

1.6 Scope of the Study

7

2008 has been considered as the pre-recession phase and the period from 1 April 2008 to 31 March 2015 has been considered as the post-recession phase. 3. The study also includes an in-depth analysis of all three indices and all their constituents on the basis of firms’ age, industry-classification and period of recession. 4. The study also captures the responses to a management survey, carried out on sample companies and includes an in-depth analysis of two sample companies (one best performer and other the worst performer, in terms of three indices developed in the study).

1.7

Methodology of the Study

The three indices put forth in the study are based on expert opinion, using Delphi technique. Additionally, risk-disclosure index also makes use of content analysis and textual analysis. The indices so developed are then empirically tested on data of Indian companies. For each of the indices, an age-wise, phase-wise and industry-wise analysis is carried out. Due to non-normal nature of data, the analysis is primarily based on non-parametric tests such as Mann–Whitney U test and Kruskal–Wallis test. Further, an attempt is made to examine the relationship between the risk index and accounting returns, and between and amongst the three indices. Owing to the panel nature of data, difference-Generalized Method of Moments (difference-GMM) is used to explore and examine these relationships. Moreover, to gain deeper insights into risk management practices prevalent in Indian companies, a questionnaire survey is administered amongst the managers of sample companies. The analysis of primary data, collected through this questionnaire, is carried out with the help of statistical techniques such as Cramer’s V. Further, the case studies carried out to corroborate the findings of secondary data analysis and survey analysis are analysed using the practice-oriented comparative case study methodology. It may be noted that the analysis of the secondary data is carried out with the help of software such as Microsoft Excel (MS Excel), Statistical Packages for Social Sciences (SPSS) version 22, Econometric views (EViews) version 9, and Stata version 12.

1.8

Organisation of the Study

The study is organised into chapters. This chapter relates to background of study. Chapter 2 presents the research methodology used to carry out the study. The indices are developed in Chaps. 3–5. Chapter 3 presents the development of the risk index. Risk-disclosure index is developed in Chap. 4. Development of risk

8

1 Introduction to the study

governance index is discussed in Chap. 5. Chapter 6 is devoted to the examination of risk–return relationship. Moderating effect of governance on risk-disclosure relationship is highlighted in Chap. 7. Chapter 8 depicts practitioners’ perception on risk and Chap. 9 presents an analysis of two comparative case studies. Chapter 10 contains the concluding observations

1.9

Concluding Observations

The Chapter provides a brief outline of the study. It introduces the various dimensions of risk that are explored and examined in the study. Further, it presents a summary of the main objectives, methodology adopted to achieve these objectives, scope, and significance of the study. Moreover, it contains the chapter plan, delineating the organisation of the study.

References Committee of Sponsoring Organizations of the Treadway Commission (COSO). (2012). Risk assessment in practice. Retrieved from https://www.coso.org/…/COSO-ERM-RiskAssessment-in-Practice-Thought-Paper-Oct. Dione, G. (2013). Risk management: History, definition, and critique. Risk Management and Insurance Review, 16(2), 147–166. Gillan, S., & Starks, L. T. (1998). A survey of shareholder activism: Motivation and empirical evidence. Journal of Financial Economics, 57(1), 275–305. Institute of Chartered Accountants of India (ICAI). (2016). Institute of chartered accountants of India. Various Publications. www.icai.org/new_category.html?c_id=362. International Finance Corporation. (2012). When do companies need a board level risk management committee? Retrieved from www.ifc.org/wps/wcm/connect/…/PSO%2B31.pdf? MOD=AJPERES. Merna, T., & Al-Thani, F. (2008). Corporate risk management (2nd ed.). West Sussex: Wiley.

Chapter 2

Research Methodology

2.1

Introduction

The present chapter delineates the objectives and hypotheses based on the gaps identified from the literature reviewed, data used to test these hypotheses and the methodology that has been used in the present research study. The study aims to develop normative frameworks on various aspects of risk and attempts to determine the relationships that may exist amongst these varied aspects. Accordingly, the study makes use of both primary data (capturing the managerial views on risk management) and secondary data (involving the components of balance sheet, profit and loss account and annual reports). The research techniques used in the study are in line with the leading research initiatives on the subject. The rest of the chapter is organised into four sections. Section 2.2 presents the research objectives and hypotheses. The data used to examine these objectives has been summarised in Sect. 2.3. Section 2.4 describes the methodology used and the scope of the study. The chapter concludes with Sect. 2.5, concluding observations.

2.2 2.2.1

Research Objectives and Hypotheses Research Objectives

The present study has specifically identified the following objectives: Objective 1: To develop a risk index as a comprehensive measure of major corporate risks. Subsequent to the development of such an index, the analysis of the following parameters is proposed to be carried out: • Computation of the risk index for the sample companies for the entire period of the study. © Springer Nature Singapore Pte Ltd. 2019 M. V. Shivaani et al., Understanding Corporate Risk, India Studies in Business and Economics, https://doi.org/10.1007/978-981-13-8141-6_2

9

10

2 Research Methodology

• Detailed analysis comprising of the following sub-dimensions: – – – – –

Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group

• In-depth analysis of the index on risk by risk basis: – – – –

For the sample as a whole for the entire period of the study: Phase-wise analysis Age-wise analysis Industry-sector-wise analysis

Objective 2: To develop a risk disclosure index as a measure of the quality of risk disclosures in annual reports. Once the index is developed, the following analysis is also proposed to be carried out: • Computation of the risk disclosure index for the sample companies for the entire period of the study. • Detailed analysis comprising of the following sub-dimensions: – – – – –

Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group

• Twofold in-depth analysis of the index: – On the basis of frequency: Top 10% of most frequently disclosed risk items: For the sample companies for the entire period of study: Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group Bottom 10% of most frequently disclosed risk items: For the sample companies for the entire period of study: Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group

2.2 Research Objectives and Hypotheses

11

– On the basis of ‘richness’/‘comprehensiveness’ Top 10% of most ‘comprehensively’ disclosed risk items: For the sample companies for the entire period of study: Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group Bottom 10% of most ‘comprehensively’ disclosed risk items: For the sample companies for the entire period of study Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group • Analysis based on items pertaining to ‘risk management’: – – – – – –

For the sample companies for the entire period of study: Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-sector-wise analysis Phase-wise analysis of each of the industry-sector group

Objective 3: To develop a risk governance index as a measure of the quality of governance structure of a company. Once the index is developed, the following analysis is also intended to be carried out: • Computation of the risk governance index for the sample companies for the entire period of the study. • Detailed analysis consisting of the following sub-dimensions: – – – – –

Phase-wise analysis Age-wise analysis Phase-wise analysis of each of the age groups Industry-wise analysis Phase-wise analysis of each of the industry-sector group

• In-depth analysis of the index on a component by component basis: – – – –

For the sample as a whole for the entire period of the study: Phase-wise analysis Age-wise analysis Industry-sector-wise analysis

12

2 Research Methodology

Objective 4: To examine the relationship between risk index and accounting returns as measured by • Return on assets (ROA) – Without risk governance index as a control variable – With risk governance index as a control variable • Return on equity (ROE) – Without risk governance index as a control variable – With risk governance index as a control variable Objective 5: To explore the relationship between/among: • • • •

risk risk risk risk

disclosure index and risk index disclosure index and risk governance index. index and risk governance index index, risk governance index and risk disclosure index

Objective 6: To gain an insight into risk management practices prevalent in Indian companies on the basis of: • survey analysis • case study analysis

2.2.2

Hypotheses of the Study

To achieve the above-stated objectives, the following hypotheses have been formulated for the proposed study: 1. Hypotheses related to risk index Hypothesis I: Recession has had an impact on risk exposure levels. Hypothesis II: Age has an impact on risk exposure levels. Hypothesis III: Industry classification has an impact on risk exposure levels. 2. Hypotheses related to risk disclosure index Hypothesis IV: Recession has had an impact on the quality of/level of risk disclosures. Hypothesis V: Age has an impact on the quality of/level of risk disclosures. Hypothesis VI: Industry classification has an impact on quality of/level of risk disclosures. 3. Hypotheses related to risk governance index: Hypothesis VII: Recession has had an impact on the quality of governance structure. Hypothesis VIII: Age has an impact on the quality of governance structure.

2.2 Research Objectives and Hypotheses

13

Hypothesis IX: Industry classification has an impact on the quality of/governance structure. 4. Hypotheses related to risk–return relationship Hypothesis X: Risk (as measured by risk index) has a negative relationship with ROA. Hypothesis XI: Risk (as measured by risk index) has a negative relationship with ROE. 5. Hypotheses related to relationship among three indices. Hypothesis XII: Risk (exposure) index is negatively related to the quality of governance structure. Hypothesis XIII: Risk disclosure index is positively related to risk index. Hypothesis XIV: Risk disclosure index is positively related to the quality of governance structure.

2.3

Data

The study has attempted to analyse secondary as well as primary data pertaining to risk measurement, disclosures and governance structure. The secondary data has been taken from Bloomberg database, Ace equity database, and companies’ annual reports. Primary data has been collected through a survey conducted among the sample companies.

2.3.1

Secondary Data

The secondary data considered for the analysis can be broadly classified as follows: (i) data on the risk-free rate of return, (ii) daily closing values of the NIFTY 500 index as well as that of the constituent companies, (iii) data derived from financial statements namely, balance sheet and profit and loss account, (iv) data handpicked from annual reports from the sections in annual reports pertaining to (a) auditor’s opinion from auditor’s report, (b) number of women directors, executive/ non-executive status of Chairman, constitution of risk management committee, and existence/non-existence of whistle blower policy from the corporate governance report, (c) appointment of chief risk officer and presence/absence of CEO duality from relevant sections in the annual report and (d) risk disclosures made in ‘risks and concerns’ section and ‘opportunities and threats’ section in the management discussion and analysis part of the annual reports, and risk disclosures made in ‘risk management’ section (wherever provided).

14

2 Research Methodology

2.3.2

Primary Data

In addition to the secondary data analysis, the study has used primary data to corroborate the findings from the secondary data. For the purpose, a questionnaire has been administered amongst the sample companies. The questionnaire captured the opinion of the respondents on eight major aspects related to, (i) attitude towards risk, (ii) risk identification and prioritisation, (iii) risk measurement, (iv) risk handling, (v) risk communication, (vi) risk management authority, (vii) risk and return and (viii) evaluation of risk management practices.

2.4 2.4.1

Research Methodology and Scope Research Methodology

The objectives of the study have been addressed using a comprehensive approach. The approach to address the risk management aspects covered in the study has been demonstrated in Fig. 2.1. Form the figure, it may be deciphered that the objectives 1 through 5 have been addressed using secondary data and objective 6 by using primary data. Further, the objectives 1 through 3 are based on the normative frameworks and generally accepted methodology for the construction of composite indicators (indices). Composite indicators are increasingly becoming popular in view of their

ObjecƟve 1,2,3, 4, 5 & 6

ObjecƟve 1,2,3

ObjecƟve 4

Fig. 2.1 Outline of research methodology

ObjecƟve 6

ObjecƟve 5

2.4 Research Methodology and Scope

15

ability to capture data pertaining to more than one indicator (Tangian 2004, Hudrlikova and Petkovova 2013). As per Organisation for Economic Co-operation and Development (OECD) (2008), composite indicators are capable of summarising complex, multidimensional realities and are easy to interpret (enhancing their utility for decision makers). Further, they are of immense utility in inter-temporal studies and reducing the visible size of the set of indicators without loss of information. They enable effective comparison and analysis of complex dimensions. Despite the above-mentioned advantages, composite indicators (CI) also have some shortfalls; selection of indicators and weights may be subjected to debate, CI may be misused to put forth a particular policy implication and bias, and, above all, lack of transparency in the construction process itself may disguise failings of some dimensions (OECD 2008). It is worth mentioning that scales proposed for various indicators used in the construction of risk index (Objective 1) have been shared with select industry and academic experts and duly verified by them. The risk disclosure index (as per Objective 2) is based on content analysis and textual analysis of information provided in management discussion and analysis section of annual reports. Content analysis is a technique that facilitates coding and interpretation of written material, thereby enabling the researcher to draw valid and reliable inferences from the narratives under consideration (Holsti 1968; Krippendorff 1980; Weber 1985). It usually involves counting the number of relevant words, sentences or themes. The textual analysis focuses on the content of narrative in terms of its semantic attributes (Jia et al. 2016). In context risk disclosure studies, Ntim et al. (2012), in a first of its kind attempt, have attempted to bring together elements of both content analysis and textual analysis. The risk governance index developed (in pursuance of Objective 3) is based on the definition of risk governance as put forth by International Finance Corporation (IFC 2013). The index also draws heavily on popular corporate governance indices such as GIM index (Gomper et al. 2003), Gov-score (Brown and Caylor 2006) and entrenchment index (Bebchuk et al. 2008). The empirical analysis based on these indices has been carried out with the help of descriptive statistics and inferential statistics. Mann–Whitney U test has been carried out for phase-wise analysis, phase-wise analysis for each of the age groups, and phase-wise analysis for each of the industry classifications. Similarly, Kruskal–Wallis test has been used to carry out age-wise and industry-wise analysis. It may be noted, that Mann–Whitney U test and Kruskal–Wallis test can be seen as non-parametric counterparts of independent-t test and one-way ANOVA. The rationale for using non-parametric test lies in the non-normal distribution of the data under analysis. The data pertaining to Objectives 4 and 5 falls in the category of panel data. Therefore, panel data regression appears to be the appropriate technique to gauge the relationship between risk index and returns (Objective 4) and to explore the relationships among the three indices (Objective 5). Panel data analysis provides several advantages over pooled ordinary least squares (OLS) regression. It facilitates consideration of individual/firm-specific heterogeneities that may be having an impact on the dependent variable, provides more informative data, more degrees of

16

2 Research Methodology

freedom and more efficiency (Hsiao 2003; Baltagi 2008). Further, Wintoki et al. (2010) suggest three potential sources of endogeneity that may exist in panel data structures: (i) Dynamic endogeneity—it occurs when the preceding periods’ values of a variable influence its current period values. This form of endogeneity has been often observed in studies in the area of finance and accounting (Hermalin and Weisbach 1998). (ii) Simultaneity—It occurs when two variables simultaneously affect each other, resulting in their co-determination. (iii) Unobserved heterogeneity —It is a phenomenon where the relationship between two variables is affected by some third unobservable variable. In general, these may be attributed to firm-specific characteristics or firm-fixed effects (Haubrich 1998). The most common solution to deal with endogeneity problems is the use of lagged dependent variables or instrumental variables. The estimation techniques that may be employed are OLS, fixed effects or dynamic panel data generalized method of moments (GMM). If OLS is used for estimation, it typically results in an upward bias in the coefficient of lagged dependent variable (Bond 2002). Similarly, in the context of unobservable firm heterogeneities, Baltagi (2008) discourages the use of fixed effects model (particularly, when the panel is a short panel). He suggests that the lagged dependent variable may end up being correlated with error-term, resulting in biased coefficients. Further, the coefficients of the lagged dependent variable, obtained through fixed effects estimation may have a downward bias (Nickel 1981). To overcome these problems, Holtz-Eakin et al. (1988) proposed generalized method of moments (GMM) panel specifications, which has drawn greater attention of researchers after its use by Arellano and Bond (1991), Arellano and Bover (1995) and Blundell and Bond (1998). It is worth mentioning that Arellano and Bond (1991) first-difference the panel data to remove the time-invariant fixed effect and show that the lagged dependent variables’ values (levels) constitute legitimate variable, provided that the residuals are free from second-order serial correlation. It is noteworthy that the significant relation between lagged dependent variable and dependent variable in each of the regression models developed in pursuance of Objective 4 and Objective 5 justifies the use of GMM. Diff.-GMM can be best explained in the words of Schultz et al. (2010) as follows: ‘The GMM procedure, as developed by Hansen (1982), provides a non-parametric approach to estimating model parameters. GMM is an appealing approach to modelling the relationships of interest in finance as it imposes no distributional assumptions on the model specification. As such, GMM standard errors are robust to autocorrelation and heteroskedasticity of unknown form. Holtz-Eakin et al. (1988) and Arellano and Bond (1991) develop the difference GMM specification for dynamic panel datasets that produces consistent parameter estimates in the presence of endogeneity. These estimates are robust to dynamic endogeneity, firm-fixed effects, endogenous regressors, heteroskedasticity and serial correlation. A desirable property of the difference GMM is the use of internal instruments that are embedded in the existing dataset, and hence, are readily available to the econometrician. Moreover, all variables are time differenced so that

2.4 Research Methodology and Scope

17

unobservable firm characteristics are eliminated without the necessity for strict exogeneity assumptions, allowing for the inclusion of the lag of the dependent variable on the right-hand side to account for any possible dynamic endogeneity, or: DP ¼ L  Da þ DGb þ DXg þ DE

ð2:1Þ

where: L is a one period lag operator; DP is a (N − I)  1 vector of the differenced dependent variable across N observations and I firms; a is a 1  1 scalar of the coefficient for the lag time differenced dependent variable, LDY, across N observations; DG is a (N − I)  H matrix of the H differenced independent variables, across N observations and I firms; b is a H  1 vector of coefficients for the H differenced independent variables; DX is a (N − I) Q matrix of the Q differenced firm control variables across N observations and I firms; η is a Q  1 vector of coefficients for the Q differenced firm control variables; and, DE is a (N − I)  1 vector of error terms across N observations and I firms. To derive the instrument set for the GMM estimation of Eq. 2.1, sequential exogeneity is assumed (Wooldridge 2002, and illustrated in Wintoki et al. 2010). Sequential exogeneity is a much weaker condition than strict exogeneity, and provides a more realistic setting for the analysis of the relationships in the finance domain. Additionally, it is assumed that there is some form of persistence in the dependent, independent and control variables, such that past realisations of firm characteristics may serve as valid instrumental variables. Given the preceding assumptions, valid instruments for the orthogonality conditions of Eq. 2.1 include the lags of dependent, independent and control variables. Selected lags have the desirable instrumental variable properties of being correlated to the regressors, yet uncorrelated with contemporaneous errors. For potentially endogenous variables such as G, lags of 2 or more are valid instruments. For predetermined regressors, such as X and L  P, lags of 1 or more are available as instruments for the estimation of Eq. 2.1. Using lagged levels rather than lagged differences as instruments maximises the number of observations available for estimation and readily allows for the overidentification of moment restrictions. Given the linear relation, Eq. 2.1, and the available instruments, the following moment conditions are specified: 3 ðDP  L  DPa  DGb  DXgÞ  L2  GÞ .. 7 6 7 6 . 7 6 6 ðDP  L  DPa  DGb  DXgÞ  LJ  GÞ 7 7 6 6 ðDP  L  DPa  DGb  DXgÞ  L  XÞ 7 7 6 1 X6 7 .. gT ¼ 7 6 . 7 6 N 6 ðDP  L  DPa  DGb  DXgÞ  LB  XÞ 7 7 6 6 ðDP  L  DPa  DGb  DXgÞ  L2  XÞ 7 7 6 7 6 .. 5 4 . 2

ðDP  L  DPa  DGb  DXgÞ  LV  PÞ

ð2:2Þ

18

2 Research Methodology

where: L is a one period lag operator; LS  X is the Sth-lag operator on variable X; DP is an (N − I)  1 vector of the differenced dependent variable, across N observations and I firms; a is a 1  1 scalar of the coefficient for the lag time differenced dependent variable, L  DY, across N observations; DG is an (N − I) H matrix of the H differenced independent variables, across N observations and I firms; b is a H  1 vector of coefficients for the H differenced independent variables; DX is an (N − I) Q matrix of the Q differenced firm control variables, across N observations and I firms; η is a Q  1 vector of coefficients for the Q differenced firm control variables; and, J, B, and V are the maximum lag length of instruments created by the independent, control, and dependent variables, respectively. Although all the moment conditions in Eq. 2.2 should theoretically converge to zero if correctly specified, this is impossible when the number of moment conditions exceeds the number of parameters to be estimated. That is, when the model is overidentified. Therefore, the goal is to minimise Eq. 2.2 based on a weighting matrix, M:   min gTT MgT

ð2:3Þ

Let h equal the vector of coefficients, a, b, and η, for Eq. 2.2. For the efficient estimation of h, M must equal the inverse of the variance–covariance matrix of Eq. 2.2, X−1, as highly correlated orthogonality conditions and those with a high degree of variation will be allocated lower weights due to their lesser informational content. While the researcher is unable to observe X−1 without initial values of h, a consistent though not necessarily efficient estimate of h can be derived by using an arbitrary estimate of X, such as one assuming homoskedasticity. Defining D as a simplistic estimate of X that assumes the errors to be independent and identically distributed, implementing D1 as the weighting matrix in Eq. 2.3 will result in a h D to estimate consistent estimate of the model parameters, b h D . One can then use b b Therefore, the final the variance–covariance matrix of the moment conditions X. estimation of h is achieved by h i b 1 g min gTT X T

ð2:4Þ

This procedure is known as two-step GMM estimation. The resultant estimator for the difference GMM model, b h DGMM , is consistent and efficient in the presence of heteroskedasticity and serial correlation of unknown form in the panel dataset. b h DGMM , is also robust to the potential unobservable heterogeneity, simultaneity and dynamic endogeneity present in the corporate governance and performance relation (Roodman 2009)’. Objective 6 involves twofold analysis, (i) survey analysis, and (ii) case study analysis. The survey analysis includes descriptive analysis and hypothesis testing. Hypotheses testing has been carried out with the intent of corroborating the findings of secondary data analysis. Due care has been exercised to avoid the problem of

2.4 Research Methodology and Scope

19

non-response bias. Further, practice-oriented comparative case study methodology has also been employed to gain insights into risk management practices prevalent in Indian companies. The analysis involves an in-depth assessment of two of the sample companies. The analysis involves descriptive statistics and hypothesis testing. For an in-depth analysis, 28 hypotheses covering various aspects of risk measurement, disclosure and governance and performance have also been tested.

2.4.2

Scope of the Study

The study is confined to non-financial companies (429) that constituted Nifty 500 index as on 31 March 2014. NIFTY 500 represents more than 90% of the free float market capitalization of the stocks listed on the National Stock Exchange of India (NSE). It may be noted that financial and banking companies have been excluded (i) because of the specialised nature of their operations, (ii) as they are governed by different laws, rules and regulations. The study covers a time span of 10 years from 1 April 2005 to 31 March 2015. For the analysis of the impact of the recession, the study has been segregated into two phases, namely, pre-recession (2005–2008) and post-recession (2009–2015). The management survey has been attempted for all 429 non-financial companies that form a part of the sample.

2.4.3

Use of Statistical Software

The study has made extensive use of Microsoft Office (Microsoft Word and Microsoft Excel) and data processing software such as Statistical Package for Social Sciences 22.0 (SPSS), econometric views 9.0 (EViews), and Stata 12.0.

2.5

Concluding Observations

This chapter describes the methodology used in the present study. The identified research gaps, objectives of the study, and hypotheses have also been outlined in the chapter. Moreover, the details pertaining to the data used, techniques employed for investigation and analysis and statistical tools used have also been provided. The methodology is chosen on the basis of sound logic, the outcome of previous research works and, above all, its practicality. The selection of methodology is done keeping in mind the availability (and constraints) of data for Indian corporates. Based on the outline presented in the chapter, the empirical analysis has been carried out in forthcoming chapters.

20

2 Research Methodology

References Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58(1), 277–297. Arellano, M., & Bover, O. (1995). Another look at the instrumental variable estimation of error components models. Journal of Econometrics, 68(1), 29–51. Baltagi, B. (2008). Econometric analysis of panel data (4th ed.). Chichester, UK: Wiley. Bebchuk, L., Cohen, A., & Ferrel, A. (2008). What matters in corporate governance? The Review of Financial Studies, 22(2), 783–827. Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143. Bond, S. R. (2002). Dynamic panel data models: A guide to micro data methods and practice. Portuguese Economic Journal, 1(2), 141–162. Brown, L., & Caylor, M. (2006). Corporate governance and firm valuation. Journal of Accounting and Public Policy, 25(4), 409–434. Gompers, P., Ishii, J., & Metrick, A. (2003). Corporate governance and equity prices. Quarterly Journal of Economics, 118(1), 107–155. Hansen, L. P. (1982). Large sample properties of generalized method of moments estimators. Econometrica: Journal of the Econometric Society, 1029–1054. Haubrich, J. G. (1998). Bank diversification: Laws and fallacies of large numbers. Economic Review, Federal Reserve Bank of Cleveland, 34(2), 2–9. Hermalin, B., & Weisbach, M. (1998). Endogenously chosen boards of directors and their monitoring of the CEO. American Economic Review, 88(1), 96–118. Holsti, O. R. (1968). Content analysis. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology (2nd ed., Vol. II, pp. 596–692). New Delhi: Amerind Publishing Co. Holtz-Eakin, D., Newey, W., & Rosen, H. (1988). Estimating vector auto regressions with panel data. Econometrica, 56(6), 1371–1395. Hsiao, C. (2003). Analysis of panel data. London, UK: Cambridge University Press. Hudrlikova, L., & Petkovova, L. (2013). Sustainable development—Czech region ranking using multi-criteria decision analysis. European Journal of Sustainable Development, 2(4), 253–262. International Finance Corporation. (2013). Risk culture, risk governance, and balanced incentives: Recommendationsforstrengthening riskmanagementin emergingmarketbanks.Washington,DC: IFC. Jia, J., Munro, L., & Buckby, S. (2016). A finer-grained approach to assessing the quality (quantity and richness) of risk management disclosures. Managerial Auditing Journal, 31(8/9), 770–803. Krippendorff, K. (1980). Validity in content analysis. In E. Mochmann (Ed.), Computer analyse (pp. 69–112). Frankfurt, Germany. Retrieved from http://repository.upenn.edu/asc_papers/291. Nickell, S. (1981). Biases in dynamic models with fixed effects. Econometrica, 49(6), 1417–1426. Ntim, C. G., Opong, K. K., & Danbolt, J. (2012). The relative value relevance of shareholder versus stakeholder corporate governance disclosure policy reforms in South Africa. Corporate Governance: An International Review, 20(1), 84–105. Organisation for Economic Co-operation and Development. (2008). Handbook on constructing composite indicators: Methodology and user guide. OECD, European Commission, Joint Research Centre OECD Publishing. Available at https://www.oecd.org/std/42495745.pdf. Roodman, D. (2009). How to do xtabond2: An introduction to difference and system GMM in Stata. The Stata Journal, 9(1), 86–136. Schultz, E. L., Tan, T. D., & Walsh, K. D. (2010). Endogeneity and the corporate governance– performance relation. Australian Journal of Management, 35(1), 145–163. Tangian, A. (2004). Defining the flexicurity index in application to European countries. WSI-Diskussions papier, No. 122. Available at http://www.boeckler.de/pdf/p_wsi_diskp_122.pdf. Weber, R. P. (1985). Basic content analysis. Series: Quantitative Applications in the Social Sciences. A Sage University Paper, Series/No. 07-049. Beverly Hills: Sage Publications. Wintoki, M. B., Linck, J. S., & Netter, J. M. (2010) Endogeneity and the dynamics of internal corporate governance. In CELS 2009 4th Annual Conference on Empirical Legal Studies. Available at SSRN: http://ssm.com/abstract=970986. Wooldridge, J.M. (2010). Econometric analysis of cross section and panel data. Massachusetts: MIT Press.

Chapter 3

Normative Framework for Risk Index and Its Empirical Analysis

What gets measured, gets managed. —Peter Drucker

Abstract This chapter aims to provide a normative framework (primarily based on accounting information) for measurement of corporate risk. The index is based on nine risks and has been developed using expert opinion. In the process two new ratios, namely, modified defensive interval ratio and contingency coverage ratio have been developed. The index has been then empirically computed for sample 429 non-financial companies for each of the 10 years from 2005 to 2015. In addition an aggregative analysis, focussing on phase-wise, age-wise, and industry-wise analysis has been carried out. Further, a dis-aggregative (risk-wise) detailed analysis has been carried out to develop deeper understanding of risks surrounding the Indian corporates.




Keywords Risk index Market risk Accounting risk Operating risk




3.1


Liquidity risk
Solvency risk


Introduction

Risk has long perplexed managers and investors alike. Much of the ambiguity in the domain stems from the use of different definitions of risk for different purposes. The present study considers an all—encompassing definition of risk, recognising both positive and negative connotations. In light of the worldwide corporate debacles like that of Enron and Satyam, focus on risk management has been gaining momentum. Literature is replete with studies making use of accounting data for bankruptcy predictions. Most of the studies focus on exploring variables that best describe this outcome. But, bankruptcy is just one of the possible outcomes if the vulnerabilities to which the company is exposed materialise. Depending on the severity and area of impact of an event, risks may result in decline in market share, temporary shut-down, management turnover, insolvency, winding-up, to name a few © Springer Nature Singapore Pte Ltd. 2019 M. V. Shivaani et al., Understanding Corporate Risk, India Studies in Business and Economics, https://doi.org/10.1007/978-981-13-8141-6_3

21

3 Normative Framework for Risk Index …

22

consequences. Further, very few studies provide an explicit interpretation of ratios as a measure of risk. Thus, recognising the fundamental strength of ratios and need for a risk measurement tool, the present study aims to develop a corporate risk index. The index is proposed to be in the form of a normative framework. Based on nine major risks that corporates face, it is expected to provide all stakeholders an easily comprehensible and applicable tool of corporate risk measurement. The chapter has been organized into seven sections. Section 3.2 elaborates the methodology employed for index construction. Section 3.3 describes the sample used and sources of data. Section 3.4 examines the findings and presents the analysis of the same. This is followed by a Sect. (3.5) on implications for various stakeholders. Section 3.6 deals with limitation of the study. Last but not the least, Sect. 3.7 contains the concluding observations.

3.2

Methodology

Sometimes risk is viewed as a product of exposure, impact and frequency. However, it is almost impossible to gauge the impact and frequency of a particular threat for a particular company, purely based on publicly available information. The only part of risk that can be estimated with some accuracy is the exposure, using financial information provided in the financial statements. To this end, financial ratios can be appropriately used. In fact, most of the studies on financial distress and bankruptcy prediction have made use of ratios. The success can be gauged from widespread use of Ohlson’s O score and Altman’s Z score (both of which are exclusively based on ratios). Also, the interpretation of ratios is consistent with them being peculiar representatives of exposure levels. For example, the most commonly used measure of solvency of a company is debt to equity ratio. If a company has a debt equity ratio of 0.8 (D/E = 0.8/1), it is equivalent to saying that if the value of equity falls by more than 20%, then the investments of debenture-holders may be in danger. In other words company may not be able to pay its debenture-holders in full. Being an external party to firm, one cannot gauge the probability of this event, particularly, when little or no historical data of such event is available. Also, one cannot claim with certainty the severity or impact of this event. In an extreme case it may lead to shut down of company and in more general cases it may lead to loss of reputation following a legal suit and claim for damages. The only part one can say with surety is, exposure, as probability and severity may be conditional and dependent on situations and events beyond public domain. In view of the above, it is very appropriate to state that ratios are excellent indicators of various risks. Pursuing this stream of thought, and taking cue from work of Tamari (1966), a risk index (more appropriately an ‘exposure index’) has

3.2 Methodology

23

been attempted. This may also be viewed as normative ‘tolerance limits’ for various risks. The first step in construction of a risk index is to identify the risks that affect a company. For the purpose, it would be in order to include all possible risks that may affect a company. Since companies are unique in terms of their organizational features, culture, risk appetite, tolerance, management practices; ‘one size fits all’ approach sounds unrealistic. Therefore, based on literature and expert opinion, only those risks have been focused upon which are believed to be pervasive and material. Even COSO propagates management by exception; in operational terms, focus should be on the most important risks only. Accordingly, following risks were identified for the study: accounting risk, attrition risk, competition risk, credit risk, customer satisfaction, exchange rate risk, interest rate risk, liquidity risk, market risk, operational risk, risk of over dependence on a product or division, risk related to innovation, solvency risk and taxation risk. But owing to unavailability of accurate and reliable data on some of the risks (for majority of the companies), only following nine risks could be considered for the construction of the index; market risk, accounting risk, competition risk, contingency risk, credit risk, exchange rate risk, liquidity risk, operating risk, and solvency risk. It is pertinent to note that Muzzy (2008) also advises the use of a filtering process to limit the number of manageable risks to 10 or fewer. Most finance theories assert that only systematic risk is relevant for investors. But, the present study considers both systematic and unsystematic risk, mainly for two reasons. First, unless the investors are aware of various firm specific risks, they cannot diversify them effectively. Second, the study intends to propose a normative framework that is expected to be of use to all stakeholders from shareholders to prospective investors, regulators, rating agencies, fund managers and most importantly the risk managers in company itself. It is pertinent to state that every risk has been coded on a scale of 1–5; where 1 represents the least risky proposition and 5, the most risky. Thus, with nine risks, each on a scale of 1–5, the minimum risk index value that a company can have is 9 and the maximum is 45. It may further be noted that every risk score and its corresponding risk scale has been duly verified and validated by experts (both academic and industry). Further, the index values obtained for various companies have been expressed in percentage terms. The maximum possible index value of 45 has been taken as the base to express the computed index values in percentage form. In other words, if a company scores 1 on each of the nine risks, its risk index value will be 9; but while expressing it as a percentage of maximum possible risk index value (i.e. 45), it will be 20% (i.e. 9/45). Similarly, if a company has 80% risk exposure, it will imply that it has a risk index value of 36 (i.e. 80% of 45); the lower is the index value, the lower is the risk and vice versa. A brief about the nine risks included in the study and the scale construction for each of these risks is described as given below.

3 Normative Framework for Risk Index …

24

Risks: Modern finance theory is based on the premise that risk surrounding a company can be divided in two parts—systematic risk and unsystematic risk. 1. Market risk: Market risk or systematic risk is that part of company risk which cannot be diversified away. Its sources include inflation risk, political risks, interest rate risk, etc. These risks affect almost all the companies in the economy, but in varying degrees. The most common measure of market risks is, beta. It basically measures the relation between company’s return and returns of market as a whole. A beta equal to 1 represents average level of risk, whereas, a beta more than 1 is usually associated with higher degrees of risk and volatility. Such stock will move with market but to a greater extent. The following model was used to calculate beta for each company for each year. Rt ¼ a þ bðRm  Rf Þ þ et

ð3:1Þ

where, Rt Rm Rf b

Weekly equity return of a company Weekly return of NIFTY 500 index Weekly risk-free rate based on 91 day T-bill Sensitivity of company’s returns to market returns

It is noteworthy that Eq. 3.1 has been run for every company for every year separately, subject to availability of data. Thus, close to 4,000 regression equations resulted in the betas required for every firm-year observation. It may be noted that only the companies with stationary returns have been considered for the purpose of regression. Further, companies having negative betas have not been considered for the purpose of the study. The reason being, they are supposed to be a result of peculiar events that may have happened during the period of estimation. These events may be in form of law suits and takeovers that may have disturbed company’s correlation with the market. As these are rare events and no entire sector has ever recorded negative beta, its underlying factors remain a puzzle (Damodaran 2009).

Exhibit 3.1 Scoring in relation to market risk

Beta range

Risk score

0 1/2

1 3 5

appointed to bring to the board: independence, impartiality, wide experience, special knowledge and personal qualities (Financial Stability Board 2013). Further, they are expected to provide strategic direction to company, monitor its performance, communicate effectively with outside parties and manage risk. In the context of risk management, they are expected to ensure the integrity of financial controls and robustness as well as defensibleness of risk management systems (Institute of Directors 2010). Therefore, if the board has less than legally required number of non-executive directors, it has been given a score of 1, and if it has more than the prescribed legal minimum then a score of 5. The same has been presented in Exhibit 5.3. 4. Executive/non-executive Chairman Legal status—There is no legal requirement regarding whether the Chairman should be an executive or non-executive director. Perspective from literature—Chairman is expected to act as a link between the management and shareholders. Further, in the event of an equality of votes on any issue, he gets a casting vote. Higgs report (2003) outlines the duties of the Chairman. These include upholding standards of integrity and probity, promotion of communication between the executive and non-executive directors, coherent leadership of the company to name a few. Therefore, in a bid to have transparency and fairness in governance structure, it seems desirable to have a non-executive director as Chairman. Therefore, the following scoring (Exhibit 5.4) in relation to the status of Chairman has been proposed. 5. Proportion of independent directors2 Legal status—149(4) of Companies Act 2013 requires that every Board shall have at least one-third of the total directors as independent directors. 2

As per section 149(6) of Companies Act 2013, an independent director in relation to a company, means a director other than a Managing Director, Whole-Time Director Or Nominee Director, (a) who in the opinion of the board is a person of integrity and possess relevant expertise and experience; (b) who is or was not a promoter of the company, who is not related to promoters/ directors of company; (c) who has or had no pecuniary relationship with the company; (d) none of whose relatives has or had pecuniary relationship or transaction with the company; (e) who neither himself nor any of his relatives i. holds or has held the position of key managerial personnel, ii. is or has been an employee or proprietor or a partner, in any of the three financial years preceding, iii. holds together with his relative 2% or more of the total voting power of the company; or iv. is a Chief Executive or director, of any nonprofit organisation, or who possesses such other qualifications as may be prescribed.

174 Exhibit 5.4 Scoring in relation to executive– non-executive status of chairman

5 Normative Framework for Risk Governance Index … Status of chairman

Score

Executive chairman Non-executive chairman

3 5

Whereas, the SEBI requirements are that where the Chairman of the board is a non-executive director, at least one-third of the board should comprise of independent directors and in the case of an executive Chairman at least one-half of the board of the company shall consist of independent directors. Perspective from the literature—Boyer and Stern (2012) depict independent boards as a good governance feature and expect firms with more independent boards to pay a lower premium. The role of the independent director is to advise the board on strategic matters and provide an independent voice to monitor the balance of executive power. Further, they have incentives and the expertise to act in ways that maximize the returns to shareholders. Prior studies suggest that there are some positive shareholder outcomes associated with independent boards including lower instances of earnings manipulation and fraud (Dechow et al. 1996; Beasley 1996; Klein 2002); superior decision-making (Dahya and McConnell 2005); and greater levels of disclosure (Karamanou and Vafeas 2005). In view of the regulatory requirements and perspective from the literature, the following scoring (Exhibits 5.5 and 5.6) has been proposed in the context of independent directors (a) When the Chairman is an executive director. (b) When the Chairman is a non-executive director 6. CEO duality Legal status—There is no legal provision pertaining to CEO duality, however, Section 203 of Companies Act 2013 mandates the appointment of managing director (MD)/ whole-time director (WTD)/Manager/CEO. Further, Clause 49 amended, introduced a non-mandatory provision which states that ‘separation is required unless articles of the company permits otherwise or the company does not have multiple businesses’. Therefore, if the MD/WTD/Manager/CEO also serves as the Chairman, then it is a case of CEO duality. Perspective from the literature—When the CEO also serves as the Chairman, the board’s ability to fulfil its supervisory function is significantly reduced due to conflict of interests (Brickley et al. 1997). Further, Rechner and Dalton (1991) suggest the absence of CEO duality facilitates effective monitoring of the activities of top management and results in a reduction in agency costs. CEO duality is often

Exhibit 5.5 Scoring in relation to the proportion of independent directors, with executive chairman

Proportion of independent directors on board (PoID)

Score

0 < PoID < 1/2 PoID = 1/2 PoID > 1/2

1 3 5

5.2 Methodology

175

Exhibit 5.6 Scoring in relation to proportion of independent directors, with non-executive chairman

Proportion of independent directors on board (PoID)

Score

0 < PoID < 1/3 PoID = 1/3 PoID > 1/3

1 3 5

Exhibit 5.7 Scoring in context of CEO duality

CEO duality

Score

Yes No

3 5

seen as a red flag for entrenchment. Therefore, CEO non-duality is often preferred for strategic as well as operational reasons. Therefore, the scoring for CEO duality (Exhibit 5.7) is proposed as follows: 7. Chief risk officer (CRO) Legal status—There is no legal provision in this regard. Perspective from the literature—Appointment of a CRO is often linked with likely implementation of enterprise-wide risk management (ERM) (Beasley et al. 2005; Daud et al. 2010). ERM is an integrated risk management approach that focuses on the attainment of organisational goals while managing its risks. Further, it is believed that CRO will act as a supporting pillar in the development of risk management policies, frameworks and analysis (Aabo and Simkins 2005). Given the importance of CRO in the context of risk management, his absence has been equated with non-compliance of law and hence a score of 1 and his presence has been awarded a score of 5 (shown in Exhibit 5.8). 8. Whistle-blower policy Legal status—Revised clause 49 of the listing agreement wide circular no. CIR/ CFD/POLICY CELL/2/2014 dated April 17, 2014 makes mandatory the formulation of a whistle-blower policy w.e.f. 1 October 2014. The clause requires a ‘vigil mechanism to report about unethical behaviour, actual or suspected fraud, violation of the company’s code’. Perspective from the literature—Implementation of a whistle-blower policy and protection of whistle-blowers has been advocated by several regulations and legislations worldwide (e.g.: The Public Interest Disclosure Act, 1998, in the UK; Sarbanes Oxley Act, 2002, in the US). With the increasing thrust on operational risks and incidence of frauds and corrupt practices, the existence of a vigil mechanism seems imperative. Whistle-blowing at the right time may save the company Exhibit 5.8 Scoring in the context of the appointment of a CRO

Appointed CRO

Score

No Yes

1 5

5 Normative Framework for Risk Governance Index …

176 Exhibit 5.9 Scoring in context of the existence of a whistle-blower policy

Implemented a whistle-blower

Score

No Yes

1 5

from financial loss, scathing publicity or costs of litigation (Rothschild and Miethe 1999). Therefore, the existence of a whistle-blower policy has been assigned 5 points and its absence, a score of 1 (Exhibit 5.9). 9. Risk management committee Legal status—Revised clause 49 of the listing agreement wide circular no. CIR/ CFD/POLICY CELL/2/2014 dated April 17, 2014 w.e.f. 1 October 2014, requires companies to constitute a risk management committee. Such committee is not required under Companies Act 2013. Perspective from the literature—Risk management committee (RMC) has been defined as a committee that is charged with the responsibility of organisational risk, advising the Board on firm’s overall current and future risk appetite and risk strategy and implementation of that strategy (FSB 2013). RMC can be viewed as the central node connecting all the risk parameters like appetite, tolerance limits, strategy and its implementation. Therefore, the scoring for RMC has been proposed as shown in Exhibit 5.10. 10. Compulsory committees Legal status—As per Companies Act 1956, only the audit committee and investor grievance committee were required. But, as per Companies Act 2013, w.e.f. 1.4. 2014, the provisions regarding compulsory committees are as follows: Section 177 continues with the mandatory status of the audit committee. Further, Section 178 makes mandatory the constitution of a nomination and remuneration committee, Section 178 also mandates stakeholder relationship committee (if no. of shareholders, debenture-holders and deposit-holders and any other security holders is greater than 1000). Corporate social responsibility committee has been made mandatory under Section 135 (1) read with rule 3 of Companies (Corporate Social Responsibility Policy) Rules, 2014). In addition, revised clause 49 of the listing agreement wide circular no. CIR/ CFD/POLICY CELL/2/2014 dated 17 April 2014 w.e.f. 1 October 2014, requires companies to constitute a risk management committee. Keeping in view the intent behind making certain committees mandatory; if a company defaulted in constituting even a single compulsory committee, it has been

Exhibit 5.10 Scoring in context of the existence of risk management committee

Existence of a risk management committee

Score

No Yes

1 5

5.2 Methodology Exhibit 5.11 Scoring in the context of compulsory committees

177 Number of compulsory committees

Score

At least 1 compulsory committee is not in place All compulsory committees are in place

1 5

accorded a score of 1. It is only when all the compulsory committees have been constituted by a company (during the relevant financial year) that a score of 5 has been assigned (Exhibit 5.11). 11. Voluntary committees Legal status—There is no legal provision and limits on the constitution of voluntary committees. A company is free to constitute as many committees as it considers useful and relevant. Based on a perusal of 4033 Annual reports, 33 committees were identified as existing in practice. A number of committees had different names but had the same functions and role. Therefore, for the purpose of the study, the committees have been classified into five broad categories, namely, shares-related committees (includes committees like share transfer, ESOP, etc.); finance-related committees (includes borrowing committee, financial management committee, etc.); human resource (HR) related committees (e.g.: screening committee, conflict resolution committee, etc.); management-related committees (e.g. compliance committee, corporate management committee, etc.) and miscellaneous category (with committees like information security committee, innovation committee, etc.). If a company has no voluntary committee, it will have a score of 1 and if a company has committees belonging to all 5 categories, it will be given a score of 5. The scoring can be illustrated as follows: If a company has, say, 1 committee by the name of Finance committee, it will be given a score of 1.8. In addition, if a company has 2 committees namely, bond committee and financial management committee, it will still be given a score of 1.8, as both these committees belong to the same category, i.e. finance committees. Whereas, if a company has two committees namely, ESOP committee and borrowing committee, it will be given a score of 2.6 as these committees belong to two different categories. The scales have been defined in Exhibit 5.12.

Exhibit 5.12 Scoring in context of number of voluntary committees constituted

Constituted voluntary committees

Score

None Only in 1 category In 2 categories In 3 categories In 4 categories In all 5 categories

1 1.8 2.6 3.4 4.2 5

5 Normative Framework for Risk Governance Index …

178

5.3

Sample

The sample consists of non-financial companies that constitute NIFTY 500 index as on 31 March 2014. The study covers a period of 10 years from 1 April 2005 to 31 March 2015. The period of study is of particular importance as it includes the recession period, which impacted the world economy towards the second half of 2008. As per the United Nations Council on Trade and Development (UNCTAD), Investment Brief (November 1, 2009), the year 2008 marked the end of a growth cycle in global foreign direct investment. Worldwide flows came down by more than 20%. This global financial crisis reduced access to financial resources both internally as well as externally (Singh et al. 2012). Thus, the study considers two phases, Phase I (pre-recession period) 1 April 2005 to 31 March 2008 (2006–2008) and 1 April 2008 to 31 March 2015 (2009–2015) as Phase II (post-recession period). As the findings of past studies suggest that company age and industry classification influence governance structure of a company, the study also looks at these aspects in the analysis. Therefore, the companies have been divided into three age categories—(based on year of incorporation): young companies (bottom 25% (Quartile 1)), middle-aged companies (i.e. middle 50% companies, falling in Quartile 2 and Quartile 3, and old and established companies (i.e. top 25% companies or companies in Quartile 4). Further, for an industry-wise analysis, companies have been regrouped into 15 industry groups, namely, agriculture, capital goods, chemical, diversified, fast moving consumer goods (FMCG), healthcare, housing, information and communication technology (ICT), media, metal, miscellaneous, oil and gas, power, textile and transport. The variables of interest were collected from Bloomberg database and missing variables were completed from the annual reports of sample corporates.

5.4

Empirical Evidence

The normative framework for measuring risk has been empirically tested, using the sample described in Sect. 5.4. The scope of the analysis is depicted in Fig. 5.1.

5.4.1

Aggregative Analysis

5.4.1.1

Full-Sample Analysis

Table 5.1 depicts the year-wise descriptive statistics for the risk governance index. It is evident from Table 5.1 that the quality of governance structure has steadily and persistently increased from about 68% in 2006 to 79% in 2015. This may be

5.4 Empirical Evidence

179

Fig. 5.1 Scope of analysis

Table 5.1 Descriptive statistics of year-wise risk governance index on full sample, 2005–2015 Year

N

Mean (%)

Median (%)

Std. deviation (%)

Minimum (%)

Maximum (%)

2005– 2006 2006– 2007 2007– 2008 2008– 2009 2009– 2010 2010– 2011 2011– 2012 2012– 2013 2013– 2014 2014– 2015

387

68.41

70.20

8.46

41.60

93.80

379

68.81

70.20

8.97

36.00

93.80

397

69.74

71.20

8.67

41.60

93.80

399

70.50

71.20

8.54

41.60

93.80

406

71.27

71.60

8.45

45.60

95.00

413

71.61

71.60

8.31

44.00

95.00

416

71.96

71.90

8.42

47.20

95.00

421

72.33

73.20

8.37

48.60

95.00

415

72.99

73.20

9.04

48.60

93.40

400

79.30

80.60

9.05

52.00

95.40

attributed to an increased focus on corporate governance and risk management. An index score in the range of 70–80% is indicative of most parameters being in the range of 3.5–4 (out of 5). In other words, on an average, Indian companies have a near ideal index, based on the normative framework developed above. Further, an increase in minimum index levels is also indicative of growing consciousness on the part of Indian companies with respect to the importance of risk and risk governance. In addition, the maximum risk index exhibited in each of the ten years is

5 Normative Framework for Risk Governance Index …

180 120.00% 100.00% 80.00%

Mean

60.00%

Median

40.00%

Std. DeviaƟon

20.00%

Minimum Maximum

0.00%

Fig. 5.2 Descriptive statistics of risk governance index, 2005–2015

above 90%. This clearly indicates the importance (that at least) some companies have been giving to risk governance structures. In addition, a low standard deviation in the range of 8–9% is suggestive of somewhat similar structures in the majority of companies (Fig. 5.2). In sum, the Indian corporate sector appears to be conscious of the benefits of strong governance structure. Ultimately, it is the governance structure and mechanism that will enable companies to manage risks, endure difficulties and leverage the opportunities.

5.4.1.2

Phase-Wise Analysis

As is evident from Table 5.2 the mean governance index has risen from approximately 69% in the pre-recession period and close to 73% in the post-recession period. Whether the increase is significant or not has been tested using the following hypothesis: Hypothesis: Recession has an impact on risk governance structure of companies Mann–Whitney U test has been applied to test the hypothesis. The statistically significant improvement (Table 5.3) in governance structure can be attributed to increased awareness towards the need for more effective and efficient governance. Also, the period around the years 2005–2006 was a transitory period for governance

Table 5.2 Descriptive statistics of risk governance index, phase-wise (pre-recession (2005–2008); post-recession (2008–2015))

Particulars

Pre-recession

Post-recession

N Mean (%) Std. deviation (%)

1163 68.99 8.71

2870 72.84 9.01

5.4 Empirical Evidence

181

Table 5.3 Mann–Whitney U test for difference of mean risk governance index (pre-recession, post-recession) Particulars

Pre-recession (2005–2008)

N Mean rank Sum of ranks

1163 1687.66 1,962,750.00

Mann–Whitney U Wilcoxon W Z Asymp. sig. (2-tailed) Note *** Significant at 1% level of significance

Post-recession (2008–2015)

2870 2150.46 6,171,811.00 Risk governance index 1,285,884.000 1,962,750.000 −11.437 0.000***

legislations and hence, the true effects may have been realised in subsequent periods. Further, several studies have shown that good governance results in better firm performance. Therefore, firms may be trying to signal that strong governance will act as a catalyst for strong future performance. In addition, post-recession period was beset with dwindling investor confidence; therefore, by committing to higher standards of governance, firms may have been trying to showcase their will and ability to endure tough times.

5.4.1.3

Age-Wise Analysis

Table 5.4 shows that amongst the three age groups, old and established companies (falling in Q4) have led the pack with the highest average governance score of 73% during the sample period. But, young and middle-aged companies are not far behind with average governance scores of 71% and 72%, respectively. Accordingly, the next hypothesis is as follows: Hypothesis: Age has an impact on risk governance structure Kruskal Wallis test (Table 5.5) indicates that there is a statistically significant difference between governance levels of young and old and established companies, Table 5.4 Descriptive statistics of risk governance index, age-wise (2005–2015) Age class

Mean (%)

Median (%)

Std. deviation (%)

Minimum (%)

Maximum (%)

Skewness

Kurtosis

Young (Q1) Middle-aged (Q2 and Q3) Old and established (Q4)

70.80 71.67

71.20 71.60

8.93 8.94

41.60 36.00

95.00 95.40

−0.11 −0.07

−0.06 −0.17

72.85

73.20

9.44

44.00

95.40

−0.21

−0.15

182 Table 5.5 Kruskal–Wallis test for mean difference in risk governance index among different age groups

5 Normative Framework for Risk Governance Index … Age class

N

Mean rank

Young 1072 1902.44 Middle-aged 1973 2002.44 Old and established 988 2170.38 Total 4033 Test statistics Chi-square 27.843 df 2 Asymp. sig. 0.000*** Note *** Significant at 1% level of significance

and middle-aged and old and established companies (with the governance quality of old companies being statistically significantly higher in both the cases). The fact that older companies are better governed may be attributed to the fact that as firms get older, they learn how to do things better (Jovanovic 1982). Further, the results are in contrast to the results of studies that show that older firms exhibit rent-seeking behaviour and poorer corporate governance (Loderer and Waelchli 2011). In addition, younger firms may be expected to be filled with zeal and vigour, thus, desiring less governance (Fig. 5.3). Further, a phase-wise analysis for all three age groups (Tables 5.6 and 5.7) reveals, that there has been a statistically significant increase in risk governance of companies, in all three age groups, in the post-recession period. This suggests that recession had a learning impact on Indian companies, notwithstanding their age. The learning is with regard to the importance of governance structure as a source of garnering investor confidence and more importantly, as a fundamental force behind effective risk management.

Fig. 5.3 Pair-wise comparison of risk governance index of companies based on age

5.4 Empirical Evidence

183

Table 5.6 Descriptive statistics of risk governance index, age-wise (pre-recession (2005–2008); post-recession (2008–2015)) Particulars

Time frame

Young

Middle-aged

Old and established

N

Total Pre-recession Post-recession Total Pre-recession Post-recession Total Pre-recession Post-recession

1072 301 771 70.80 67.56 72.07 8.93 8.27 8.86

1973 579 1394 71.67 68.90 72.82 8.94 8.62 8.82

988 283 705 72.85 70.70 73.72 9.44 9.08 9.45

Mean (%)

Std. deviation (%)

Table 5.7 Mann–Whitney U test for difference of mean risk governance index (pre-recession, post-recession) for each age category Particulars

Phase

Young

Middle-aged

Old and established

N

Pre-recession Post-recession Pre-recession Post-recession Pre-recession Post-recession

301 771 425.39 579.88 128,041.50 447,086.50 82,590.500 128,041.500 −7.345 0.000***

579 1394 828.27 1052.93 479,569.50 1,467,781.50 311,659.500 479,569.500 −7.978 0.000***

283 705 432.12 519.54 122,289.50 366,276.50 82,103.500 122,289.500 −4.355 0.000***

Mean rank Sum of ranks

Mann–Whitney U Wilcoxon W Z Asymp. sig. (2-tailed) Note *** Significant at 1% level of significance

5.4.1.4

Industry-Wise Analysis

It is evident from Table 5.8 that FMCG, power, diversified and ICT industry have maximum governance with an average of more than 73%; in contrast, media and healthcare industry have minimum risk governance index with an average of about 68%. FMCG and ICT industry may be viewed as the most competitive industries both in terms of product and labour markets. Therefore, in order to cope with unforeseen uncertainties and to keep abreast with trends of the industry, the companies in these industries may be expected to be proactive managers of risk; and in order to have effective risk management, a robust risk governance structure is pertinent. Similarly, the power industry is highly regulated, and to gain legitimacy and to

5 Normative Framework for Risk Governance Index …

184

Table 5.8 Descriptive statistics of risk governance index, industry-wise (2005–2015)

Agriculture Capital goods Chemical Diversified FMCG Health care Housing and construction ICT telecom Media Metals Miscellaneous Oil and gas Power Textile Transport

Mean (%)

Median (%)

Std. deviation (%)

Minimum (%)

Maximum (%)

72.84 70.24 72.15 73.33 73.63 69.61 71.01

74.20 71.20 71.70 74.60 74.20 71.20 72.30

9.22 10.86 8.87 8.69 8.33 9.07 7.91

55.20 45.60 45.60 51.60 48.60 47.20 36.00

90.60 95.00 93.40 95.40 92.20 92.20 93.40

73.36 67.67 72.65 70.89 71.35 73.40 70.48 72.43

73.20 66.80 73.00 71.20 71.20 73.20 70.90 71.60

8.68 9.35 9.80 8.97 9.67 9.41 8.80 8.58

45.60 41.60 51.60 45.60 44.00 45.60 51.60 49.60

95.40 91.80 95.00 93.40 93.40 93.80 87.60 92.20

ensure continued access to essential resources (resource dependence theory), companies in this industry may be expected to exhibit strong governance structures. However, lower levels of governance in media and healthcare industry remain a puzzle. Schmalensee (1985) and Wernerfelt and Montgomery (1988) conjecture that industry effects have a strong influence on the firm’s success. This leads to the next hypothesis: Hypothesis: Industry-sector has an impact on risk governance structure of companies A pair-wise comparison of mean risk governance levels in industries (Fig. 5.4) reveals that healthcare industry has statistically significant different risk governance than six other industries (namely, FMCG, ICT, power, diversified, agriculture and transport). Similarly, the media industry has risk levels, which are significantly different from that of eight other industries. These may be attributed to the peculiar nature of these industries. Amongst all industry groups, healthcare is one of the most research focussed and regulated industry. In sharp contrast, the media industry has been observed as the one with minimum constraints and regulations. It is worth mentioning that risk is an all-pervasive and material aspect of the business; therefore, it appears reasonable to assume that all entities would have appropriate risk management governance structures. In view of the above statement, the fact these industries have substantially different levels of risk vis-à-vis other industries, is surprising.

5.4 Empirical Evidence

185

Fig. 5.4 Pair-wise comparison of risk governance index of companies based on industry

It is noteworthy, that a phase-wise analysis (Tables 5.9 and 5.10) indicates that only 3 (namely, diversified, media and textile) out of 15 industries did not witness a statistically significant increase in risk governance index in the post-recession period. The remaining industries have demonstrated greater commitment to risk governance (in the post-recession period), statistically significant.

5.4.2

Dis-aggregative Analysis

As is evident from aggregative analysis that Indian companies are scoring an average of about 3.5–4 (out of 5) on each of the 11 governance characteristics, it would be useful to investigate whether any of the characteristics are dominating this scenario. Therefore, an attempt has been made to gauge the frequencies of companies for each of the scores, 1, 2, 3, 4 and 5, for each of the 11 governance characteristics.

5.4.2.1

Full-Sample and Phase-Wise Analysis

About half the firm-year observations have optimum Board size, i.e. eight–ten directors (Table 5.11). This is indicative of consciousness on the part of Indian companies regarding the disadvantages of large unmanageable Boards as well as perils of smaller/concentrated Boards. Further, there are no significant differences in board size in pre- and post-recession periods. It is noteworthy that the legal requirement of minimum 50% non-executive Directors has been violated only in isolated cases (2.3%) (Table 5.11). Equally important to note is that in over 90% of observations more than half the Boards

Time frame

Pre-recession

Pre-recession

Post-recession

Post-recession

Particulars

Mean (%)

Mean (%)

Standard deviation (%)

Standard deviation (%)

9.69

7.31

74.07

69.93

Agriculture

11.12

9.50

71.46

67.08

Capital goods

8.84

8.53

73.13

69.84

Chemical

8.84

8.53

73.13

69.84

Diversified

8.32

7.67

74.80

70.76

FMCG

9.16

8.29

70.69

66.94

Healthcare

7.65

7.68

72.27

67.78

Housing and construction

8.41

8.72

74.49

70.44

ICT

8.89

10.24

68.58

65.08

Media

10.10

8.53

73.81

69.96

Metal

8.66

9.10

72.04

68.05

Miscellaneous

9.63

8.72

72.91

67.52

Oil and gas

8.03

11.77

74.50

70.76

Power

8.94

8.46

70.94

69.39

Textile

Table 5.9 Descriptive statistics of risk governance index, industry-wise, phase-wise (pre-recession (2005–2008); post-recession (2008–2015))

8.50

7.99

73.70

69.29

Transport

186 5 Normative Framework for Risk Governance Index …

Pre-recession Post-recession Pre-recession Post-recession Pre-recession Post-recession

N N Mean rank Mean rank Sum of ranks Sum of ranks Mann–Whitney U Wilcoxon W Z Asymp. Sig. (2-tailed) Particulars

N Pre-recession 35 N Post-recession 100 Mean rank Pre-recession 59.41 Mean rank Post-recession 71.01 Sum of ranks Pre-recession 2079.50 Sum of ranks Post-recession 7100.50 Mann–Whitney U 1449.500 Wilcoxon W 2079.500 Z −1.510 Asymp. Sig. (2-tailed) 0.131 Note ***,* Significant at 1% and 10% level of significance,

78 181 110.96 138.20 8655.00 25,015.00 5574.000 8655.000 −2.686 0.007*** respectively

51 118 78.50 87.81 4003.50 10,361.50 2677.500 4003.500 −1.136 0.256

Diversified

120 295 175.19 221.35 21,023.00 65,297.00 13,763.000 21,023.000 −3.555 0.000***

Miscellaneous

98 230 142.26 173.98 13,941.00 40,015.00 9090.000 13,941.000 −2.774 0.006***

Chemical

Metal

70 180 105.15 133.41 7360.50 24,014.50 4875.500 7360.500 −2.776 0.006***

Capital goods

Media

50 118 66.49 92.13 3324.50 10,871.50 2049.500 3324.500 −3.128 0.002***

Agriculture

Time frame

Time frame

Particulars

47 115 64.85 88.30 3048.00 10,155.00 1920.000 3048.000 −2.890 0.004***

47 111 74.71 81.53 3511.50 9049.50 2383.500 3511.500 −.857 0.391

Textile

135 349 186.07 264.33 25,119.50 92,250.50 15,939.500 25,119.500 −5.522 0.000***

Housing and construction

60 145 90.95 107.99 5457.00 15,658.00 3627.000 5457.000 −1.873 0.061*

Power

89 220 128.17 165.85 11,407.50 36,487.50 7402.500 11,407.500 −3.358 0.001***

Healthcare

Oil and gas

84 204 115.90 156.27 9736.00 31,880.00 6166.000 9736.000 −3.741 0.000***

FMCG

106 262 146.94 199.69 15,576.00 52,320.00 9905.000 15,576.000 −4.315 0.000***

Transport

93 242 138.31 179.41 12,863.00 43,417.00 8492.000 12,863.000 −3.480 0.001***

ICT

Table 5.10 Mann Whitney U test for difference of mean risk governance index, industry-wise, phase-wise (pre-recession (2005–2008); post-recession (2008– 2015))

5.4 Empirical Evidence 187

5 Normative Framework for Risk Governance Index …

188

Table 5.11 Summary of frequency distribution of risk governance characteristics, phase-wise (pre-recession (2005–2008); post-recession (2008–2015)) Particulars Governance characteristics Size of the board

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

Time frame Score

Total

Pre-recession

Post-recession

Percent

Cumulative percent

Percent

Cumulative percent

Percent

Cumulative percent

1 2 3 4 5 1 3 5 3 5 3 5

0.0 0.1 3.7 52.0 44.2 2.3 7.4 90.3 36.4 63.6 48.0 52.0

0.0 0.1 3.8 55.8 100.0 2.3 9.7 100.0 36.4 100.0 48.0 100.0

0.0 0.0 3.4 52.5 44.1 2.8 8.0 89.3 36.2 63.8 47.2 52.8

0.0 0.0 3.4 55.9 100.0 2.8 10.7 100.0 36.2 100.0 47.2 100.0

0.0 0.2 3.8 51.8 44.3 2.1 7.2 90.7 36.5 63.5 48.3 51.7

0.0 0.2 4.0 55.7 100.0 2.1 9.3 100.0 36.5 100.0 48.3 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5

60.1 39.8 0.0 0.0 12.4 16.2 71.3 4.4 95.6 86.4 13.6

60.1 100.0 100.0 100.0 12.4 28.7 100.0 4.4 100.0 86.4 100.0

70.0 29.8 0.2 0.0 15.6 15.0 69.4 5.8 94.2 95.2 4.8

70.0 99.8 100.0 100.0 15.6 30.6 100.0 5.8 100.0 95.2 100.0

56.1 43.9 0.0 0.0 11.2 16.7 72.1 3.8 96.2 82.8 17.2

56.1 100.0 100.0 100.0 11.2 27.9 100.0 3.8 100.0 82.8 100.0

1 5 1 5 1 1.8 2.6 3.4 4.2 5

96.4 3.6 65.0 35.0 2.1 24.4 45.7 17.1 8.5 2.3

96.4 100.0 65.0 100.0 2.1 26.5 72.2 89.2 97.7 100.0

97.1 2.9 78.3 21.7 2.9 27.9 48.5 14.8 4.7 1.2

97.1 100.0 78.3 100.0 2.9 30.8 79.3 94.1 98.8 100.0

96.1 3.9 59.5 40.5 1.7 23.1 44.5 18.0 10.0 2.7

96.1 100.0 59.5 100.0 1.7 24.8 69.3 87.3 97.3 100.0

consist of non-executive directors. This would ensure unbiased and fair perspective on matters of concern. It is worth mentioning that findings are similar in both pre-recession and post-recession phase.

5.4 Empirical Evidence

189

As per Table 5.11 in about two-thirds of the observations, there is a separation of the role of CEO and Board Chairman. This implies that majority of Indian companies subscribe to the view that non-duality promotes Board independence, ensures effective oversight and facilitates separation of decision management and decision control (Lorsch and MacIver 1989; Millstein 1992). Interestingly, almost the same proportion of companies had CEO duality in both phases. Surprisingly, the number of observations having an executive Chairman is similar to that of having a non-executive Chairman (Table 5.11). This finding could be attributed to the fact that the suitability of the candidate for the position of Chairman is being judged on the basis of his expertise, knowledge, experience and vision rather than his executive or non-executive status. It is noteworthy, that considering the entire period of study; more than 60% of the companies did not have even a single woman director on their board. In light of the fact, that the legal requirement of minimum one woman director was not in force during the period of study, this newly inserted clause appears to be a much-needed step to ensure gender diversity on Boards. As a result, the phase-wise analysis (Table 5.11) reveals that the proportion of companies appointing at least one woman director has increased from 30 to close to 44% from the pre-recession phase to the post-recession phase. Though about three-fourths of the observations have a majority of independent directors, the fact that in about one-eighth of cases, the legal requirement regarding a minimum number of independent directors has not been complied with is startling (Table 5.11). The reason being that the sample companies represent more than 90% of India’s corporates, in terms of market capitalisation. Therefore, such findings raise a concern about the commitment of companies towards the principles of fairness and transparency. Further, a phase-wise analysis does not reveal any significant difference with regard to board independence. In tune with the intuition, a vast majority (96%) of firm-year observations had all mandatory committees in place. But, the fact that close to 4% observations did not constitute at least one of the mandatory committees raises serious concern about their willingness and ability to provide good governance. It is incidence like these that may be expected to lead to loss of motivation on part of employees, resulting in poor performance of the company and eventually its winding up. The continuing evidence of such trends, despite the attention around good governance is a cause of concern. Further, the existence of risk management committee has been observed in only 14% firm-year observations. Similarly, the appointment of a CRO has been observed in only 4% of firm-year observations. These two facts indicate the laxity that is prevalent in risk governance structures in Indian companies. Given the findings in Table 6.11 the recently enforced rules and regulations in this regard appear to be the need of the hour. However, an increase of 12% points (from Phase I to Phase II) in context of companies having a risk management committee is indicative of an increasing desire, for effective risk management, on the part of Indian companies.

5 Normative Framework for Risk Governance Index …

190

In addition only about 35% observations had a whistle-blower policy (Table 5.11). This may be due to two reasons: one, the companies are confident of the ethical conduct of their employees and thus believe that no such mechanism is required, or second, the management of the company is insecure that the irregularities that they may be involved in will be exposed. Further, the fact that the proportion of companies that have put in place a whistle-blower policy has doubled from phase I to phase II is indicative of the efforts that Indian companies are putting into coming across as fair, transparent and accountable organisations. In the context of voluntary committees (Table 5.11), almost half of the companies have constituted at least two voluntary committees. Most of these are related to financial decisions and human resource issues. The results are, by and large, similar for both the phases.

5.4.2.2

Age-Wise Analysis

To gain a better insight into prominent aspects of risk governance structure, an age-wise analysis has been attempted (Table 5.12). It is noteworthy that all the middle-aged companies had more than three directors on their board. Further, more than two-fifths of the observations in all the age groups had an optimal board size, i.e. eight to ten directors. This implies that Indian companies are aware of the pros and cons of smaller as well as excessively larger boards, notwithstanding their age. Interestingly, almost 90% observations in all three age groups had more than 50% of the directors belonging to the class of non-executive directors. This clearly indicates that Indian companies acknowledge the advantages of non-executive directors. Non-executive directors not only bring independence, but also the experience, knowledge and fairness in dealings. It is noteworthy that about two-thirds of the firm-year observations in all age groups have separate persons for the position of CEO and Chairman. CEO duality has often been considered a primary cause of the decline of major US giants like General Motors and IBM (White and Ingrassia 1992). Therefore, evidence of non-duality in Indian context assures stakeholders of distributed authority resulting in effective and efficient risk monitoring. About half of young as well as middle-aged companies have executive Chairman. Whereas, close to three-fifths of the old and established companies have a non-executive Chairman. Since, the Chairman can be seen as a central figure in decision hierarchy, his independence will always be appreciated. As old and established companies have the relevant and intense experience, they may have recognised the benefits of having non-executive chairman, better than their younger counterparts. In a startling finding, close to two-thirds of the young companies did not have even a single women director on their board. The finding is revealing as young companies are expected to be more conscious of their social responsibilities. They are expected to have fair and balanced boards with representations from all

5.4 Empirical Evidence

191

Table 5.12 Summary of frequency distribution of risk governance characteristics, age-wise Particulars Governance characteristics Size of the board

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

Age class Score

Young companies Cumulative percent

Middle-aged companies Percent Cumulative percent

Old and established companies Percent Cumulative percent

Percent

1 2 3 4 5 1 3 5 3 5 3 5

0.0 0.1 3.7 53.0 43.2 1.5 8.5 90.0 38.4 61.6 50.0 50.0

0.0 0.1 3.8 56.8 100.0 1.5 10.0 100.0 38.4 100.0 50.0 100.0

0.0 0.0 3.5 52.2 44.2 3.0 7.5 89.5 38.2 61.8 50.1 49.9

0.0 0.0 3.5 55.8 100.0 3.0 10.5 100.0 38.2 100.0 50.1 100.0

0.0 0.4 3.9 50.4 45.2 1.6 6.2 92.2 30.7 69.3 41.5 58.5

0.0 0.4 4.4 54.8 100.0 1.6 7.8 100.0 30.7 100.0 41.5 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5

65.0 35.0 0.0 0.0 8.8 22.4 68.8 6.9 93.1 90.1 9.9

65.0 100.0 100.0 100.0 8.8 31.2 100.0 6.9 100.0 90.1 100.0

58.4 41.5 0.1

58.4 99.9 100.0

58.2 41.8 0.0

58.2 100.0 100.0

14.3 14.9 70.8 2.9 97.1 86.1 13.9

14.3 29.2 100.0 2.9 100.0 86.1 100.0

12.7 12.3 75.0 4.7 95.3 83.0 17.0

12.7 25.0 100.0 4.7 100.0 83.0 100.0

1 5 1 5 1 1.8 2.6 3.4 4.2 5

97.5 2.5 66.4 33.6 1.3 19.7 54.0 14.8 8.3 1.9

97.5 100.0 66.4 100.0 1.3 21.0 75.0 89.8 98.1 100.0

96.2 3.8 64.3 35.7 2.5 26.8 41.9 17.3 8.7 2.8

96.2 100.0 64.3 100.0 2.5 29.3 71.2 88.5 97.2 100.0

95.5 4.5 64.7 35.3 1.9 24.9 44.2 19.0 8.4 1.5

95.5 100.0 64.7 100.0 1.9 26.8 71.1 90.1 98.5 100.0

stakeholders. In contrast, only about half of the companies in the middle-aged and old group have no female representation on Board. Campbell and Vera (2008) evidence that gender diversity in boards may generate higher economic gains. Therefore, Indian companies need to take steps to promote gender diversity in organisations, in general, and on boards, in particular.

192

5 Normative Framework for Risk Governance Index …

Majority of Indian companies, notwithstanding their age, have more than 50% independent directors. Old and established companies lead the pack with more than three-fourths of the companies having independent boards. This clearly indicates the commitment of Indian companies towards independent, fair and accountable governance. Though more than 90% of the observations in all age groups have constituted all the mandatory committees, the fact that close to 7% of the observations in the category of young companies fail to do so, is startling. In light of legitimacy theory and resource dependence theory, younger companies are expected to have complete compliance with legal and statutory requirements. Despite increasing attention on risk management, more than four-fifths of the observations in all the three age groups have not constituted a risk management committee. Interestingly, the maximum proportion of companies having a risk management committee has been found in the old and established group. Apparently, these companies believe in the power of having a formal structure for more effective risk management. It is pertinent to note that about in 95% of the observations in all three age groups did not have a CRO. The findings are noteworthy in view of the fact that the period under study witnessed a crisis, increasing uncertainty in a complex business environment; yet, such large number of companies undermines the importance of having a CRO. Almost one-third of the observations in all age groups have a defined whistle-blower policy in place. This shows their commitment to ensure a corruption-free, fair and impartial work culture. Such mechanisms play a pivotal role in curbing risks related to embezzlement and fraud. It is noteworthy that at least 70% observations in all groups have voluntary committees belonging to at least two categories. This highlights companies’ vigilance and agility in the context of potential risks in various areas.

5.4.2.3

Industry-Wise Analysis

In terms of board size, all the companies in all the industries with the exception of the capital goods industry had at least three directors on the board (Table 5.13). Further, more than two-fifths of the observations in all the industries had an ideal board size, i.e. 8–10 directors. This clearly indicates that across industries, Indian companies avoid large/unmanageable boards as well as smaller concentrated Boards. On the one hand, all the industries, with the exception of 8 of them, had at least one company that failed to comply with the legal requirement regarding non-executive Director; on the other hand, more than four-fifths of the companies in all the industries had more than half their Board comprising of non-executive directors (Table 5.13). This can be viewed as a step towards curbing agency problems and conflict of interests.

5.4 Empirical Evidence

193

Table 5.13 Summary of the frequency distribution of risk governance characteristics, industry-wise Particulars Governance characteristics

Industry Score

Agriculture Per Cumulative cent per cent

Capital goods Per Cumulative cent per cent

Chemical Per Cumulative cent per cent

Size of the board

1 2 3 4 5 1 3 5 3 5 3 5

0.0 0.0 0.0 53.6 46.4 2.4 8.9 88.7 23.8 76.2 29.2 70.8

0.0 0.0 0.0 53.6 100.0 2.4 11.3 100.0 23.8 100.0 29.2 100.0

0.0 1.6 4.8 44.4 49.2 3.2 5.2 91.6 40.4 59.6 50.0 50.0

0.0 1.6 6.4 50.8 100.0 3.2 8.4 100.0 40.4 100.0 50.0 100.0

0.0 0.0 2.4 50.9 46.6 1.2 5.2 93.6 25.9 74.1 32.6 67.4

0.0 0.0 2.4 53.4 100.0 1.2 6.4 100.0 25.9 100.0 32.6 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5 1 5 1 5 1 1.8 2.6 3.4 4.2 5

60.7 39.3 0.0 0.0 8.9 11.3 79.8 0.6 99.4 86.3 13.7 93.5 6.5 75.6 24.4 0.0 25.6 62.5 8.3 3.6 0.0

60.7 100.0 100.0 100.0 8.9 20.2 100.0 0.6 100.0 86.3 100.0 93.5 100.0 75.6 100.0 0.0 25.6 88.1 96.4 100.0 100.0

62.0 38.0 0.0 0.0 20.0 18.4 61.6 10.8 89.2 84.4 15.6 99.2 0.8 60.0 40.0 4.8 28.4 50.8 8.0 4.0 4.0

62.0 100.0 100.0 100.0 20.0 38.4 100.0 10.8 100.0 84.4 100.0 99.2 100.0 60.0 100.0 4.8 33.2 84.0 92.0 96.0 100.0

58.2 41.8 0.0 0.0 12.8 8.2 79.0 3.7 96.3 91.8 8.2 98.8 1.2 73.5 26.5 0.3 36.0 32.0 15.2 14.0 2.4

58.2 100.0 100.0 100.0 12.8 21.0 100.0 3.7 100.0 91.8 100.0 98.8 100.0 73.5 100.0 0.3 36.3 68.3 83.5 97.6 100.0 (continued)

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

5 Normative Framework for Risk Governance Index …

194 Table 5.13 (continued) Particulars Governance characteristics Particulars Governance characteristics Size of the board

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

Industry Score

Diversified Per Cumulative cent per cent

FMCG Per Cumulative cent per cent

Healthcare Per Cumulative cent per cent

Industry Score 1 2 3 4 5 1 3 5 3 5 3 5

Diversified Per Cumulative cent per cent 0.0 0.0 0.0 0.0 5.9 5.9 53.8 59.8 40.2 100.0 0.6 0.6 3.0 3.6 96.4 100.0 31.4 31.4 68.6 100.0 46.7 46.7 53.3 100.0

FMCG Per Cumulative cent per cent 0.0 0.0 0.0 0.0 0.7 0.7 45.5 46.2 53.8 100.0 1.7 1.7 10.4 12.2 87.8 100.0 37.5 37.5 62.5 100.0 44.4 44.4 55.6 100.0

Healthcare Per Cumulative cent per cent 0.0 0.0 0.0 0.0 2.9 2.9 46.3 49.2 50.8 100.0 1.3 1.3 15.9 17.2 82.8 100.0 55.7 55.7 44.3 100.0 71.2 71.2 28.8 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5 1 5 1 5 1 1.8 2.6 3.4 4.2 5

57.4 42.6 0.0 0.0 13.0 13.6 73.4 6.5 93.5 73.4 26.6 89.3 10.7 71.0 29.0 5.3 18.9 42.0 21.9 11.2 0.6

56.3 43.8 0.0 0.0 5.6 20.1 74.3 3.5 96.5 92.0 8.0 93.4 6.6 50.7 49.3 0.3 21.2 48.3 22.2 5.6 2.4

58.3 41.7 0.0 0.0 7.4 20.4 72.2 4.5 95.5 82.2 17.8 97.7 2.3 71.8 28.2 2.9 30.1 45.6 11.3 10.0 0.0

57.4 100.0 100.0 100.0 13.0 26.6 100.0 6.5 100.0 73.4 100.0 89.3 100.0 71.0 100.0 5.3 24.3 66.3 88.2 99.4 100.0

56.3 100.0 100.0 100.0 5.6 25.7 100.0 3.5 100.0 92.0 100.0 93.4 100.0 50.7 100.0 0.3 21.5 69.8 92.0 97.6 100.0

58.3 100.0 100.0 100.0 7.4 27.8 100.0 4.5 100.0 82.2 100.0 97.7 100.0 71.8 100.0 2.9 33.0 78.6 90.0 100.0 100.0 (continued)

5.4 Empirical Evidence

195

Table 5.13 (continued) Particulars

Industry

Governance characteristics Particulars

Score

Governance characteristics Size of the board

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

Housing and construction Per Cumulative cent per cent

ICT

ICT

1 2 3 4 5 1 3 5 3 5 3 5

Housing and construction Per Cumulative cent per cent 0.0 0.0 0.0 0.0 3.3 3.3 48.8 52.1 47.9 100.0 0.8 0.8 11.0 11.8 88.2 100.0 42.8 42.8 57.2 100.0 64.5 64.5 35.5 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5

55.4 44.6 0.0 0.0 9.3 27.3 63.4 1.4 98.6 90.5 9.5

55.4 100.0 100.0 100.0 9.3 36.6 100.0 1.4 100.0 90.5 100.0

60.0 40.0 0.0 0.0 5.4 13.4 81.2 6.0 94.0 87.2 12.8

60.0 100.0 100.0 100.0 5.4 18.8 100.0 6.0 100.0 87.2 100.0

60.7 39.3 0.0 0.0 16.3 23.7 60.0 4.4 95.6 97.8 2.2

1 5 1 5 1 1.8 2.6 3.4 4.2 5

96.7 3.3 65.3 34.7 0.8 14.3 51.2 21.1 10.1 2.5

96.7 100.0 65.3 100.0 0.8 15.1 66.3 87.4 97.5 100.0

100.0 0.0 52.2 47.8 3.0 24.2 44.5 19.4 6.0 3.0

100.0 100.0 52.2 100.0 3.0 27.2 71.6 91.0 97.0 100.0

100.0 0.0 74.1 25.9 1.5 19.3 46.7 21.5 11.1 0.0

Industry Score

Per cent

Per cent 0.0 0.0 2.4 60.6 37.0 0.0 1.8 98.2 32.5 67.5 48.1 51.9

Media Cumulative per cent

Per cent

Cumulative per cent

Media Cumulative per cent 0.0 0.0 2.4 63.0 100.0 0.0 1.8 100.0 32.5 100.0 48.1 100.0

Per cent 0.0 0.7 7.4 61.5 30.4 5.2 9.6 85.2 45.2 54.8 63.0 37.0

Cumulative per cent 0.0 0.7 8.1 69.6 100.0 5.2 14.8 100.0 45.2 100.0 63.0 100.0 60.7 100.0 100.0 100.0 16.3 40.0 100.0 4.4 100.0 97.8 100.0 100.0 100.0 74.1 100.0 1.5 20.7 67.4 88.9 100.0 100.0 (continued)

5 Normative Framework for Risk Governance Index …

196 Table 5.13 (continued) Particulars Governance characteristics Particulars Governance characteristics Size of the board

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

Industry Score

Metal Per Cumulative cent per cent

Miscellaneous Per Cumulative cent per cent

Oil Per cent

Cumulative per cent

Industry Score 1 2 3 4 5 1 3 5 3 5 3 5

Metal Per cent 0.0 0.0 2.7 60.6 36.7 2.7 7.7 89.6 31.3 68.7 44.0 56.0

Cumulative per cent 0.0 0.0 2.7 63.3 100.0 2.7 10.4 100.0 31.3 100.0 44.0 100.0

Miscellaneous Per Cumulative cent per cent 0.0 0.0 0.0 0.0 1.9 1.9 52.0 54.0 46.0 100.0 4.1 4.1 9.2 13.3 86.7 100.0 37.3 37.3 62.7 100.0 46.0 46.0 54.0 100.0

Oil Per cent 0.0 0.0 9.3 53.1 37.7 7.4 8.6 84.0 35.8 64.2 35.8 64.2

Cumulative per cent 0.0 0.0 9.3 62.3 100.0 7.4 16.0 100.0 35.8 100.0 35.8 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5 1 5 1 5 1 1.8 2.6 3.4 4.2 5

46.3 53.7 0.0 0.0 18.9 18.5 62.5 2.7 97.3 88.4 11.6 90.0 10.0 59.1 40.9 1.2 37.1 30.5 15.4 13.1 2.7

46.3 100.0 100.0 100.0 18.9 37.5 100.0 2.7 100.0 88.4 100.0 90.0 100.0 59.1 100.0 1.2 38.2 68.7 84.2 97.3 100.0

58.3 41.7 0.0 0.0 14.2 16.9 68.9 5.8 94.2 83.1 16.9 98.8 1.2 72.3 27.7 5.1 17.8 46.0 23.9 5.3 1.9

66.7 33.3 0.0 0.0 37.0 3.7 59.3 7.4 92.6 82.7 17.3 99.4 0.6 51.2 48.8 0.6 19.8 39.5 21.0 10.5 8.6

58.3 100.0 100.0 100.0 14.2 31.1 100.0 5.8 100.0 83.1 100.0 98.8 100.0 72.3 100.0 5.1 22.9 68.9 92.8 98.1 100.0

66.7 100.0 100.0 100.0 37.0 40.7 100.0 7.4 100.0 82.7 100.0 99.4 100.0 51.2 100.0 0.6 20.4 59.9 80.9 91.4 100.0 (continued)

5.4 Empirical Evidence

197

Table 5.13 (continued) Particulars Governance characteristics Particulars Governance characteristics Size of the board

Proportion of non-executive directors CEO duality Executive/ non-executive chairman Proportion of women on board

Proportion of independent directors Compulsory committees Risk management committee CRO Whistle-blower policy Voluntary committees

Industry Score

Power Per Cumulative cent per cent

Textile Per Cumulative cent per cent

Transport Per Cumulative cent per cent

Industry Score 1 2 3 4 5 1 3 5 3 5 3 5

Power Per cent 0.0 0.0 8.3 48.8 42.9 4.9 2.9 92.2 32.7 67.3 37.1 62.9

Cumulative per cent 0.0 0.0 8.3 57.1 100.0 4.9 7.8 100.0 32.7 100.0 37.1 100.0

Textile Per cent 0.0 0.0 3.8 55.7 40.5 3.2 3.8 93.0 43.0 57.0 51.3 48.7

Cumulative per cent 0.0 0.0 3.8 59.5 100.0 3.2 7.0 100.0 43.0 100.0 51.3 100.0

Transport Per Cumulative cent per cent 0.0 0.0 0.0 0.0 5.7 5.7 52.7 58.4 41.6 100.0 0.8 0.8 4.1 4.9 95.1 100.0 28.3 28.3 71.7 100.0 40.2 40.2 59.8 100.0

1 2.5 3.5 5 1 3 5 1 5 1 5

71.2 28.8 0.0 0.0 16.6 9.3 74.1 0.0 100.0 76.1 23.9

71.2 100.0 100.0 100.0 16.6 25.9 100.0 0.0 100.0 76.1 100.0

68.4 30.4 1.3 0.0 7.0 19.6 73.4 8.2 91.8 88.0 12.0

68.4 98.7 100.0 100.0 7.0 26.6 100.0 8.2 100.0 88.0 100.0

71.5 28.5 0.0 0.0 9.8 9.8 80.4 3.5 96.5 86.7 13.3

71.5 100.0 100.0 100.0 9.8 19.6 100.0 3.5 100.0 86.7 100.0

1 5 1 5 1 1.8 2.6 3.4 4.2 5

91.2 8.8 71.7 28.3 0.5 19.0 37.1 24.4 13.7 5.4

91.2 100.0 71.7 100.0 0.5 19.5 56.6 81.0 94.6 100.0

100.0 0.0 62.7 37.3 5.1 27.2 50.6 10.8 5.1 1.3

100.0 100.0 62.7 100.0 5.1 32.3 82.9 93.7 98.7 100.0

95.1 4.9 65.5 34.5 0.3 29.3 55.4 8.7 6.0 0.3

95.1 100.0 65.5 100.0 0.3 29.6 85.1 93.8 99.7 100.0

198

5 Normative Framework for Risk Governance Index …

Close to three-fifth of the observations in all the industries (Table 5.13) have separation of roles relating to CEO and Chairman of Board. It is only in the case of 6 industries that the proportion of companies having CEO duality is more than that of companies without CEO duality. CEO duality is often associated with abuse of managerial discretion (Balinga et al. 1996). Therefore, evidence of persisting non-duality in Indian corporates may be viewed as a mechanism to control operational risks. Majority of the companies in two-thirds of industries (namely agriculture, chemical, diversified, FMCG, ICT, metal, miscellaneous, oil, power and transport) have a non-executive Chairman (Table 5.13). This is indicative of the fact that Indian companies, by and large, desire independence and tend to avoid agency costs. Also, the objectivity, fairness and transparency brought about by the presence of a non-executive Chairman, have well been recognised by the Indian companies. The absence of women directors in more than 50% observations across almost all the industries is a cause of concern. The findings are a clear reflection of the patriarchal societal structure prevalent in Indian society. Brammer and Pavelin (2007) suggest that a higher representation of women is usually observed in industries that have close proximity to consumers like the retail industry and media. But contrary to their observation, the findings of this study show the highest proportion of women in the metal industry; with more than 50% observations having at least one woman director. It is noteworthy, that none of the companies in none of the years had more women director than men. In the light of these findings, the recently enacted legislation mandating at least one woman director on every Board appears to be a much-needed move by the Government. It is startling to note (Table 5.13) that at least one company in every industry has violated the legal requirement regarding a minimum number of independent directors. Further, more than one-third of the observations in the oil and gas industry have violated this requirement. Given the highly regulated nature of this industry and the kind of scrutiny its companies are usually subjected to, such findings are noteworthy. It is pertinent to note that despite such violations, more than three-fourths of the companies in all the industries have independent boards. In other words, they have more independent directors than required by the law. This clearly indicates that Indian companies are aware of the benefits of transparency and fairness brought about by Independent directors. It is noteworthy that in almost all the industries, at least some of the companies have failed to constitute the mandatory committees. It is only in the power industry that all the companies have constituted all the mandatory committees (Table 5.13). The intent behind mandatory committees is to ensure effective and efficient governance. For instance, the audit committee has been mandated under Companies Act 2013 with the aim to reduce the risk of misstatement in financial statements and financial reporting. Similarly, the objective of remuneration and compensation committee is to ensure a commensurate reward to key managerial persons so as to avert agency problems. Therefore, non-constitution of such committees raises serious questions on the commitment of companies.

5.4 Empirical Evidence

199

In addition, the role of risk management committees (RMC) is pivotal in a risk governance structure. It is pertinent to note that despite the importance of risk management committee, more than four-fifths of the companies in each of the industries did not have such a committee. It is only in 4 industries that about one-fourth of the observations (Table 5.13) have constituted a formal risk management structure. In view of the increased attention on risk and risk management in the past decade, such findings are startling. It is believed that the recent mandate to constitute RMC will prove to be a stepping stone in the risk management process. On similar pattern are the findings related to the appointment of CRO. Almost in 90% observations in all the industries, there was no appointed designated CRO (Table 5.13). Keeping in mind Fayol’s principle of ‘unity of command’ and ‘unity of direction’, the absence of such designated post in most organisations is alarming. It is believed that it could jeopardise the Indian companies’ ability to manage risks well. Existence of whistle-blower policy is viewed as a check for operational risks. It is noteworthy that more than two-fifths of the observations in capital goods, FMCG, metal and oil and gas industry had a whistle-blower policy in place. But, in remaining industries only about one-third of the observations show any such policy. This could be viewed as either an indication of collusion at the top level to perpetuate irregularities or of confidence on part of companies, on employees’ and managements’ ethics. As previously stated, the intent behind the constitution of committees is to facilitate direct supervision and control in significant areas. Therefore, the constitution of certain committees has been left to the will of the companies. For the purpose of the study, these committees have been divided into five categories, namely, shares related, finance related, human resource related, and management related and miscellaneous. It is noteworthy that all the companies in the Agriculture industry had at least one such voluntary committee. Further, most of the companies in most of the industries have about two voluntary committees. It is pertinent to mention that evidence of constitution of all 5 voluntary committees is scanty. This clearly indicates that the majority of the companies focus on limited areas/functions. Therefore, there is a need to create awareness regarding enterprise-wide risk management; it will enable companies to move away from myopic management practices.

5.5

Implications

1. For policymakers—First and foremost, SEBI and Ministry of Corporate Affairs can take pride in implementing the much-needed rules/regulations (like mandating the constitution of a risk management committee) in the area of risk management. However, the fact that there is non-compliance with a number of legal/statutory provisions, calls for more stringent penal provisions in the case of

5 Normative Framework for Risk Governance Index …

200

2.

3.

4.

5.

such non-compliance. Further, to enhance the efficiency of the risk management process in organisations, SEBI may mandate the appointment of a risk advisory director, who will be a risk specialist with commensurate qualifications. In addition, instead of ensuring the presence of a woman Director on Board, efforts should be directed towards enabling balanced Boards with an equal number of male and female directors. For investors—Indian companies, by and large, have strong risk governance structures and are conscious of the importance of risk management. Such governance structures signal limited agency problems, more transparency, fairness and accountability. Therefore, investors may be confident about the risk-handling capabilities of these companies. For company itself—Balanced Boards and effective risk management structures will enable companies limit agency problems, promote transparency, fairness and accountability. In turn, this will help with better risk management. Further, it may boost investors’ confidence; reduce their required rate of return, resulting in a lower cost of capital and ease of borrowing for companies. However, the fact that the majority of companies shy away from constituting voluntary committees, needs attention. For other stakeholders—Strong governance structures signal the will and ability to manage risks well. Therefore, stakeholders like lenders and employees may be assured of timely identification and management of risks. For academics—As the attempt is perhaps first of its kind of attempt, there is a scope for cross-country analysis. Further, the relationship between risk structures and risk levels and/ risk management practices of firms may be explored.

5.6

Concluding Observations

Literature is replete with corporate governance studies and various versions of the corporate governance index are available. However, construction of a risk governance index (as proposed in this study) is perhaps the first of its kind attempt. For effective and efficient risk management, a risk governance structure is pertinent. Indian companies have effective risk management structure with mean index scores of 78%. Most companies have an optimal board size with 8–10 directors. Further, the Boards are, by and large, independent with adequate number of experts by way of non-executive directors. The under-representation of women in the upper echelons of management continues to be a cause of concern. In addition, absence of CRO in most companies is contrary to Fayol’s principle3 of ‘authority and responsibility’ and requires 3 As per Fayol’s principle of authority and responsibility in order to get things done, management should have the power to give orders to employees/subordinates. With this power comes the responsibility, which can be traced back on the basis of performance.

5.6 Concluding Observations

201

attention. Similarly, the infringement of a number of provisions is a cause of concern. It calls for more stringent penalties in the case of dereliction of rules. Recently introduced initiatives by SEBI and Companies Act appear to be the need of the hour and are expected to improve the structures further. The uncertainties experienced during the crisis period seem to have strengthened the risk governance structures. Similarly, age and industry have had profound effects on governance structures, during the period of study. Considering the fact that risks are all-pervasive and governance is a ubiquitous requirement, such results are puzzling. Thus, the study is believed to have important implications for regulators, investors as well as for management of companies. In sum, Indian corporates have strong governance structures, enabling effective and efficient risk management. However, these could further be strengthened with collaborative efforts of corporates and regulatory bodies.

References Aabo, A., & Simkins, B. J. (2005). Interaction between real options and financial hedging: Fact or fiction in managerial decision-making. Review of Financial Economics, 14(3), 353–369. Balinga, B. R., Moyer, R. C., & Rao, R. S. (1996). CEO duality and firm performance: What’s the fuss? Strategic Management Journal, 17(1), 41–51. Beasley, M. S. (1996). An empirical analysis of the relation between the board of director composition and financial statement fraud. Accounting Review, 71(4), 443–465. Beasley, M. S., Clune, R., & Hermanson, D. R. (2005). Enterprise risk management: An empirical analysis of factors associated with the extent of implementation. Journal of Accounting and Public Policy, 24(6), 521–531. Blau, P. M. (1977). Inequality and heterogeneity: A primitive theory of social structure. New York, NY: The Free Press. Boyer, M. M., & Stern, L. H. (2012). Is corporate governance risk valued? Evidence from directors’ and officers’ insurance. Journal of Corporate Finance, 18(2), 349–372. Brammer, S., & Pavelin, S. (2007). Factors influencing the quality of corporate environmental disclosure. Business Strategy and the Environment, 17(2), 120–136. Brickley, J., Coles, J., & Jarrell, G. (1997). The leadership structure: Separating the CEO and chairman of the board. Journal of Corporate Finance, 3(1), 189–220. Campbell, K., & Vera, A. M. (2008). Gender diversity in the boardroom and firm financial performance. Journal of Business Ethics, 83(1), 435–445. Chen, A., Kao, L., Tsao, M., & Wu, C. (2007). Building a corporate governance index from the perspectives of ownership and leadership for firms in Taiwan. Corporate Governance: An International Review, 15(2), 251–261. Companies Act (India). (2013). Act no. 18 of 2013. Available at indiacode.nic.in/acts-in-pdf/ 182013.pdf. Dahya, J., & McConnell, J. J. (2005). Outside directors and corporate board decisions. Journal of Corporate Finance, 11(1–2), 37–60. Daud, W. N., Yazid, A. S., & Hussin, M. R. (2010). The effect of chief risk officer (CRO) on enterprise risk management (ERM) practices: Evidence from Malaysia. International Business & Economics Research Journal, 9(11). Retrieved from http://dx.doi.org/10.19030/iber.v9i11.30.

202

5 Normative Framework for Risk Governance Index …

Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1996). Causes and consequences of earnings misstatement: An analysis of firms subject to enforcement actions by the SEC. Contemporary Accounting Research, 13(1), 1–36. Deszo, C. L., & Ross, D. G. (2012). Does female representation in top management improve firm performance? A panel data investigation. Strategic Management Journal, 33(9), 1072–1089. Eagly, A. H., & Johnson, B. T. (1990). Gender and leadership style: A meta-analysis. CHIP Documents, 11. Retrieved from http://digitalcommons.uconn.edu/chip_docs/11. Eisenberg, T., Sundgren, S., & Wells, M. (1998). Larger board size and decreasing firm value in small firms. Journal of Financial Economics, 48(1), 35–54. Financial Stability Board. (2013). Global shadow banking monitoring report. Retrieved from www.fsb.org/wp-content/uploads/r_131114.pdf?page_moved=1. Higgs, D. (2003). Review of the role and effectiveness of non-executive directors. Report of the Department of Trade and Industry, London, United Kingdom. Available at www.dti.gov.uk/ cld/non_exec_review. Institute of Directors. (2010). Business risk a practical guide for board members. London: Airmic Ltd and Chartis Europe Ltd. Retrieved from https://www.director.co.uk/wp-content/uploads/ 2015/02/Business-Risk.pdf. International Finance Corporation. (2012). When do companies need a board level risk management committee? Retrieved from www.ifc.org/wps/wcm/connect/…/PSO%2B31.pdf? MOD=AJPERES. Jensen, M. C., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency cost, and ownership structure. Journal of Financial Economics, 3(4), 305–360. Jensen, M. C. (1993). The modern industrial revolution, exit and the failure of internal control systems. The Journal of Finance, 48(3), 831–880. Jovanovic, B. (1982). Truthful disclosure of information. Bell Journal of Economics, 13(1), 36–44. Karamanou, I., & Vafeas, N. (2005). The association between corporate boards, audit committees, and management earnings forecasts: An empirical analysis. Journal of Accounting Research, 43(3), 453–486. Klein, A. (2002). Economic determinants of audit committee effectiveness. The Accounting Review, 77(2), 435–454. Lipton, M., & Lorsch, J. W. (1992). A modest proposal for improved corporate governance. Business Lawyer, 48(1), 59–77. Loderer, C., & Waelchli, U. (2011). Firm age and performance. Discussion Paper No. 09-18, German Economic Association of Business Administration—GEABA. Retrieved from geaba. de/DP/DP-09-18.pdf. Lorsch, J. W., & MacIver, E. (1989). Corporate governance and investment time horizons. In M. E. Porter (Ed.), Capital choices: Changing the way America invests in industry. Washington, DC: Council on Competitiveness. Millstein, I. (1992). The limits of corporate power: Existing constraints on the exercise of corporate discretion. New York, NY: Macmillan. Rechner, P. L., & Dalton, D. R. (1991). CEO duality and organizational performance: A longitudinal analysis. Strategic Management Journal, 12(2), 155–160. Rosener, J. B. (1995). America’s competitive secret: Utilizing women as a management strategy. New York: Oxford University Press. Rothschild, J., & Miethe, T. D. (1999). Whistle-blower disclosures and management retaliation. Work and Occupations, 26(1), 107–128. Sarbanes Oxley Act (USA). (2002). 107th Congress Public Law 204, July 30th. Available at https://www.congress.gov/bill/107th-congress/house-bill/3763. Schmalensee, R. (1985). Do markets differ much? American Economic Review, 75(3), 341–351. Securities and Exchange Board of India. (2014, April 17). SEBI circular no. CIR/CFD/POLICY CELL/2/2014. Retrieved from http://www.sebi.gov.in/cms/sebi_data/attachdocs/ 1397734478112.pdf.

References

203

Singh, S., Jain, P. K., & Yadav, S. S. (2012). Capital budgeting decisions: Evidence from India. Journal of Advances in Management Research, 9(1), 96–112. The Public Interest Disclosure Act (UK). (1998). C 23 of 1998. Available at www.legislation.gov. uk/ukpga/1998/23/contents. United Nations Council on Trade and Development (UNCTAD), investment brief (November 1, 2009). Available at unctad.org/…/Results.aspx?…write%3E%3D%2203%2F06%2F2012% 22%20unctadla. Wernerfelt, B., & Montgomery, C. A. (1988). Tobin’s q and the importance of focus in firm performance. The American Economic Review, 78(1), 246–250. White, J., & Ingrassia, P. (1992). Board ousts managers at GM: Takes control of crucial committee. The Wall Street Journal, 13(1), 15–31. Yermack, D. (1996). Higher market valuation for firms with a small board of directors. Journal of Financial Economics, 40(2), 185–211.

Chapter 6

Examining Risk–Return Relationship

All models are wrong, but some are useful. —George E. P. Box

Abstract This chapter aims to explore the relationship between (accounting based) risk index developed in the study and accounting returns. In view of the possible endogeneity problem, diff-GMM regression has been used. The results contradict the widely accepted hypothesis of ‘higher the risk, higher the return’; and lend credence to the fact that by following the normative risk index developed in Chap. 3 , and by keeping lower risk levels, firms may generate higher returns. Keywords Risk–return

6.1


Risk index
ROA
ROE
Endogeneity
Diff-GMM

Introduction

The chapter attempts to determine the relationship between risk levels as measured by risk index developed in Chap. 4 and accounting returns. Yet, there is no consensus on the nature of risk–return relationship. Much of the ambivalence in the domain may be ascribed to various risk–return measures in vogue. The chapter has been organized into seven sections. Section 6.2 presents a brief background of risk–return relationship. Section 6.3 describes the sample used. Section 6.4 elaborates the methodology employed. Section 6.5 examines the findings and presents the analysis of the same. This is followed by a Sect. 6.6 on implications for various stakeholders. Section 6.7 presents concluding observations.

6.2

Background and Hypotheses Development

There are two categories of risk–return measures (i) accounting-based and (ii) market-based. From the market’s perspective, return is measured as equity returns and from accounting perspective as return on assets (ROA), return on equity © Springer Nature Singapore Pte Ltd. 2019 M. V. Shivaani et al., Understanding Corporate Risk, India Studies in Business and Economics, https://doi.org/10.1007/978-981-13-8141-6_6

205

206

6 Examining Risk–Return Relationship

(ROE) or cash flows to total assets. Risk is generally defined as variance or standard deviation of the return or is proxies by leverage (Coskun and Kulali 2016). Literature is rife with studies that have examined the risk–return relationship. Most studies provide evidence in support of positive risk–return relationship (Fama and French 1992). But, since the seminal work of Bowman (1980), a large number of studies have observed negative risk–return relationship as well. Therefore, the true relationship between risk and return continues to be a puzzle. It is worth noting that much of the ambiguity regarding the nature of risk–return relationship is attributed to the use of different risk–return measures and the varying context in which they are used. Lehner (2000) and Brick et al. (2015) suggest that another possible reason for conflicting results on risk–return relationship is the use of end of the period (EOP) versus beginning of the period (BOP) measures. Apart from the abovementioned reasons, the following theories are often used to provide justification for the nature of risk–return relationship. Behavioural theory states that managerial decisions are based on mangers’ aspirations and expectations of performance. If the expectations fall short of the aspirations, a negative risk-relationship is expected and if expectations surpass the aspirations, managers will exhibit risk-averse behaviour and a positive risk-relation is indicated (Bromiley 1991; Fiegenbaum et al. 1996; Greve 1998; Palmer and Wiseman 1999; Li et al. 2014). Prospect theory (Kahneman and Tversky 1979) suggests that managerial decisions are based on prospects (good/bad) of the company as well as the current performance of the company. If the managers have reached their target rate of return they will tend to be risk-averse and a positive risk–return relationship may be expected (Jegers 1991; Sinha 1994; Wiseman and Catanach 1997). Startegic conduct hypothesis (Siggelkow 2001) thrives on the concept of strategic fit and heterogeneity in firms’ strategic capabilities. As per this theory, firms that have unique and dynamic capabilities, and that form unique strategic groups (Fiegenbaum and Thomas 2004) tend to achieve higher returns with lower risks (Miller and Chen 2003; Andersen et al. 2007). Another stream of literature resorts to implicit cost hypothesis to explain the negative risk–return relationship. Deephouse and Wiseman (2000) suggest that high risk will increase company’s operating cost, causing reduction in its operating income which, in turn, leads to poor firm performance. In view of these contrasting results, it has been desired to examine the risk–return relationship, where the measure of risk is the risk index developed in Chap. 4. It may be noted that the risk index has been put forth with the objective of developing a normative framework that should enable firms to keep their exposure levels in tenable limits. The framework is normative in the sense that it has its genesis in sound tenets of theory and generally accepted accounting and finance principles; in operational terms, it is believed that the companies that do well in terms of this normative framework will also be the companies that will (or are likely to) perform well in terms of generally accepted performance indicators such as ROA and ROE. In other words, a negative relationship should be expected between

6.2 Background and Hypotheses Development

207

the risk index and returns of companies. Theoretically, higher operating risk (say, in terms of higher DOL), should get reflected in terms of lower operating returns. As the higher risk here reflects that business has not been managed well it should result in correspondingly lower returns. In view of the foregoing discussion, the following two hypotheses have been formulated: H1: There is a negative relationship between risk index (developed in Chap. 4) and return on assets. H2: There is a negative relationship between risk index (developed in Chap. 4) and return on equity.

6.3

Sample

The sample consists of non-financial companies that constitute Nifty 500 index as on March 31, 2014. The study covers a period of 10 years from April 1, 2005 to March 31, 2015. The data has been obtained from databases like Bloomberg and Ace-equity and companies’ annual reports.

6.4

Methodology

The main objective of the Chapter is to examine the relationship between the scores obtained by the sample companies on the risk index (developed in Chap. 4) and returns earned by companies. As the risk index is largely based on accounting measures, therefore, the Chapter makes use of accounting measures of return, namely, return on assets (ROA) and return on equity (ROE). Also, it is well established in literature that equity returns of a company are affected not just by company-specific factors but also by global, economic and industry factors as well; thus, equity returns may not be a suitable proxy of returns for the purposes of the current study. The variables used in the Chapter are as follows: Dependent variables: Return on assets (ROA)—ROA is one of the most popular measures of return in studies dealing with risk–return relationship (Bowman 1980; Andersen et al. 2007). It helps to understand how efficiently the funds have been applied and also to understand the operating efficiency of firms. ROA has been measured as Net profit after tax þ Interest Average total assets

ð6:1Þ

208

6 Examining Risk–Return Relationship

Return on equity (ROE)—As per Singh et al. (2016), ROE is not only suggestive of the economic efficiency of the firm but is also indicative of gainful employment of resources and operating competence of the firm. ROE presents the financial performance of the firm from a view that is more relevant to shareholders. It has been measured as Net profit after tax  Preference dividend Average equity shareholders0 funds

ð6:2Þ

Independent variable: Risk Index (RI)—It refers to the risk index developed in Chap. 4; this risk index is proposed to be used as a proxy for the risk (exposures) level in a company. The index comprises of nine risks namely, market risk, accounting risk, competition risk, contingency risk, credit risk, foreign exchange rk, liquidity risk, operating risk and solvency risk. Control Variables—It is often observed that there are certain variables that extraneously affect the relationship of interest. Therefore, in order to unravel the true relationship among the relevant variables, it is pertinent to control for certain unimportant factors (Carlson and Wu 2012; Spector and Brannick 2011). Based on literature review, following control variables have been incorrated in the regression models used in this chapter. Risk Governance Index (RGI)—It is the governance index developed in this chapter. Strategic endowment hypothesis suggests that certain firm-specific strategic factors such as managerial ability may enable a company to achieve high returns with low-risk levels (Kim et al. 1993; Andersen et al. 2007). Taking cue from the said hypothesis, quality of governance structure has been proposed to be used as a control variable. Age—Age may have a positive impact on firm performance as it enables firms to move up the learning curve (Arrow 1962; Ericson and Pakes 1995), but, with experience firms may become too rigid leading to a negative impact on firm performance (Leonard and Barton 1992). Thus, it seems imperative to control for firms’ age. It has been measured as the number of years a firm has been in existence (since its inception). Growth—Growth opportunities available to a firm are expected to have a substantial influence on firms’ performance (Fama and French 2006; Loderer and Waelchli 2009). Annual growth in sales has been used as a proxy for firm’s growth. Size—Coskun and Kulali (2016) suggest that firm size may have significant influence on risk–return profile of a firm. They suggest that managerial capabilities, scale of operations and access to resources vary significantly in terms of firm’s size. Therefore, it seems pertinent to control for size of firms. It has been measured as natural log of total assets of the firm.

6.4 Methodology

209

Recession—It has been captured by introduction of a dummy variable for pre-recession period (2005–2008) and post-recession period (2008–2015). Given the panel nature of data, panel data regression appears to be the appropriate technique. Panel data analysis has been considered to be more advantageous than OLS regression. Panel data analysis facilitates consideration of firm-specific heterogeneities that may be having an impact on the dependent variable, inclusion of more informative data, and more efficient analysis (Hsiao 2003; Baltagi 2008). Further, risk and return are so intimately connected that there are possibilities of endogentites (Xiaodong and Lee 2013). Wintoki et al. (2010) suggest three potential sources of endogeneity that may exist in panel data structures: (i) Dynamic endogeneity—it happens when the current period values of a variable are influenced by values of preceding periods. (ii) Simultaneity—It happens when two variables simultaneously affect each other, resulting in their co-determination. (iii) Unobserved heterogeneity—It is a situation where some third unobservable variable affects the relationship between two variables of interest. The most common solution to deal with endogeneity problems is the use of lagged dependent variables or instrumental variables. In case OLS is used for estimation, it typically results in an upward bias in the coefficient of lagged dependent variable (Bond 2002). Similarly, in the context of unobservable firm heterogeneities, Baltagi (2008) discourages the use of fixed effects model (particularly, when the panel is a short panel). He suggests that the lagged dependent variable may end up being correlated with error-term, resulting in biased coefficients. Further, the coefficients of the lagged dependent variable, obtained through FE estimation may have a downward bias (Nickell 1981). To overcome these problems, Holtz-Eakin et al. (1988) proposed generalized method of moments (GMM) panel specifications, which was later popularised by Arellano and Bond (1991), Arellano and Bover (1995) and Blundell and Bond (1998). It is worth mentioning that Arellano and Bond (1991) first-difference the panel data to remove the time-invariant fixed effect and show that the lagged dependent variables’ values (levels) constitute legitimate variables; provided that the residuals are free from second-order serial correlation. As a company is a going concern, it seems reasonable to assume that performance of a firm in a year may be (rather should be) affected by the firm’s performance in preceding years. Further, Bowman (1980) quotes that distressed firms take higher risk which results in poor firm performance and so on, and it is a vicious circle. As GMM takes into account problems of endogeneity, Bowman’s (1980) argument that troubled firms take more risk is also factored in. Hence, given the panel nature of data and following past studies, the study proposes to use ‘difference GMM’ for estimating various relationships.

210

6.5 6.5.1

6 Examining Risk–Return Relationship

Empirical Evidence Descriptive Analysis

It is evident (Table 6.1) that the mean risk index of the sample companies has increased from 41% in 2005–06 to 44% in 2014–2015. Though the increase is of only 3% points or 1.42 points out of maximum possible 45 points; it is a sign of increasing distress. The marginal fall in average risk index, in 2 years immediately preceding the period of recession is puzzling. But almost a steady increase in average risk index in each of the years in post-recession period is indicative of lagged effects of recession. The risk index in the range of 40–45% is suggestive of moderate risk levels of the sample companies. In other words, an average risk index in this range indicates that on an average the companies have a risk score of 2 (out of 5) on each of the nine risks. This could be viewed as decent risk management, as risk scores in this range appear to be a result of ‘calculated risks’ rather than unattended risks. It is not surprising that Indian companies had these moderate risk levels in pre-recession period as well during the recession (Fig. 6.1). This is so because had they had excessive exposures, they might not have been able to survive the downturn. Given the fact that these are India’s topmost companies and they could endure the gloomy market period, it seems unreasonable to assume that they would have had exorbitant levels of risk exposure. But, a steady increase in risk levels of Indian companies, with the highest average risk index being in 2014–2015, is alarming. This could be viewed as both an indicator, and a consequence of ‘contagion’. With increasing globalisation and cross border trades, it would be unjustified to assume that companies work in isolation or closed environment of the country that they are operating in.

Table 6.1 Descriptive statistics of risk index Year

Mean (%)

Median (%)

Std. deviation (%)

Minimum (%)

Maximum (%)

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Total

41.01 40.69 40.52 42.01 42.59 41.71 42.61 42.90 43.21 44.18 42.23

41.11 40.83 40.00 41.39 41.94 40.83 42.22 42.22 42.50 43.33 41.94

6.71 6.43 6.70 7.36 7.21 7.74 8.04 8.85 8.62 9.50 7.92

25.28 26.39 21.94 25.56 26.11 24.17 22.78 21.94 24.17 22.78 21.94

69.72 67.78 61.67 64.44 66.94 77.78 64.44 73.33 66.67 69.72 77.78

6.5 Empirical evidence 90.00% 80.00%

211

Risk index

70.00% 60.00%

Mean

50.00%

Minimum

40.00%

Maximum

30.00%

Std. Deviation

20.00% 10.00% 0.00% 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Total

Fig. 6.1 Year-wise descriptive statistics of risk index

Companies’ fundamentals, no matter how strong they are, cannot be cushioned against world-wide booms and depressions without adequate risk management. It is heartening to note that the minimum risk index has declined during the period of study. It is suggestive of dynamic risk management, at least on part of some companies. Similarly, the maximum risk index for a particular year has by and large remained the same. It is only in the year 2010–2011 that the risk index peaked to a maximum of 78% (approx.). It is noteworthy, as the period of recession is believed to be from 2008–2009, hence its true effects on accounting figures could only be observed in the financial year with a one year lag. Further, a low standard deviation (within each year) is indicative of the fact that the companies have similar exposures to the risks considered in the study. Since, these exposures implicitly incorporate the risk handling mechanism employed for concerned risk, it is also suggestive of somewhat similar risk appetite and tolerance by Indian companies. Table 6.2 shows that the mean ROA of the sample companies has been steadily declining since 2007–08. Such a decline may be attributed to the global financial crisis that began to grapple the world economy towards the second half of 2007–08 (UNCTAD 2009). Though the decline of about 3% points in ROA should be a cause of worry, the fact that the mean ROA is still 11.38% is encouraging. In simple words, it implies that the Indian firms, on an average, in current turbulent times, full of uncertainty and pitfalls, are able to generate a return of more than Rs. 11 on every Rs. 100 worth of assets. Further, the maximum return generated by a company in any particular year has by and large remained the same throughout the period of study (approx. 35%). While the fact that the lowest ROA experienced by a company happens to be in the latest year under consideration (i.e. 2014–15) should be a cause of concern (Fig. 6.2). In the context of ROE, it can be seen in Table 6.3 that it has considerably reduced from 20.22% in 2006 to 12% in 2015. Though the decline is worrisome,

212

6 Examining Risk–Return Relationship

Table 6.2 Descriptive statistics of return on assets Year

Mean (%)

Median (%)

Std. deviation (%)

Minimum (%)

Maximum (%)

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Overall

13.40 14.25 13.69 12.53 12.88 12.05 11.91 11.65 11.53 11.38 12.44

12.51 12.99 12.55 11.79 12.24 11.42 10.83 10.68 10.41 10.06 11.56

5.70 6.72 6.60 6.65 6.55 5.88 6.54 6.44 6.56 6.61 6.50

1.79 1.25 −3.20 −7.31 −2.81 −2.94 −6.29 −6.48 −5.27 −1.68 −7.31

31.25 37.43 38.36 35.78 36.06 33.50 37.99 34.48 34.15 36.96 38.36

50.00%

ROA

40.00% 30.00%

Mean

20.00%

Std. Deviation Minimum

10.00%

Maximum 0.00% -10.00%

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Total

Fig. 6.2 Year-wise descriptive statistics of return on assets

the fact that world over, economies are still experiencing the tremors of global financial crisis cannot be overlooked; in light of this fact, an average ROE of about 16% (for the period) seems satisfactory. In addition, the fact that there is at least one company, in each of the years under the study, which has been able to generate a return of over 50% for its equity-holders, is emboldening. Further, these findings provide evidence about the fundamental strength of Indian companies, which have been able to endure and sustain themselves in the aftermath of recession. In contrast, the finding of negative ROE by some companies is alarming and needs to be probed (Fig. 6.3). The descriptive statistics indicate that the risk and return for the sample period have moved in opposite directions. On the one hand, risk appears to have been on the rise, on the other hand, returns seem to have faced a steady decline. Therefore, it seems imperative to gauge whether the returns are dependent on the risk (as measured by the risk index). For the purpose, the following analysis has been carried out.

6.5 Empirical evidence

213

Table 6.3 Descriptive statistics of return on equity Year

Mean (%)

Median (%)

Std. deviation (%)

Minimum (%)

Maximum (%)

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Overall

20.22 21.44 20.17 15.76 16.89 15.50 13.92 13.38 12.37 11.93 15.87

19.27 19.88 18.48 14.68 16.29 14.29 12.71 12.13 11.09 11.18 14.66

10.86 12.30 13.01 12.29 11.26 11.41 11.34 10.70 11.37 11.45 12.00

0.61 −11.36 −32.14 −23.56 −25.23 −22.41 −29.27 −32.46 −32.71 −33.09 −33.09

58.77 62.52 66.06 63.57 64.41 65.22 55.44 54.33 53.21 50.23 66.06

80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% -10.00% -20.00% -30.00% -40.00%

ROE Mean Std. Deviation Minimum Maximum

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Total

Fig. 6.3 Year-wise descriptive statistics of return on equity

6.5.2

Relationship Between Risk Index and ROA (Model 1)

The model attempts to examine the impact of risk index on ROA. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation. ROAit ¼ a þ ROAit1 þ RIit þ Ageit þ Sizeit þ Growthit þ Recession dummyit þ eit ð6:3Þ The results of the estimation have been presented in Table 6.4. It can be observed from Table 6.4 that the ROA of immediately preceding previous year has a statistically significant impact on the ROA of the current year. Further, risk index appears to have a negative and statistically significant relationship with the ROA of the same year.

214

6 Examining Risk–Return Relationship

Table 6.4 Results of (Arellano-Bond) GMM estimation of ROA on first lag of ROA, RI and control variables Variables

Coefficient

ROA (lag) 0.1618408 RI −0.004004 Age 0.0018 Size −0.0499 Growth −0.0007 Recession dummy −0.0053 Constant 0.4734 Number of observations: 2362 Number of instruments: 42

Standard error

p-value

0.0580 0.0216 0.0055 0.0056 0.0007 0.0034 0.42

0.005*** 0.064* 0.001*** 0.000*** 0.294 0.116 0.000***

Statistic v2(6) 190.93 v2(35) 42.4446 First order −3.8922 Second order −1.453 ***, * significant at 1% and 10% level of significance, respectively Wald test Sargan test Autocorellation

p-value 0.0000*** 0.1809 0.0001 0.1462

As discussed earlier, the finding of a negative risk–return relationship is often termed as Bowman’s paradox (Bowman 1980). It may be noted that even the studies that have documented a negative risk–return relationship (Armour and Teece 1978; Bowman 1980, 1982; Fiegenbaum and Thomas 1986; Coskun and Kulali 2016) have by and large, relied on the same measures of risk as have the studies that have found a positive relation; the measure of risk being, ‘variability’ of some sort of returns. But, the current study considers risk in terms of exposure levels, as measured by the risk index constructed in Chap. 4, and not in terms of variability of returns. It may further be noted, that the said risk index is basically a normative framework (that attempts to put forth a Basel like framework for non-financial companies) where a score of 1 on a particular risk is indicative of least risky levels in terms of sound tenets of theory and a score of 5 indicates an alarming or distressing situation. Therefore, lower risk scores are indicative of financial and operational strength of the firm. Intuitively, such efficient and effective risk management should get reflected in terms of better firm performance. In other words, firms with lower risk should exhibit higher returns and vice versa. Thus, the result of a negative risk–return relationship as captured in Table 6.4 provides supporting evidence for the normative framework; the result strengthens the reliability of the index. This result could be better explained with the following illustration—if a firm is operating at the least risky levels in terms of say, liquidity risk (as measured by a score of 1 on inverse of acid-test ratio) it means that the firm is well-equipped to meet its current liabilities on time. As the firm has adequate resources to meet its current liabilities, it is saved from penalties and unreasonably high costs of

6.5 Empirical evidence

215

financing, to which it may have been subjected to, had it not have been able to meet its liabilities (currently maturing) on time. Therefore, this efficient and effective utilisation of resources should ultimately get reflected in terms of higher ROA. The finding of a negative risk–return relationship when viewed in conjunction with the moderate risk levels (found for Indian companies), highlights how rational the decision makers are. It gives a novel dimension to the whole debate on risk– return relationship. Generally, the negative relationship between risk and return is associated with ‘risk seeking’ behaviour, but such an inference may not be tenable in the present context. First, that explanation is in context of ex-ante or expected returns. Second, the measure of risk used in such cases is usually ‘standard deviation of returns’. It is noteworthy that, in terms of control variables, age and size exhibit a statistically significant relationship with returns. On the one hand, age has a positive relationship, on the other hand, size (Greene 2003; Brealey et al. 2008; Brick et al. 2015) has a negative relationship. It may further be noted that the p-value of 0.000 for Wald test indicates that the model used is a good fit. The GMM specifications are well-specified based on the Sargan test of overidentifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors.

6.5.3

Relationship Between Risk Index and ROA (Model 2)

The model attempts to examine the impact of risk index on ROA while controlling for quality of risk governance structure, among other variables. As mentioned earlier, in view of ‘strategic endowment hypothesis’ it seems pertinent to control for quality for governance structure. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: ROAit ¼ a þ ROAit1 þ RIit þ RGIit þ Ageit þ Sizeit þ Growthit þ Recession dummyit þ eit

ð6:4Þ The results of the estimation have been presented in Table 6.5. As in the case of Table 6.4, Table 6.5 shows that the first lag of ROA is significantly and positively related to current levels of ROA. In other words, companies’ current ROA is significantly related to immediately preceding year’s ROA. Despite controlling for quality of risk governance, the main variable of interest, risk index, has a negative and significant impact on ROA. Interestingly, quality of governance structure does not have a significant relationship with ROA. In terms of control variables, age, and size continue to have a significant relationship with ROA.

216

6 Examining Risk–Return Relationship

Table 6.5 Results of (Arellano-Bond) GMM estimation of ROA on first lag of ROA, RI, RGI and control variables Variables

Coefficient

ROA (lag) 0.1586705 RI −0.040552 RGI −0.0132 Age 0.0018 Size −0.0496 Growth −0.0007 Recession dummy −0.0053 Constant 0.4787 Number of observations: 2362 Number of instruments: 42

Standard error

p-value

0.0578 0.02165 0.0199 0.0005 0.0057 0.0007 0.0034 0.0454

0.006*** 0.061* 0.507 0.001*** 0.000*** 0.294 0.113 0.000***

Statistic v2(7) 190.76 v2(35) 42.3777 First order −3.8755 Second order −1.4657 ***, * significant at 1% and 10% level of significance, respectively Wald test Sargan test Autocorellation

p-value 0.0000*** 0.1827 0.0001 0.1427

It is noteworthy that the p-value of 0.000 for Wald test indicates that the model used is a good fit. The GMM specifications are well-specified based on the Sargan test of over-identifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors.

6.5.4

Relationship Between Risk Index and ROE (Model 3)

The model attempts to examine the impact of risk index on ROE. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: ROEit ¼ a þ ROEit1 þ RIit þ Ageit þ Sizeit þ Growthit þ Recession dummyit þ eit ð6:5Þ The results of the estimation have been presented in Table 6.6. Results in Table 6.6 indicate that ROE of immediately preceding previous year has a significant negative impact on ROE of the current year; the finding may be attributed to the fact that there has been a steady increase in interest rates in Indian context and the total debt to equity ratio of firms has also increased during the

6.5 Empirical evidence

217

Table 6.6 Results of (Arellano-Bond) GMM estimation of ROE on first lag of ROE, RI and control variables Variables

Coefficient

ROE (lag) −0.1472577 RI −0.2492839 Age −0.0185031 Size 0.0339411 Growth −0.0012386 Recession dummy −0.336393 Constant 0.7570026 Number of observations: 2362 Number of instruments: 42

Standard error

p-value

0.0183004 0.0404904 0.0032447 0.0170743 0.0021449 0.0083236 0.0784474

0.000*** 0.000*** 0.000*** 0.047** 0.564 0.000*** 0.000***

Statistic v2(6) 153.32 v2(35) 43.99361 First order −1.5746 Second order 0.77395 ***, ** significant at 1% and 5% level of significance, respectively Wald test Sargan test Autocorellation

p-value 0.000*** 0.1417 0.1154 0.4390

period under study. Further, as noted in the case of ROA, risk index is significantly and negatively related to ROE as well. In other words, firms that have lower risk levels are able to generate higher returns for their equity-holders. The rationale and justification for these results are similar to that for the results of Model 1 and Model 2. Lower risk levels signal effective and efficient utilisation of resources resulting in higher profits and consequently higher ‘earnings available for equity-shareholders’. In terms of control variables, age, size of firm and recessionary conditions exhibit a significant relationship with ROE. The results indicate that older firms tend to have lower ROE than younger firms, and bigger firms tend to generate higher returns for their shareholders than smaller firms. Further, the post-recession period has had a negative impact on ROE. It is noteworthy that the p-value of 0.000 for Wald test indicates that the model used is a good fit. The GMM specifications are well-specified based on the Sargan test of over-identifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors.

6.5.5

Relationship of Risk Index ROE (Model 4)

In order to gain a holistic view of the relationship between risk index and ROE, it appears reasonable to introduce RGI as a control variable. Using Arellano and Bond

218

6 Examining Risk–Return Relationship

Table 6.7 Results of (Arellano-Bond) GMM estimation of ROE on first lag of ROE, RI, RGI and control variables Variables

Coefficient

ROE (lag) −0.1395419 RI −0.2543248 RGI −0.0000734 Age −0.0182354 Size 0.0324281 Growth −0.0012283 Recession dummy −0.032552 Constant 0.7556955 Number of observations: 2362 Number of instruments: 42 v2(7) v2(35) First order Second order *** significant at 1% level of significance Wald test Sargan test Autocorellation

Standard error

p-value

0.0175503 0.0396392 0.0452627 0.0034374 0.0178052 0.0021345 0.0084353 0.079088

0.000*** 0.000*** 0.999 0.000*** 0.069*** 0.565 0.000*** 0.000***

Statistic

p-value

155.94 43.54476 −1.5802 0.78041

0.000*** 0.1523 0.1141 0.4351

(1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: ROEit ¼ a þ ROEit1 þ RIit þ RGIit þ Ageit þ Sizeit þ Growthit þ Recession dummyit þ eit

ð6:6Þ As in the case of ROA (Model 2), introduction of RGI as a control variable in the Model for ROE does not alter the results (Table 6.7). ROE of immediately preceding previous year, risk index, age, size and recession continue to have a significant impact on ROE. Further, a p-value of 0.000 in the case of Wald test suggests that the model used has a good fit. Also, the statistics pertaining to Sargan test indicate that there is no mis-specification in the model. Moreover, the Arellano– Bond test statistics indicate that there exists no autocorrelation in the errors.

6.6

Implications

For policy-makers—The results obtained in the chapter provide support for the methodology used in the construction of risk index and strengthens the validity of the normative framework on which the index is based. In view of these findings, it seems reasonable to recommend that Institute of Chartered Accountants of India,

6.6 Implications

219

Securities and Exchange Board of India and Ministry of Corporate Affairs should come up with a Basel like framework for non-financial companies. For companies—As the results of the chapter lend credence to the normative framework put forth in Chap. 4, companies may adapt the said framework as a part of their risk management spectrum. For investors—This easy to compute risk measure is expected to help investors in evaluation of firms. It is also expected to be useful in gauging the risk appetite and risk tolerance levels of various firms. Consequently, the investors may choose securities of companies whose appetite aligns best with theirs. For academics—The index may be used as risk measure in strategic management literature, where variability of returns as a measure of risk is often criticised. Further, cross-country studies may be conducted to further strengthen the robustness of the results.

6.7

Concluding Observations

The chapter examines the relationship between the risk index developed in Chap. 4 and accounting returns as measured by return on assets (ROA) and return on equity (ROE). The findings indicate a negative risk-return relationship. In other words, lower a company scores on the risk index, the higher are its ROA and ROE. These results lend credence to the normative framework developed in Chap. 4 and provide support for the risk index based on the said framework. The results seem counter-intuitive as the generally accepted proposition in financial economics is ‘higher the risk, higher the return’. But, as the index is based on a normative framework and is recommendatory in nature, companies that fare well in terms of risk should also be the companies that are able to generate higher returns for their stakeholders. These results seem similar to the results of the studies that belong to the domain, popularly known as ‘Bowman’s paradox’. Most of these studies have used the variability of returns as a measure of risk and have consequently sought the explanation of negative risk–return relationship in behavioural theory and prospect theory. But, as the scope, context and purpose of the risk index used in the chapter are substantially different from those of the most used measure of risk (variability of returns), it may not be reasonable to use the traditional theories to explain the results of this chapter. Therefore, the finding of the negative risk–return relationship, in the present context, should be attributed to sounds tenets of theory and financial prudence of rational decision-makers. To conclude, the risk–return relationship always needs to be viewed in some context, any generalisation is a long shot given the peculiarities, idiosyncrasies and heterogeneities involved in financial context, setting and environment.

220

6 Examining Risk–Return Relationship

References Andersen, T. J., Denrell, J., & Bettis, R. A. (2007). Strategic responsiveness and Bowman’s risk-return paradox. Strategic Management Journal, 28(1), 407–429. Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58(1), 277–297. Arellano, M., & Bover, O. (1995). Another look at the instrumental variable estimation of error components models. Journal of Econometrics, 68(1), 29–51. Armour, H. O., & Teece, D. J. (1978). Organizational structure and economic performance: a test of the multidivisional hypothesis. The Bell Journal of Economics, 9(1), 106–122. Arrow, K. (1962). Economic welfare and the allocation of resources to invention. In The Universities—National Bureau Committee for Economic Research and the Committee on Economic Growth of the Social Science Research Councils (ed) The rate and direction of inventive activity: Economic and social factors (pp 609–626). Princeton, NJ: Princeton University Press. Baltagi, B. (2008). Econometric analysis of panel data (4th ed.). Chichester, United Kingdom, UK: Wiley. Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143. Bond, S. R. (2002). Dynamic panel data models: a guide to micro data methods and practice. Portuguese Economic Journal, 1(2), 141–162. Bowman, E. (1980). A risk/return paradox for strategic management. Sloan Management Review, 21(1), 17–31. Bowman, E. (1982). Risk seeking by troubled firms. Sloan Management Review, 31(2), 33–42. Brealey, R. A., Myers, S. C., & Allen, F. (2008). Principles of corporate finance. New York, NY: McGraw-Hill. Brick, I. E., Palmon, O., & Venezia, I. (2015). On the relationship between accounting risk and return: is there a (Bowman) paradox? European Management Review, 12(2), 99–111. Bromiley, P. (1991). Testing a causal model of corporate risk taking and performance. Academy of Management Journal, 24(1), 37–59. Carlson, K. D., & Wu, J. (2012). The illusion of statistical control: Control variable practice in management research. Organizational Research Methods, 15(3), 413–435. Coskun, M., & Kulali, G. (2016). Relationship between accounting based risk and return: Analysis for Turkish companies. International Journal of Business and Management, 11(4), 240–251. Deephouse, D. L., & Wiseman, R. M. (2000). Comparing alternative explanations for accounting risk-return relations. Journal of Economic Behavior & Organization, 42(1), 463–482. Ericson, R., & Pakes, A. (1995). Markov-perfect industry dynamics: A framework for empirical work. The Review of Economic Studies, 62(1), 53–82. Fama, E. F., & French, K. R. (1992). The cross-section of expected stock returns. The Journal of Finance, 47(2), 427–465. Fama, E. F., & French, K. R. (2006). Profitability, investment and average returns. Journal of Financial Economics, 82(1), 491–518. Fiegenbaum, A., & Thomas, H. (1986). Dynamic and risk measurement, perspectives on Bowman’s risk–return paradox for strategic management: An empirical study. Strategic Management Journal, 7(1), 395–407. Fiegenbaum, A., & Thomas, H. (2004). Strategic risk and competitive advantage: An integrative perspective. European Management Review, 1(1), 84–95. Fiegenbaum, A., Hart, S., & Schendel, D. (1996). Strategic reference point theory. Strategic Management Journal, 17(1), 219–235. Greene, W. H. (2003). Econometric analysis. New York: New York University. Greve, H. R. (1998). Performance, aspirations and risky organizational change. Administrative Science Quarterly, 43(1), 58–86.

References

221

Holtz-Eakin, D., Newey, W., & Rosen, H. (1988). Estimating vector auto regressions with panel data. Econometrica, 56(6), 1371–1395. Hsiao, C. (2003). Analysis of panel data. London. United Kingdom: Cambridge University Press. Jegers, M. (1991). Prospect theory and the risk–return relation: Some Belgian evidence. Academy of Management Journal, 34(1), 215–225. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291. Kim, W. C., Hwang, P., & Burgers, W. P. (1993). Multinationals’ diversification and the risk– return trade. Strategic Management Journal, 14(1), 275–286. Lehner, J. M. (2000). Shifts of reference points for framing of strategic decisions and changing risk-return associations. Management Science, 46(1), 63–76. Leonard-Barton, D. (1992). Core capabilities and core rigidities: A paradox in managing new product development. Strategic Management Journal, 13(1), 111–125. Li, X., Yang, F., & Zhang, R. (2014). Determinants of corporate risk taking and risk-return relationship. Canadian Social Science, 10(2), 24–32. Loderer, C., & Waelchli, U. (2009). Firm age and performance. Discussion Paper No. 09-18. German Economic Association of Business Administration—GEABA. Retrieved from geaba. de/DP/DP-09-18.pdf. Miller, K. D., & Chen, W. (2003). Risk and firms’ costs. Strategic Organization, 1(4), 355–382. Nickell, S. (1981). Biases in dynamic models with fixed effects. Econometrica, 49(6), 1417–1426. Palmer, T. B., & Wiseman, R. M. (1999). Decoupling risk taking from income stream uncertainty: A holistic model of risk. Strategic Management Journal, 20(1), 1037–1062. Siggelkow, N. (2001). Change in the presence of fit: The rise, the fall, and the renaissance of Liz Claiborne. Academy of Management Journal, 44(1), 838–857. Singh, S., Jain, P. K., & Yadav, S. S. (2016). Equity markets in India: Returns. Delhi, Springer (India) Private Limited: Risk and Price Multiples. Sinha, T. (1994). Prospect theory and the risk return association: Another look. Journal of Economic Behavior & Organization, 24(1), 225–231. Spector, P. A., & Brannick, M. T. (2011). Methodological urban legends: The misuse of statistical control variables. Organizational Research Methods, 14(1), 287–305. United Nations Council on Trade and Development (UNCTAD), investment brief. Accessed 1 Nov, 2009 from unctad.org/…/Results.aspx?…write%3E%3D%2203%2F06%2F2012%22% 20unctadla. Wintoki, M. B., Linck, J. S., & Netter, J. M. (2010) Endogeneity and the dynamics of internal corporate governance. In CELS 2009 4th Annual Conference on Emperical Legal Studies. SSRN: http://ssm.com/abstract=970986. Wiseman, R. M., & Catanach, A. H. (1997). A longitudinal disaggregation of operational risk under changing regulations: evidence from the savings and loan industry. Academy of Management Journal, 40(4), 799–830. Xiaodong, L., & Lee, L. F. (2013). Two stage least squares estimation of spatial autoregressive models with endogenous regressors and many instruments. Econometric Reviews, 32(1), 734– 753.

Chapter 7

Moderating Role of Governance in Risk-Disclosure Relationship

If you can’t describe what you are doing as a process, you don’t know what you are doing. —W. Edwards Deming

Abstract This chapter aims to explore the relationships among risk disclosure index, risk governance index and risk index. In view of the possible endogeneity problem, diff-GMM regression has been used. The empirical analysis reveals that current risk levels of a company are significantly and positively associated with risk levels of immediately preceding year. It is noteworthy that risk governance index acts as a moderating variable, influencing the relationship between disclosure index and risk levels. In addition, current disclosure levels are significantly and positively associated with previous year’s disclosure. This supports the view of symbolic rather than substantive disclosures.




Keywords Risk disclosure Risk index Diff-GMM Moderating variable




7.1


Risk governance
Endogeneity


Introduction

Company is an artificial person; therefore, it cannot function on its own, its direction and control are in the hands of its directors. They are the lifeline of the company. Evidently, they are responsible for its financial and operational health. Further, the responsibility of protecting the company from internal and external risks vests with the Board. In other words, there is a close relationship between corporate governance and risk management. The process of risk management can be explained through five steps: identifying and prioritising risks, quantifying risks, managing risks, reporting risks and reviewing risks. The first step of the process is of utmost importance as the determination of companies’ strengths, weaknesses, opportunities and threats depends on it. Once the risks have been identified, it is imperative to quantify them. Due to information asymmetry between management and shareholders, it becomes © Springer Nature Singapore Pte Ltd. 2019 M. V. Shivaani et al., Understanding Corporate Risk, India Studies in Business and Economics, https://doi.org/10.1007/978-981-13-8141-6_7

223

224

7 Moderating Role of Governance in Risk-Disclosure Relationship

imperative to communicate risks to different stakeholders (government, employees, investors, etc.). Further, effective risk disclosures may avert agency costs. In recent times, regulatory bodies have also turned their attention to ‘disclosure and reporting’ practices followed by firms, in general and listed firms, in particular. It is reasonable to believe that transparent and fair disclosures facilitate informed decision-making among stakeholders. To keep pace with this ever-changing world, periodic and timely review of the process is now reckoned as a virtual necessity. It is noteworthy that the initial step, risk measurement has been captured by constructing a ‘risk index’ (RI); the reporting part has been presented by developing a ‘risk disclosure index’ (RDI). Both these aspects have been analysed in conjunction with the risk management authority, i.e. ‘risk governance’, by developing the ‘risk governance index’ (RGI). This chapter attempts to analyse the interrelationships among these three pillars of risk management. The relationships are proposed to be explored by employing the ‘difference generalised method of moments’ (GMM) technique and by constructing crosstabs. Further, the genesis of these relationships and the rationale for their examination may be better understood through Fig. 7.1. Figure 7.1 presents a holistic view of risk management process prevalent in organisations. The chapter has been organised into seven sections. Section 7.2

Fig. 7.1 Framework of interrelationships among various dimensions of risk

7.1 Introduction

225

outlines the background of interrelationships among risk disclosure index (RDI), risk governance index (RGI) and risk index (RI) and puts forth the hypotheses. Section 7.3 describes the sample used and sources of data. Section 7.4 elaborates the methodology employed for exploring the interrelationships among three variables. Section 7.5 presents the analysis and findings of the same. This is followed by Sect. 7.6, listing implications for various stakeholders. Lastly, Sect. 7.7 gives concluding observations.

7.2 7.2.1

Background and Hypothesis Development Risk Index (RI) and Risk Governance Index (RGI)

Idiosyncratic risk or firm-specific risk has been the focus of many researches (Fama and French 1993; Campbell et al. 2001). This may be attributed to the significant role that idiosyncratic risk plays. Without effective and efficient management of the risks, a firm may not even survive. Similarly, the implications of firm-specific risks for investment decisions by prospective investors cannot be overstated (Bali and Cakici 2008). Further, firm-specific risks are important to almost all the stakeholders from creditors to investors to regulators. Recognising this importance, attempts have been made to understand the factors that influence these risks. Amongst others, earnings growth (Xu and Malkiel 2003), age of firms (Bali and Cakici 2008; Fink et al. 2010), cash flow volatility (Irvine and Pontiff 2009), growth options (Cao et al. 2008), information quality (Chen et al. 2011; Rajgopal and Venkatachalam 2011) are some factors that have been found to be significant in explaining firm-specific risks. Abu-Ghunmi et al. (2015) emphasise that a plausible explanation for idiosyncratic/unsystematic risk could be looked into corporate governance mechanisms. Their arguments are in line with those of Baxter and Cotter (2009), Davidson et al. (2005), who found that composition of audit committee and proportion of non-executive directors, are significant contributors towards improvement in earnings quality. In a similar study, Huang et al. (2015) emphasised the importance of board size for firms’ risk management. Unlike most studies, that have linked specific governance indicators with firm’s riskiness, Jiraporn et al. (2015) explored the relationship between governance and risk by using a composite governance indicator. They consider two contrasting hypotheses; first, risk avoidance hypothesis and second, risk-seeking hypothesis. In the context of risk avoidance view, they posit that weak governance structures result in lower risk taking. They explain this proposition by emphasising on managerial risk aversion. Since managers’ human capital and large chunk of wealth is tied up in the firm, they may tend to play safe and avoid risky alternatives. Further, if the firm has weaker governance structure, these risk averse managers may have leeway in decision-making process, resulting in less risky decisions. In contrast, in the context

226

7 Moderating Role of Governance in Risk-Disclosure Relationship

of risk-seeking view, Jiraporn et al. (2015) submit that weaker governance leads to higher risk taking. They conjecture that strong and effective governance is expected to protect shareholders from unnecessary risk taking and weaker governance is less likely to keep managers from assuming too much risk. Further, managers have greater authority in weaker governance structures, which is likely to result in extreme decisions. These unbalanced decisions may, in turn, lead to more variability in firm performance—an indicator of higher risk (Adams et al. 2005). Pursuing the risk-seeking hypothesis Lee et al. (2016) suggest that better governance may reduce firm-specific risks. It is noteworthy that there are empirical evidences and studies focussing on corporate governance and risk, but, almost all of them have used market-based measures like variance of market-adjusted returns for stock, as proxy for idiosyncratic risk. Also, in terms of corporate governance, they have either looked at specific governance mechanisms (like, CEO duality, board size, audit committee composition, etc.) or existing composite governance measures like GIM-index, developed by Gompers et al. (2003). It is pertinent to note that corporate governance has varying definitions and encompasses a plethora of variables, whose relevance varies as per the context. In the context of risk and risk taking, the focus needs to be put on a specialised subset of corporate governance, called risk governance. Risk governance has been defined as ‘the ways in which directors authorize, optimize, and monitor risk taking in an enterprise. It includes the skills, infrastructure (i.e. organization structure, controls and information systems), and culture deployed as directors exercise their oversight’ (International Finance Corporation (IFC) 2012). Since, risk governance is the specialised arm of corporate governance that deals exclusively with risk and risk management, it is reasonable to believe that the quality of governance structure would have an impact on the risk/risk exposure levels of the company. Pursuing this stream of thought, the study proposes to explore the relationship between risk governance structure and risk levels. It is worth stating that the proxy for risk governance structure is the governance index constructed in Chap. 5. It contains the eleven variables namely, size of board, board diversity in terms of gender, proportion of executive directors, executive/ non-executive status of Chairperson, proportion of independent directors, CEO duality, Chief risk officer (CRO), risk management committee, mandatory committees, voluntary committees and existence/non-existence of whistle blower policy. Similarly, the risk is captured by the risk index constructed in Chap. 3. The risks that have been considered are: market risk, accounting risk, competition risk, contingency risk, credit risk, exchange rate risk, liquidity risk, operating risk and solvency risk. It is noteworthy that higher risk index denotes higher risk levels and higher value of governance index indicates better quality. Therefore, the first hypothesis is: H1: Risk (exposure) index is likely to be negatively related to the quality of risk governance.

7.2 Background and Hypothesis Development

7.2.2

227

Risk Disclosure Index (RDI) and Risk Index (RI)

As mentioned in the preceding paragraphs, idiosyncratic risks are important for almost all the stakeholders, therefore, their communication assumes greater importance. Annual reports are considered to be the most reliable and pandemic source of corporate communication. Considering the outreach of annual reports, it becomes imperative to gauge the relationship between risk disclosures in annual reports and firms’ risk levels. Dobler et al. (2011) postulate two contrasting hypotheses to explain the relationship between risk and risk disclosures. In the first hypothesis that is termed as disclosure motive, they suggest that firms with higher levels of risk may be inclined to provide more risk disclosure in order to explain the causes and management of that higher risk. The genesis of this view could be found in legitimacy theory, which indicates that riskier firms will disclose more to gain legitimacy. Further, agency theory and signalling theory also suggest that managers of riskier firms may be inclined to provide more details of risk. This could be viewed in the light of the fact that directors of riskier firms have greater responsibility to explain the causes of high risk. In addition, these directors may want to signal to stakeholders as to how well they are managing the risks. In the same vein, Ahn and Lee (2004) resort to stakeholder theory to suggest a positive relation between risk and risk disclosure levels. They opine that the company owes an explanation to stakeholders about the functioning and well-being of the company. In the second hypothesis, termed as concealing motive, Dobler et al. (2011) posit that riskier firms will attempt obscure disclosures in order to avoid disclosure costs. Similarly, Linsley and Shrives (2006) suggest that such firms will be reluctant in making risk disclosures as they would not want to attract undue attention towards their riskiness. In contrast, less risky firms may want to disclose more to highlight their lower levels of risk as well as their risk management systems. Further, signalling theory also suggests that riskier firms may not want to signal the inefficiency of their risk management systems by highlighting their risks. In accordance with the two contrasting motives of risk disclosure, literature provides mixed evidence on the issue. Certain studies, such as the one Alexander (1996), have found a positive and significant association between risk and disclosure levels. Whereas, certain other studies (Marshall and Weetman 2002) have noted a negative relationship. Further, a large number of studies indicate no relationship between risk and risk disclosure levels (Lev and Penman 1990). Although there are a substantial number of empirical evidences and studies focussing on risk and risk disclosures, yet, almost all of them have used single risk measures, i.e. either beta (Linsley and Shrives 2006), or leverage (Ahn and Lee 2004) as proxy for idiosyncratic risk. Also, in terms of risk disclosure, they have seldom considered quality of risk disclosures. In addition, no particular study could be found that has explicitly examined the relationship between quality of risk disclosure and a comprehensive risk measure.

228

7 Moderating Role of Governance in Risk-Disclosure Relationship

Therefore, this study attempts to explore the relationship between RDI and RI, using the RDI developed in Chap. 4 and RI developed in Chap. 3. It is worth restating that RDI is based on 69 risk items and focuses on three semantic attributes, namely, nature of disclosure (qualitative/quantitative), tense (backward-looking/ forward-looking/both, backward-looking and forward-looking) and tone (good news/bad news/both, good and bad news). In sum, following the mixed evidence from literature, the hypothesis is: H2: Risk disclosure index is not likely to be related to risk index.

7.2.3

Risk Disclosure Index (RDI) and Risk Governance Index (RGI)

Almost all the stakeholders attempt to elicit relevant corporate information. Such information would enable them to assess risk levels of companies and the efficiency with which the companies are managing these risks (Lajili and Zehgal 2005). But, separation of ownership and control leads to information asymmetry (Jensen and Meckling 1976). This information asymmetry may be reduced by increasing transparency and disclosing more information. It is worth noting that the most easily available and most credible source of corporate information is annual report. Disclosures in annual reports are largely the responsibility of the Board. Therefore, certain Board characteristics may be expected to have a significant impact on levels of risk disclosures. Bozec and Bozec (2012) and Elzahar and Hussainey (2012) quote agency theory to suggest that there may be a positive association between board size and risk disclosure levels. In contrast, Jensen (1993) suggests that, due to poor coordination among large boards, they may have a negative impact on disclosure levels and firm performance. Similarly, presence of independent directors may lead to increased transparency and accountability, translating into higher risk disclosures (Lopes and Rodrigues 2007). Abraham and Cox (2007) assert that the presence of non-executive directors on board often leads to reduced agency costs and increased motivation for voluntary disclosures. Further, there has been mixed evidence in terms of CEO duality and levels of risk disclosures. On the one hand, Gul and Leung (2004) observe a negative relation between CEO duality and levels of disclosures and, on the other hand, Cheng and Courtenay (2006) observe no such relationship. O’Sullivan and O’Dwyer (2009) subscribe to resource-dependency theory to explain the relation between risk disclosures and governance levels. They suggest that reputation is a critical resource for firm’s growth; and firms may improve their reputation by increasing their visibility and by coming across as responsible and transparent organisations. Therefore, strong governance structures, which are expected to be aware of reputational risks, may provide greater risk disclosures. Further, as per legitimacy theory, to gain legitimacy in the eyes of stakeholders, firms may want to portray a transparent and accountable image. If the risk

7.2 Background and Hypothesis Development

229

governance structures are robust, they would recognise the need to communicate effectively with the stakeholders and would have more comprehensive risk disclosures. In addition, firms with strong governance structures may be expected to be more effective and efficient in managing their risks. Therefore, they would want to apprise their stakeholders of their robust risk handling strategies and may hence disclose more. Most of the studies like that of Ntim et al. (2013) have examined the relationship between risk disclosures and specific governance characteristics like gender diversity on Board, CEO duality, etc. Further, most of the risk disclosure studies have used a dichotomous scale for constructing the risk disclosure index. Such a methodology neglects quality of disclosures. In addition, no particular study could be found that has explicitly examined the relationship between quality of risk disclosure and a comprehensive risk governance measure. In view of the above, this study attempts to explore the relationship between RDI and RGI, using the RDI developed in Chap. 4 and RI developed in Chap. 3. Therefore, based on the literature, the hypothesis is: H3: Risk disclosure index is likely to be positively related to quality of governance structure

7.3

Sample

The sample consists of non-financial companies that constitute NIFTY 500 index as on 31 March 2014. The study covers a period of 10 years from 1 April 2005 to 31 March 2015. The data has been obtained from databases, namely, Bloomberg and Ace-equity and companies’ annual reports.

7.4

Methodology

As the study covers 429 (non-financial) companies over a period of 10 years (2005–2015), the data is panel data in nature. Since, the data for all the companies is not available for all the years; it constitutes an ‘unbalanced panel’. Further, as the number of cross sections (429) is more than the number of time periods (10), it is a ‘short panel’. The analysis makes use of the following variables: Risk Index (RI)—It is the risk index developed in Chap. 3. It is a proxy for the risk level in a company. Risk Disclosure Index (RDI)—It is the risk disclosure index as per Chap. 4. It captures the quality and quantity of risk disclosures in companies’ annual reports. Risk Governance Index (RGI)—It is the governance index constructed in Chap. 5. It attempts to measure the quality of risk governance structures that exist in companies.

230

7 Moderating Role of Governance in Risk-Disclosure Relationship

Control variables: It is often observed that there are certain variables that extraneously affect the relationship of interest. Therefore, in order to uncover the true relationship between the relevant variables, it is pertinent to control for non-focal factors (Carlson and Wu 2012; Spector and Brannick 2011). Similarly, Dewey (2002) and Popper (1966) assert that to have valid inferences, it is imperative to control (or hold constant) all the variables, other than those being investigated. Based on literature review, the following four control variables have been incorporated in the regression models used in this chapter. Age—Age may influence the idiosyncratic risk of firms as well as the governance levels (Luo and Bhattacharya 2009; Mishra and Modi 2013; Bebchuk et al. 2008). It has been measured as the number of years a firm has been in existence (since its inception). Growth—John et al. (2008) observe a positive and significant relation between growth and risk levels. Similarly, Jiraporn et al. (2015) examine the relationship between growth and disclosure levels. Annual growth in sales has been used as a proxy for firm’s growth. Size—Oliveira et al. (2011) observe a positive and significant relationship between size and RDI. Whereas, Jiraporn et al. (2015) find a negative and significant relationship between the two. It is worth noting that Li et al. (2013) did not observe any relation between the size of firm and its risk. It has been measured using the natural log of total assets of the firm. Recession—Edkins (2009) examined the impact of recession on disclosure levels of banks. It has been captured by the introduction of a dummy variable for pre-recession period (2005–2008) and post-recession period (2008–2015). Given the panel nature of data, panel data regression appears to be the appropriate technique. Panel data analysis provides several advantages over pooled OLS regression. It facilitates consideration of individual/firm-specific heterogeneities that may be having an impact on the dependent variable, provides more informative data, more degrees of freedom and more efficiency (Hsiao 2003; Baltagi 2008). The most common approach to deal with dynamic panels and possible endogeneities is to use dynamic panel data generalised method of moments (GMM) regression. Further, the decision to use Arellano and Bond (1991), ‘difference GMM’, for the current study is based on the findings of Larcker and Rusticus (2010) and Petersen (2009). They suggest that companies are unique in terms of their strengths and weaknesses. This can result in a scenario whereby disclosure and governance practices are jointly and dynamically determined by unobserved company-specific heterogeneities, such as managerial talent, corporate culture and complexity (Guest 2009; Henry 2008), which simple OLS regressions may be unable to detect (Gujarati 2003; Wooldridge 2010). In addition, Hermalin and Weisbach (2003), Himmelberg et al. (1999), Davis and Kruse (2000), and Wintoki and Yang (2007) argue that firm performance and corporate governance are simultaneously

7.4 Methodology

231

determined by unobservable firm-specific factors, and that governance changes are determined by past, present and/or expected characteristics of the firm. Hence, given the panel nature of data and following past investigations, the study proposes to use ‘difference GMM’ for estimating various relationships.

7.5

Empirical Evidence

7.5.1

Descriptive Analysis

Table 7.1 presents the descriptive statistics of the variables of interest: The sample exhibits sizeable heterogeneity in terms of age, size and growth. The mean risk disclosures are around 11% (only), indicating a lack of transparency in risk reporting. In contrast, mean risk governance index of 65% is suggestive of fairly strong governance structures in Indian companies. In addition, the mean risk level, which is about 45%, indicates that Indian companies, by and large, have moderate risk levels. To explore the relationships among RI, RDI and RGI, five models have been proposed. These models have been explained and estimated in the following subsections:

7.5.2

Relationship Between Risk Index and Risk Governance Index (Model 1)

The model attempts to examine the factors influencing risk index, with a particular focus on the quality of risk governance structure. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: RIit ¼ a þ b1 RIit1 þ b2 RGIit þ b3 Ageit þ b4 Sizeit

ð7:1Þ

þ b5 Growthit þ b6 Recession dummyit þ eit

Table 7.1 Descriptive statistics of relevant variables for the period 2005–2015 Particulars

N

Minimum

Maximum

Mean

Std. deviation

Age Growth Size RDI RGI RI

3301 3301 3301 3301 3301 3301

2 −100.00% 4.45 1.16% 40.00% 22.73%

152 110.49% 14.71 36.91% 86.73% 78.18%

37.895 12.76% 7.87 11.37% 65.50% 45.06%

24.4003 2.56% 1.41 6.20% 7.99% 8.18%

232

7 Moderating Role of Governance in Risk-Disclosure Relationship

Table 7.2 Results of (Arellano–Bond) GMM estimation of RI on the first lag of RI, RGI and control variables Variables

Coefficient

RI 0.2498 RGI −0.0167 Age 0.0030 Size 0.0009 Growth 0.0003 Recession dummy 0.0061 Constant 0.2205 Number of observations: 2362 Number of instruments: 42 v2(26) v2(35) First order Second order Note *** denotes significance at 1% Wald test Sargan test Autocorrelation

Standard error

p-value

3.1700 −0.5800 7.5500 0.1900 3.2100 1.6300 4.9900

0.0020*** 0.5630 0.0000*** 0.8510 0.0010*** 0.1030 0.0000***

Statistic

p-value

143.0900 49.1357 −5.8732 0.2478

0.0000*** 0.0569 0.0000 0.8043

The results of the estimation have been presented in Table 7.2. Table 7.2 reveals that the first lag of risk index is significantly and positively related with current levels of risk index. In other words, the risk levels of companies are significantly increasing year after year. In addition, it could be viewed as an indicator of ineffective and inefficient risk management. If the risk in the previous period had been managed well, it would not have permeated to and manifested in the current year’s risk levels. Further, the main variable of interest, quality of risk governance, exhibits a negative relationship with the risk index. This suggests that better governance results in lower risk levels and weaker governance leads to higher risk levels. These results support the risk-seeking view (Jiraporn et al. 2015). Though the sign of coefficient is in tune with the intuition, the relationship is not statistically significant (p-value > 0.05). In terms of control variables, age and growth opportunities exhibit a significant and positive relationship with risk index. It is noteworthy that the p-value of 0.000 for Wald test indicates that the model used is a good fit. The GMM specifications are well-specified based on the Sargan test of over-identifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors.

7.5.3

Relationship Between Risk Disclosure Index and Risk Governance Index (Model 2)

The model attempts to examine the factors influencing risk disclosure index, with a particular focus on the quality of risk governance structure. Using Arellano–Bond

7.5 Empirical Evidence

233

Table 7.3 Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RGI and control variables Variables

Coefficient

RDI (−1) 0.3301 RGI 0.0722 Age 0.0007 Growth 0.0001 Size 0.0063 Recession dummy 0.0012 Constant −0.0463 Number of observations: 2362 Number of instruments: 42 v2(26) v2(35) First order Second order Note *** denotes significance at 1% level Wald test Sargan test Autocorrelation

Standard error

p-value

0.0942 0.0173 0.0003 0.0001 0.0023 0.0017 0.0180

0.0000*** 0.0000*** 0.0080*** 0.3670 0.0050*** 0.4740 0.0100***

Statistic

p-value

124.1700 37.4282 −4.1243 −0.2507

0.0000*** 0.3583 0.0000 0.8021

(1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: RDIit ¼ a þ b1 RDIit1 þ b2 RGIit þ b3 Ageit þ b4 Sizeit þ b5 Growthit þ b6 Recession dummyit þ eit

ð7:2Þ

The results of the estimation have been presented in Table 7.3. Table 7.3 reveals that the first lag of risk disclosure index is significantly and positively related with current levels of risk disclosure index. In other words, companies’ current risk disclosures are significantly related to immediately preceding year’s risk disclosures. The results are in tune with that of Abraham and Shrives (2014), who suggests that in order to confirm the authenticity and coherence of disclosures, they should be viewed in light of previous year’s disclosures. Further, the main variable of interest, quality of risk governance, exhibits a positive and significant relationship with the risk disclosures. This suggests that that better (poorer) governance results in higher (lower) disclosure levels. These results support the disclosure motive (Dobler et al. 2011). The results may also be viewed in light of legitimacy theory and stakeholder theory. As per legitimacy theory, Directors may be inclined to disclose more to garner investors’ confidence. In the context of stakeholder theory, if the quality of governance structure is superior, the Directors may be considered more conscious of their responsibility towards various stakeholders, which gets reflected through greater transparency. Further, the results are in tune with those of other studies like, Alexander (1996) and Abraham and Cox (2007), who found a significant

234

7 Moderating Role of Governance in Risk-Disclosure Relationship

relationship between disclosure levels and governance. In terms of control variables, age, growth opportunities and size of firm, (Linsley and Shrives 2006) exhibit a significant and positive relationship with risk disclosure index. It is noteworthy that the p-value of 0.000 for Wald test indicates that the model used is a good fit. The GMM specifications are well-specified based on the Sargan test of over-identifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors.

7.5.4

Relationship Between Risk Disclosure Index and Risk Index (Model 3)

The model attempts to examine the factors influencing risk disclosure index, with a particular focus on the levels of risk. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: RDIit ¼ a þ b1 RDIit1 þ b2 RIit þ b3 Ageit þ b4 Sizeit þ b5 Growthit þ b6 Recession dummyit þ eit

ð7:3Þ

The results of the estimation have been presented in Table 7.4. As in the case of Model 2, the lag of risk disclosure index is significantly and positively related to current disclosures levels in this model also (Table 7.4). But, a Table 7.4 Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RI and control variables Variables

Coefficient

RDI 0.3216 RI 0.0089 Age 0.0009 Growth 0.0000 Size 0.0075 Recession dummy 0.0010 Constant −0.0174 Number of observations: 2362 Number of instruments: 42

Standard error

p-value

3.5200 0.8300 2.3300 0.4000 3.2200 −0.5900 −1.0400

0.0000*** 0.4080 0.0200** 0.6860 0.0010** 0.5550 0.2980

v2(26) v2(35) First order Second order Note ***, ** denote significance at 1% and 5% level Wald Sargan test Autocorrelation

Statistic

p-value

96.8200 34.8052 −4.1591 −0.2733 respectively

0.0000*** 0.4775 0.0000 0.8124

7.5 Empirical Evidence

235

p-value > 0.05 for the variable -risk index suggests that risk levels do not have any significant impact on the quality of disclosures. It is noteworthy, that a positive coefficient on risk index definitely indicates that riskier firms tend to disclose more, but the relationship is not significant (Dobler et al. 2011). This positive relationship may also be viewed in the context of agency theory and signalling theory that suggests that managers of riskier firms need to disclose more in order to explain the causes and management of these higher risks. Further, the non-significant results are in tune with those of Lajili and Zehgal (2005), Linsley and Shrives (2006) and Amran et al. (2008). In terms of control variables, both age and size of firm exhibit a significant and positive relationship with risk disclosure index. These results are similar to those of (Black et al. 2006). It is noteworthy that the p-value of 0.000 for Wald test indicates that the model used is a good fit. The GMM specifications are well-specified based on the Sargan test of over-identifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors.

7.5.5

Relationship of Risk Disclosure Index with Risk Index and Risk Governance Index (Model 4)

In order to gain a holistic view of the factors influencing RDI, it is reasonable to introduce both the variables, RI and RGI simultaneously. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: RDIit ¼ a þ b1 RDIit1 þ b2 RIit þ b3 RGIit þ b4 Ageit þ b5 Sizeit þ b6 Growthit þ b7 Recession dummyit þ eit

ð7:4:Þ

In line with the results of Model 2 and Model 3, the lag of RDI continues to be positively and significantly related to current RDI levels (Table 7.5). Similarly, RI continues to have a positive and non-significant relationship with RDI. Whereas, RGI consistently exhibits a positive and significant relationship with RDI. Even in the context of control variables, just like Models 2 and 3, age, as well as size of firm, turns out to be positively and significantly related to RDI. It is noteworthy that the p-value of 0.000 for Wald test indicates that the model used is still a good fit. The GMM specifications are well-specified based on the Sargan test of over-identifying restrictions. Moreover, the Arellano–Bond test statistics indicate that there exists no autocorrelation in the errors. Further, an attempt has been made to explore these relationships graphically. As a first step, the risk index has been divided into four parts, based on the quartiles. In the second step, risk governance scores corresponding to each of the four quartiles of RI have been subdivided into four parts (based on quartiles) each. In other words, the governance scores corresponding to Q1 of risk index have been divided on the

236

7 Moderating Role of Governance in Risk-Disclosure Relationship

Table 7.5 Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RI and RGI and control variables Variables

Coefficient

RDI 0.3355 RI 0.0094 RGI 0.0726 Age 0.0007 Growth 0.0001 Size 0.0062 Recession dummy 0.0012 Constant −0.0489 Number of observations: 2362 Number of instruments: 43

Standard error

p-value

3.5600 0.8800 4.2100 2.4800 0.8900 2.7300 −0.7500 −2.6600

0.0000*** 0.3810 0.0000*** 0.0130** 0.3740 0.0060*** 0.4560 0.0080***

v2(26) v2(35) First order Second order Note ***, ** denote significance at 1% and 5% level Wald test Sargan test Autocorrelation

Statistic

p-value

126.0500 37.7053 −4.1519 −0.2571 respectively

0.0000*** 0.3465 0.0000 0.7971

basis of quartiles. Similarly, the governance scores corresponding to Q2 of risk index have been divided into four parts, based on the quartiles. The same procedure has been repeated for governance scores corresponding to Q3 and Q4 of the risk index. Step two yields 16 data points. Corresponding to each of these series, the mean of RDI has been computed. The methodology has been illustrated in Exhibit 7.1. Further, the computed RDI means have been plotted in Fig. 7.2. Given the specifications of Model 4 and based on the estimations of Table 7.5, a positive relationship has been established between RDI and RI, as well as RDI and RGI. In other words, with increasing RI, an increase in RDI may be expected. In graphical terms, if y-axis represents RDI and x-axis represents RI, a positive sloping, upward line may be expected. A similar relationship may be expected for RDI and RGI. At this point, it is pertinent to revisit the interpretation of regression coefficients. In multiple regressions, with two explanatory variables X1 and X2 with their corresponding betas being b1 and b2, b1 shows the relationship between X1 and the dependent variable, keeping X2 constant. Similarly, b2 shows the relationship between X2 and the dependent variable, keeping the level of X1 constant. Pursuing this theoretical premise, for a given level of risk say Q1, the relationship between RDI and RGI should be positive (due to positive coefficient −0.0726, in Table 7.5). Graphically, this should result in a hierarchy of each of the four RGI points (where each RGI point corresponds to Q1, Q2, Q3 and Q4 of RGI). In other words, the point corresponding to Q1 of RGI will be the lower most point (depicting correspondingly lower disclosures), above to it will be the point corresponding to Q2 of

7.5 Empirical Evidence

237

Exhibit 7.1 Methodology for computation mean RDI corresponding to various levels of RI and RGI Risk index

Risk governance index

Risk disclosure index

Quartile (Q)1 Q1 Q1 Q1 Q2 Q2 Q2 Q2 Q3 Q3 Q3 Q3 Q4 Q4 Q4 Q4

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4

Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding Corresponding

mean mean mean mean mean mean mean mean mean mean mean mean mean mean mean mean

of of of of of of of of of of of of of of of of

RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI RDI

Relationship among RDI, RI and RGI 16.00%

Mean risk disclosure index (RDI)

15.00%

14.00%

13.00% Indicative of interaction Q1of RGI

12.00%

Q2 of RGI Q3 of RGI

11.00%

Q4 of RGI

10.00%

9.00%

8.00%

Q1 of RI

Q2 of RI

Q3 of RI

Risk index (RI)

Fig. 7.2 Exploration of relationship among RDI, RI and RGI

Q4 of RI

238

7 Moderating Role of Governance in Risk-Disclosure Relationship

RGI; likewise, above it will be the point corresponding to Q3 of RGI and the top most point will be corresponding to Q4 of RGI. This relationship is evident at Q1, Q2 and Q3 level of risk. But, beyond Q3 level of RI, it has been observed that lines corresponding to Q2 of RGI and Q3 of RGI intersect or cross each other. Further, beyond this point of intersection, it has been observed that mean RDI for Q2 of RGI is higher than mean RDI for Q3 of RGI. These graphical results are contrary to the empirical evidence that exhibit a clear positive relation between RDI and RI, as well as RDI and RGI. In other words, the point of intersection in the graph indicates that it may be unreasonable to examine the RDI and RI relationship or RDI and RGI relationship by keeping RGI and RI constant, respectively. This is suggestive of some interaction between RI and RGI, which may be having an impact on RDI levels. In technical terms, RGI may be acting as a moderating variable, influencing the relationship between RDI and RI. It is noteworthy that a moderating variable has been defined as ‘a variable that affects the direction and/or strength of the relation between an independent, or predictor, variable and a dependent, or criterion, variable’ (Baron and Kenny 1986). Further, ‘moderation refers to the examination of the statistical interaction between two independent variables in predicting a dependent variable’ (Jose 2013). This possibility could be empirically explored by introducing an interaction term in Model 4. In operational terms, an interaction term could be viewed as RI * RGI (product of RI and RGI). Therefore, to explore the possible moderating effects of RGI on the relationship between RDI and RI, an expanded model, Model 5 has been proposed.

7.5.6

Relationship of Risk Disclosure Index with Risk Index and Risk Governance Index While Factoring for Moderating Effects of Risk Governance Index (Model 5)

To explore the moderating effects of RGI on the relationship between RDI and RI, it appears reasonable to introduce an interaction term (RI * RGI) in Model 4. Using Arellano and Bond (1991) ‘difference GMM’, the relationship has been examined by estimating the following regression equation: RDIit ¼ a þ b1 RDIit1 þ b2 RIit þ b3 RGIit þ b4 Ageit þ b5 Sizeit þ b6 Growthit þ b7 Recession dummyit þ b8 ðRI  RGIÞit þ eit

ð7:5:Þ

The results of the estimation have been presented in Table 7.6. In tune with the results of Models 2, 3 and 4 the lag of RDI continues to be significantly and positively associated with current levels of RDI (Table 7.6). Similarly, even in this model, RGI exhibits a positive and significant relationship with RDI.

7.5 Empirical Evidence

239

Table 7.6 Results of (Arellano–Bond) GMM estimation of RDI on the first lag of RDI, RI, RGI and (RI * RGI) and control variables Variables

Coefficient

RDI 0.3360 RI 0.1603 RGI 0.1804 Interaction −0.2319 Age 0.0007 Growth 0.0001 Size 0.0060 Recession dummy 0.0015 Constant −0.1192 Number of observations: 2362 Number of instruments: 44

Standard error

p-value

3.5500 1.9700 2.8700 −1.8300 2.7400 1.1000 2.6100 −0.9100 −2.8200

0.0000*** 0.0490** 0.0040*** 0.0670* 0.0060*** 0.2720 0.0090*** 0.3610 0.0050***

Statistic v2(26) 129.5500 v2(35) 37.3117 First order −4.1737 Second order −0.2964 Note ***, **, * denote significance at 1%, 5% and 10% level respectively Wald test Sargan test Autocorrelation

p-value 0.0000*** 0.3632 0.0000 0.7669

But, the noteworthy finding from Table 7.6 is that unlike Model 1 and Model 4, risk index is exhibiting a significant (and positive) relationship with RDI. This could clearly be attributed to the introduction of the interaction term. Further, the interaction term also exhibits a significant relation with RDI. These findings are a clear evidence of RGI acting as a moderating variable in the relationship between RDI and RI. The new modified relationship has been explored graphically by following the methodology postulated by Jose (2013). He recommends the estimation of following nine equations and subsequent use of the respective outputs for plotting the graph (Fig. 7.3). It is noteworthy that betas used in these nine equations are the ones derived from Model 5. 1. High RI and High RGI ½bRIðMean þ Standard deviationðSDÞ þ ½bRGI ðMean þ SDÞ þ ½bInteraction½ðMean þ SD of RI ÞðMean þ SD of RGI Þ þ constat term½0:1603ð0:4505 þ 0:08179Þ þ ½0:1804ð0:6549 þ 0:0798Þ þ ½0:2319½ð0:4505 þ 0:08179Þð0:6549 þ 0:0798Þ þ ð0:1192Þ ¼ 0:00800 ð7:6Þ

240

7 Moderating Role of Governance in Risk-Disclosure Relationship

Moderating effect of RGI on the RI-RDI relationship Ris disclosure index (RDI)

1.20% 1.00% 0.80% 0.60% 0.40% 0.20%

High RGI

0.00%

Medium RGI

-0.20%

Low RGI

-0.40% -0.60% -0.80%

Low RI

Medium RI

High RI

Risk index (RI) Fig. 7.3 Moderating effect of RGI on relationship between RDI and RI

2. High RI and Medium RGI ½bRI ðMean þ SDÞ þ ½bRGI ðMeanÞ þ ½bInteraction½ðMean þ SD of RI ÞðMean of RGI Þ þ constat term½0:1603ð0:4505 þ 0:08179Þ þ ½0:1804ð0:6549Þ

ð7:7Þ

þ ½0:2319½ð0:4505 þ 0:08179Þð0:6549Þ þ ð0:1192Þ ¼ 0:00346 3. High RI and Low RGI ½bRI ðMean þ SDÞ þ ½bRGI ðMean  SDÞ þ ½bInteraction½ðMean þ SD of RI ÞðMean  SD of RGI Þ þ constat term½0:1603ð0:4505 þ 0:08179Þ þ ½0:1804ð0:6549  0:0798Þ þ ½0:2319½ð0:4505 þ 0:08179Þð0:6549  0:0798Þ þ ð0:1192Þ ¼ 0:0010 ð7:8Þ 4. Medium RI and High RGI ½bRI ðMeanÞ þ ½bRGI ðMean þ SDÞ þ ½bInteraction½ðMean of RI ÞðMean þ SD of RGI Þ þ constat term½0:1603ð0:4505Þ þ ½0:1804ð0:6549 þ 0:0798Þ þ ½0:2319½ð0:4505Þð0:6549 þ 0:0798Þ þ ð0:1192Þ ¼ 0:00883

ð7:9Þ

7.5 Empirical Evidence

241

5. Medium RI and Medium RGI ½bRI ðMeanÞ þ ½bRGI ðMeanÞ þ ½bInteraction½ðMean of RI ÞðMean of RGI Þ þ constat term½0:1603ð0:4505Þ þ ½0:1804ð0:6549Þ

ð7:10Þ

þ ½0:2319½ð0:4505Þð0:6549Þ þ ð0:1192Þ ¼ 0:002780

6. Medium RI and Low RGI ½bRI ðMeanÞ þ ½bRGI ðMean  SDÞ þ ½bInteraction½ðMean of RI ÞðMean  SD of RGI Þ þ constat term½0:1603ð0:4505Þ þ ½0:1804ð0:6549  0:0798Þ þ ½0:2319½ð0:4505 of RI Þð0:6549  0:0798Þ

ð7:11Þ

þ ð0:1192Þ ¼ 0:0032733

7. Low RI and High RGI ½bRI ðMean  SDÞ þ ½bRGI ðMean þ SDÞ þ ½bInteraction½ðMean  SD of RI ÞðMean þ SD of RGI Þ þ constat term½0:1603ð0:4505  0:08179Þ þ ½0:1804ð0:6549 þ 0:0798Þ þ ½0:2319½ð0:4505  0:08179Þð0:6549 þ 0:0798Þ þ ð0:1192Þ ¼ 0:0096

ð7:12Þ 8. Low RI and Medium RGI ½bRI ðMean  SDÞ þ ½bRGI ðMeanÞ þ ½bInteraction½ðMean SD of RI ÞðMean of RGI Þ þ constat term½0:1603ð0:4505  0:08179Þ þ ½0:1804ð0:6549Þ

ð7:13Þ

þ ½0:2319½ð0:4505  0:08179Þð0:6549Þ þ ð0:1192Þ ¼ 0:002092

9. Low RI and low RGI ½bRI ðMean  SDÞ þ ½bRGI ðMean  SDÞ þ ½bInteraction½ðMean  SD of RI ÞðMean  SD of RGI Þ þ constat term½0:1603ð0:4505  0:08179Þ þ ½0:1804ð0:6549  0:0798Þ þ ½0:2319½ð0:4505  0:08179Þð0:6549  0:0798Þ þ ð0:1192Þ ¼ 0:001070

ð7:14Þ

242

7 Moderating Role of Governance in Risk-Disclosure Relationship

Risk governance index -0.2319

+0.1603

Risk index

Risk disclosure index

Fig. 7.4 Holistic view of relationship among RDI, RGI and RI

These computed values have been used to plot Fig. 7.3. As per Jose (2013), parallel lines indicate insignificant interaction. In other words, distinctly non-parallel lines are a clear indicator of a significant interaction. It is noteworthy that Fig. 7.3 shows distinctly non-parallel lines, thus confirming a significant interaction. Further, it can be seen that low RGI line has the steepest slope; this means that the relationship between risk index and risk disclosure index is strongest for the poorly governed companies. Further, the negative slope of High RGI line highlights the moderating effect. It indicates that in better governed companies, disclosure quality is deteriorating with increasing levels of risk. The results justify the negative coefficient (–0.2319) of interaction term. In sum, there is some interaction between RI and RGI that reverses the general positive relation between RI and RDI, for better governed and riskier firms. The final relationship among RI, RDI and RGI can be depicted through Fig. 7.4.

7.5.7

Further Analysis

To enhance the understanding of relationship between pairs of RI, RDI and RGI, crosstabs have been attempted. Cross tabulation is a statistical tool that facilitates presentation and examination of frequencies of variables in a two-dimensional setting. Usually, it involves the simultaneous consideration of two variables. It is worth mentioning that for the purpose of crosstabs, all three indices have been divided into five categories ranging from ‘very low’ to ‘very high’. The basis of categorisation of each of the indices has been presented in Exhibit 7.2. Exhibit 7.2 Methodology for categorisation of RDI, RGI, RI from ‘very low’ to ‘very high’ Particulars

RI (as a percentage of maximum possible score)

RGI (as a percentage of maximum possible score)

Very low Low Medium High Very high

RI = 20% 20% < RI 40% < RI 60% < RI 80% < RI

27% 40% 55% 70% 85%

   

40% 60% 80% 100%

< < < <