Credit Rating Agencies: Meeting the Needs of the

0 downloads 0 Views 1MB Size Report
ratings process; the perceived global domination of the market by two main players; the ... characteristics of rating agencies – ranging from technical features to ... products such as credit derivatives; (ii) criticisms of the CRA industry's seeming inability ..... analyses suggest this has been due more to a common perception that.
Credit Rating Agencies:

Meeting the Needs of the Market? Researchers: Angus Duff Sandra Einig

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

by

Angus Duff University of Paisley Sandra Einig Oxford Brookes University

Published by The Institute of Chartered Accountants of Scotland CA House, 21 Haymarket Yards Edinburgh EH12 5BH

First Published 2007 The Institute of Chartered Accountants of Scotland

© 2007 ISBN 978-1-904574-20-0 EAN 9781904574200

This book is published for the Research Committee of The Institute of Chartered Accountants of Scotland. The views expressed in this report are those of the authors and do not necessarily represent the views of the Council of the Institute or the Research Committee. No responsibility for loss occasioned to any person acting or refraining from action as a result of any material in this publication can be accepted by the author or publisher. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopy, recording or otherwise, without prior permission of the publisher.

Printed and bound in Great Britain by T. J. International Ltd.

C

ONTENTS

Foreword ..................................................................................... Acknowledgements ....................................................................... Executive Summary ..................................................................... Abbreviations .............................................................................. 1.

2.

3.

4.

i iii v xi

INTRODUCTION ............................................................

1

Background .................................................................. Purpose of the study and the research approach ............

1 6

CREDIT RATING AGENCIES ...........................................

13

A brief history of the ratings industry ............................ Ratings processes ........................................................... The demand for ratings services ..................................... Financial performance of credit rating agencies .............. Demands for self regulation of credit rating agencies ..... Barriers to entry ............................................................. Summary .......................................................................

13 14 15 16 18 29 32

LITERATURE REVIEW ....................................................

33

Market participants’ attitudes to credit rating agencies ... Audit theory and audit quality ....................................... Summary .......................................................................

33 38 47

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY ..........

49

Interviews ...................................................................... Questionnaire development ........................................... Technical features ......................................................... Service features ............................................................. Summary .......................................................................

49 51 54 59 63

CONTENTS

5.

6.

7.

8.

CONDUCT OF SURVEY ...................................................

65

Financial managers ........................................................ Other interested parties ................................................. Investors ........................................................................ Summary .......................................................................

65 69 70 73

RESULTS OF THE QUESTIONNAIRE SURVEY: ALL GROUPS ......

75

Characteristics ............................................................... Individual items ............................................................. Summary .......................................................................

75 78 85

COMPARISON BETWEEN GROUPS ..................................

87

Characteristics ............................................................... Items ............................................................................. Summary .......................................................................

87 95 101

CONCLUSIONS ..............................................................

103

Main findings and comments ........................................ 103 Policy implications ......................................................... 106 Further research ............................................................. 110 REFERENCES .........................................................................

113

APPENDIX .............................................................................

119

ABOUT THE AUTHORS ...........................................................

127

F

OREWORD

The recent turmoil in the international financial markets following the subprime mortgage crisis in the US has raised questions about the role of credit rating agencies in this crisis and as to whether credit rating agencies are meeting the needs of the market and if there is a need for increased regulation. This timely publication looks at the views of market participants before the crisis and provides an insight into the rapid development and practices of the agencies. The authors recognise the lack of research in this topical area and identify five areas for future research: to consider the relationship between the credit rating agency and the issuer; to identify whether an expectations gap exists; to gather the views of market participants in other countries, particularly in the US; to better understand the processes and methodology used by the agencies; and to see how market participants’ views have changed as a result of the turmoil encountered in the credit markets of 2007. This report commences with background information on the industry, including information on: the structure of the market; the ratings process; the perceived global domination of the market by two main players; the criticisms the agencies have faced since the late 1990s; and international calls for greater supervision and accountability of the agencies. Against this background commentators have questioned whether credit rating agencies are providing an adequate service and ‘quality ratings’. This research attempts to resolve some of these issues by gathering views of participants and interested parties to identify the components of ‘ratings quality’ and then comparing the views of three separate groups of market participants in the UK: corporate treasurers; investors; and other interested parties. The “other interested parties” group was made

ii

FOREWORD

up of finance academics, finance professionals, consultants, journalists and bankers. The report then analyses the findings from the interviews and questionnaires. The interviews were used to help identify 14 characteristics of rating agencies – ranging from technical features to service features which were used to design the survey questionnaire, which featured 71 questionnaire items. In the survey reputation, trust and values were ranked as the most important characteristic, with transparency and timeliness also ranked highly. The highest ranking items were focused on integrity, ethical standards and credibility as well as more competency based items such as the accuracy of ratings and qualification of staff. User groups did not place any importance on other services offered by the agencies. Users were confident in the independence of credit rating agencies they dealt with. Finally, the report concludes on the findings of the study and policy implications for the industry. The policy implications are that: there is no evidence to suggest a need for formal regulation of the agencies, as existing codes of conduct are considered to be adequate; the codes, and compliance with them, should be monitored closely; the agencies themselves need to maintain and improve their commitment to quality and improve their communication; agencies’ ancillary services should be formally separated from their ratings work; and any unsolicited ratings should be clearly identified as such. This project was funded by the Scottish Accountancy Trust for Education and Research (SATER). The Research Committee of The Institute of Chartered Accountants of Scotland (ICAS) has also been happy to support this project. The Committee recognises that the views expressed do not necessarily represent those of ICAS itself, but hopes that the project will contribute to current debate on credit rating agencies and their role in the financial markets. David Spence Convener of Research Committee October 2007

A

CKNOWLEDGEMENTS

The monograph benefits from comments of two anonymous reviewers, together with John Grout of the Association of Corporate Treasurers and Christine Helliar of The Institute of Chartered Accountants of Scotland. We would like to thank the 16 individuals who agreed to be interviewed and shared their experience and perspectives of the ratings industry. We are indebted to those individuals who agreed to pilot our questionnaire and who made helpful suggestions and comments. Most significantly, we are grateful to those respondents who found time to complete our questionnaire. Finally, the Research Committee and the researchers are grateful for the financial support of the Trustees of the Scottish Accountancy Trust for Education and Research without which the research would not have been possible.

E

XECUTIVE

SUMMARY

Background In the past decade, interest in the activities of credit rating agencies (CRAs) has grown significantly. This reflects: (i) the global growth of ratings services, driven by new debt issues and structured finance products such as credit derivatives; (ii) criticisms of the CRA industry’s seeming inability to forecast the Asian crisis of 1997, and the wellpublicised corporate scandals of the early part of this century, including the recent subprime mortgage crisis creating turmoil in financial markets; and (iii) the perceived duopolistic nature of competition in the CRA industry, with domination by two CRAs: Moody’s, and Standard & Poor’s (S&P). In response to these issues, commentators have suggested greater self regulation of the industry, with the United States debating the need for formal regulation. However, only a small amount of research exists which considers what constitutes quality in the provision of ratings services. This is in contrast to studies of the audit and assurance industry, where an accumulation of literature has developed the concept of audit quality.

Purpose of study and research approach The purpose of this investigation is to apply the audit quality literature to conceptualise those desirable qualities of CRAs. These theoretical qualities were refined by undertaking 16 interviews with issuers of debt who use ratings services, debt investors, and a range of other interested parties. The interviews and audit quality and related literatures were used to develop a 71 item questionnaire relating

EXECUTIVE SUMMARY

VI

to potential CRA qualities. The 71 items relate to 14 underlying characteristics of ratings quality, labelled: co-operation; expertise; independence; investor orientation; internal processes; issuer orientation; methodology; reputation; responsiveness; timeliness; transparency; trust; values; and service portfolio. This investigation was conducted in the period 2005 to 2006. The questionnaire was sent to 2,400 interested parties representing three primary groups using CRA services: financial managers in UK corporations that are issuers of debt; debt investors (bond holders); and other interested parties (accounting and finance academics, accountants in corporate finance functions, corporate and investment bankers, financial journalists, and treasury consultants). Key findings Relative importance of ratings quality characteristics

There was a broad consensus across the three groups regarding the usefulness of the 14 ratings quality characteristics. These characteristics were generally rated in the following order: reputation, trust, and values, followed by the relatively clustered characteristics of transparency, timeliness, expertise, investor orientation, methodology, co-operation, independence, issuer orientation, internal process, and responsiveness, with service portfolio ranking last and well-behind the other characteristics. Relative importance of 71 individual questionnaire items

The items rated most highly by all stakeholder groups tended to be related to the integrity of the CRA, their ethical standards and credibility. Rated almost as highly were items which directly related

EXECUTIVE SUMMARY

to the competence of CRA outputs such as accuracy of ratings and associated reports, and the education and qualifications of staff. The lowest ranking items related to the ability of the CRA to provide specialised ancillary services. Absolute importance of 14 ratings quality characteristics

In general, the financial manager group rated the 14 characteristics as more important than the ‘other interested parties’, with debt investors located between these two groups. This result perhaps reflects that the primary relationship is between the financial manager of the issuer and the CRA, with debt investors and other interested parties somewhat detached from this process. Differences between the three groups

When making inter-group comparisons between the three groups, the most polarised differences exist between issuers and debt investors. Other interested parties views fall between the views of issuers’ financial managers and debt investors. As expected, financial managers rate issuer orientation higher than debt investors, and debt investors rate investor orientation higher than issuers. Financial managers also rate the characteristics of trust and methodology higher than debt investors. These characteristics reflect the one-to-one relationship that financial managers usually have, in their role as issuers, with the CRA. Debt investors and financial managers have differing views of what ‘timely’ means in terms of rating updates. Debt investors would like to see a rating updated as soon as possible. Issuers prefer stability in their rating, as volatility in ratings may breach financial covenants in bank borrowing agreements. When financial managers are compared with other interested parties, differences relate to four characteristics: issuer orientation;

VII

EXECUTIVE SUMMARY

VIII

responsiveness; trust; and timeliness. Similar to the financial managerinvestor comparison, issuer orientation, responsiveness and trust relate to the close relationship between issuers of rated debt and the CRA. Also, timeliness means different things to financial managers and other interested parties, with financial managers preferring stability in ratings and other interested parties preferring ratings to reflect current events. Finally the differences between debt investors and other interested parties are in items that are investor-specific or issuer-specific. That is, those items that focus on the service and attention that the particular group receives from CRAs. Financial manager issues

Sixty one per cent of responding financial managers issue debt in public markets. Two-thirds of financial managers maintain a relationship with a CRA, either for debt issuance or investment purposes. Moody’s and/or S&P are used by 95 per cent of financial managers maintaining a relationship with a CRA. However, the position is far from a duopoly, with 52 per cent of respondents indicating that they use Fitch or another CRA, either with Moody’s and/or S&P, or as their sole CRA. Multiple CRA relationships are commonplace: 48 per cent of respondents use two CRAs and 33 per cent use three CRAs. Debt investor issues

Only one-third of responding debt investors indicate that it is policy to require a rating before making a debt-related internal investment. Although a majority indicate that a rating is important, 54 per cent indicate ‘it was just one input of many’.

EXECUTIVE SUMMARY

Other interested parties issues

Two-thirds of other interested parties indicate that they make regular use of ratings produced by CRAs. This validates expectations that a wide range of stakeholders make use of ratings information. The most common uses are: research, teaching and training (18 per cent); advising clients (17 per cent); assessing the creditworthiness of clients/suppliers (17 per cent); and for investment decisions (14 per cent). Discussion and policy implications Background research conducted for this project identifies the strong growth in the ratings industry enjoyed by just a few CRAs. This growth, combined with recent corporate scandals of a global nature not foreseen by CRAs and the recent subprime market crisis, has stimulated demands for greater scrutiny and regulation of CRAs work. This project explores relevant background issues to the ratings industry, and elicits what constitutes quality in CRAs work. The questionnaire survey identifies 14 quality characteristics desirable in CRAs. A great deal of consensus exists among the three groups suggesting that each of these stakeholders are looking for similar qualities in CRAs. As the most important characteristics relate to the reputation, credibility and integrity of the CRA, it is essential that CRAs do not allow their standards to slip in the light of their expansion. Behind those characteristics relating to the standing of the CRA in the market are those aspects that reflect: technical qualities of the CRAs service provision; the timeliness of ratings upgrades/downgrades; the transparency of CRA decision-making; and their expertise and methodologies. CRAs interviewed went to great lengths to explain that their methodologies were disseminated via their websites and available to all parties. However, issues relating to communication were apparent in many interviews. CRAs could therefore benefit from making additional

IX

X

EXECUTIVE SUMMARY

efforts to explain their methodologies to stakeholder groups. These could include: presentations to issuers about the conditions needed for ratings upgrades, and what might precipitate a downgrade; presentations to investor groups about their distinctive methodologies; and ‘open doors’ sessions to other members of the financial community to create a greater understanding of their work. Surprisingly, those items relating to the characteristic of independence did not score highly in the surveys. However, it would be dangerous to interpret this as meaning that independence is not an important characteristic of a CRA. Interviewees all believed that CRAs were independent, and that it was one aspect of their work that they were entirely confident about. The most appropriate interpretation of these findings is that, independence is a sine qua non of CRA work. This is one finding that CRAs and all market participants must take great comfort from. The characteristic of CRAs ranked least important was their service portfolio. The provision of added-value ancillary services by CRAs was not seen as particularly valuable by any of the user groups. The fallout from the subprime crisis has reiterated calls to separate ratings and consulting arms. CRAs either need to better communicate the value of these services, or stick to their core business of providing ratings information. A finding of this research is that a large number of stakeholder groups make use of ratings information for a variety of purposes. However, it is apparent that many individuals do not have a significant understanding of what the ratings industry does or of the ratings process. Issues of communication have already been addressed. However, one avenue for CRAs to explore would be to educate potential members of the financial community, and develop stronger links with universities and professional bodies in the accounting and financial services sector. As well as improving an understanding of the industry, courses or programmes focusing on the analysis of debt securities and structured finance could increase the supply of junior analysts.

A

BBREVIATIONS

ABBREVIATIONS USED IN REPORT ACT AFP AFTE AMF CEBS CESR CRA CRARA DBRS EC EU FEE Fitch IGTA IOSCO Moody’s NRSRO NYSE OIP SEC S&P UK US

Association of Corporate Treasurers Association of Finance Professionals (US) Association Francaise des Tresories D’Enterprises (France) Authorité Des Marchés Financiers (French Securities Regulator) Committee of European Banking Supervisors Committee of European Securities Regulators Credit Rating Agency Credit Rating Agency Reform Act Dominion Bond Rating Services European Commission European Union Federation of European Accounting Experts Fitch Inc., a division of Fimalac International Group of Treasury Associations International Organization of Securities Commissions Moody’s Investors Services Inc. Nationally Recognized Statistical Rating Organization New York Stock Exchange Other interested party Securities and Exchange Commission Standard and Poor’s, a division of The McGraw-Hill Companies Inc. United Kingdom United States of America

1

INTRODUCTION

Credit rating agencies play a vital role in global securities and banking markets. It is essential, therefore, that they consistently provide ratings which are independent, objective, and of the highest possible quality. (European Commission, 2006 p.C59/2)

This chapter outlines the background to the study, the conceptualisation of ratings quality, and the research approach together with the structure of the report.

Background The role of credit rating agencies Credit rating agencies (CRAs), usually at the request of the issuers of a debt security, provide a rating which provides an indication of the probability of investors receiving their money back in accordance with the terms under which they invested. Ratings can be applied to a wide range of debt securities including long-term debt, medium-term notes, commercial paper, bank loans and preference shares. The CRA assesses the likelihood of borrowers (debt issuers) defaulting on their repayments. A rating reflects the credit worthiness of the issuer, together with any credit enhancement feature attached to the security, such as a guarantor, a letter of credit provider, or a bond insurer. The assessment is delivered in a letter-based rating, and provides issuers, investors, and other stakeholder groups with an objective criterion to assess the credit risks associated with a particular debt. Therefore, ratings influence the interest rate issuers will pay, and their ultimate cost of capital.

2

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

The ratings industry is dominated by two agencies: Moody’s Investors Service Inc. (Moody’s), and Standard & Poor’s (S&P), a division of The McGraw-Hill Companies Inc. Fitch Inc. (Fitch), a division of Fimalac, S.A., is a third CRA providing global coverage, but as yet does not have the same market share as Moody’s and S&P. Each agency uses their own scale and nomenclature – see Table 1.1. Panel A of this table summarises the international long-term ratings scales of the three major CRAs. CRAs also assign short-term credit ratings to debt obligations that have original maturities of one year or less, for example, commercial paper. Panel B summarises the international short-term scales of the three major CRAs. Investment grade ratings signal a low probability of default and high likelihood of repayment. At the other end of the credit spectrum, speculative (junk) ratings indicate a high probability of default or that default has already occurred. Table 1.1

Long-term and short-term credit ratings of the three largest CRAs (adapted from Voizey, 2006)

Panel A: International long-term credit ratings Investment grade Highest quality/ best quality/ extremely strong Very high quality/high quality/very strong High quality/upper medium grade/ strong Good quality/medium grade/adequate Non-investment grade Speculative/lower medium grade/speculative – less vulnerable Highly speculative/low grade/more vulnerable Poor quality/currently vulnerable High default risk/highly speculative/currently highly vulnerable High default risk/extremely poor/imminent default In default

Fitch

Moody’s

S&P

AAA AA A BBB

Aaa Aa A Baa

AAA AA A BBB

BB

Ba

BB

B CCC

B Caa

B CCC

CC

Ca

CC

C D

C C

C D

Note: Fitch and S&P append rating with ‘+’ or ‘-‘ to denote relative status within major rating categories. Moody’s append ratings with ‘1’, ‘2’, or ‘3’ to denote relative status.

3

INTRODUCTION

Table 1.1

Long-term and short-term credit ratings of the three largest CRAs (adapted from Voizey, 2006) (Continued)

Panel B: International short-term credit ratings Highest/superior/strong Good/strong/satisfactory Fair/acceptable/adequate

Fitch F1+, F1 F2 F3

Speculative/not prime/speculative

B

High default risk/ - /vulnerable Default/ - /default

C D

Moody’s P1 P2 P3 Not prime -

S&P A1 A2 A3 B C D

The structure of the CRA market A recent Basel Committee survey of CRAs indicates there are more than 130 agencies worldwide (Basel Committee, 2000). These CRAs vary considerably in terms of their size, the industries they cover, their regional focus and methodologies used. The three largest CRAs: Moody’s, S&P, and Fitch operate on an international basis and boast global coverage for their ratings. Each of these CRAs offers ratings for sovereign and corporate borrowers, and specific issues of fixed-income securities. The three largest CRAs derive the majority of their revenue in the form of issuer fees for ratings. These ratings are then made publicly available, with summary information on the rationale for their decision available free-of-charge via their websites. Additional income is earned by fees charged to subscribers who are allowed detailed access to ratings information, and by the provision of ancillary business services. Smaller CRAs operate by offering a service to specific industries (e.g. insurance) or operating in developing countries by analysing local gradations of issuer credit risk that would otherwise be overwhelmed by country risk, that all issuers in that market face (IOSCO, 2003). In contrast to the larger agencies, smaller CRAs derive most of their income from subscriber fees.

4

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

CRAs have enjoyed considerable growth in the past two decades. For example, in the 21st century revenues for Moody’s, S&P, and Fitch have grown by 17.6 per cent per annum – see chapter two and figure 2.1. This growth reflects two related processes (Sinclair, 2005). First, disintermediation, where bonds and notes sold on capital markets have replaced bank loans as a source of debt finance. Second, securitisation where mortgages, credit card receivables and bank loans are transformed into securities that can be traded in capital markets. These processes provide companies, institutions, and non-profit organisations with the ability to access debt markets, rather than borrowing from a traditional relationship bank. Therefore, issuance of debt has grown dramatically over the past two decades. The growth in the holding of bonds relative to loans by institutional investors in the UK, is shown in Figure 1.1. Figure 1.1 Trends in financial assets held by institutional investors in the UK in US Dollars (Millions) (OECD, 2004)

2500 2000 Other

1500

Shares

1000

Loans

500

Bonds

0 1980

1984

1988

1992 1996

2000

INTRODUCTION

Recent criticisms of credit rating agencies However, despite the utility of information produced by CRAs, their activities have drawn criticism from a range of interested parties including regulators, politicians, issuers, investors, and professional bodies such as corporate treasurers. A significant source of criticism of CRAs has been their inability to predict large-scale failures such as the 1997 Asian financial crisis or well-publicised corporate failures of Parmalat, Enron, or WorldCom. Each of the three leading CRAs Moody’s, Standard & Poor’s, and Fitch, provided Enron with investment grade ratings until four days prior to its collapse. Such events have led commentators to question whether CRAs are providing adequate service to market participants. Despite access to ‘insider’ information, there is a feeling that CRAs did not question Enron sufficiently: The Committee staff has concluded that the credit rating agencies’ approach to Enron fell short of what the public had a right to expect… It is difficult not to wonder whether the agencies’ practical immunity to lawsuits and non-existent regulatory oversight – is a major problem. (US Congress Committee, 2002 p.116) How reasonable such criticisms are, given certain issuers’ motivation to mislead CRAs, is open to question. Other criticisms of CRAs made by various commentators at the start of the 21st century include: •

the lack of competition within the CRA market, which is dominated by two CRAs, reflects high barriers to entry within the CRA market;

5

6

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?



the excessive profitability of CRAs, reflects the lack of competition within the CRA market and lack of leverage issuers possess to negotiate tariffs; and



the lack of transparency in the ratings process and the use of unsolicited ratings.

The next section identifies the purpose of this report and the research approach adopted.

Purpose of the study and the research approach Research issues The strong growth in the market for ratings, with questions of the ability of CRAs to predict corporate failures, together with the concerns of regulators, politicians, issuers, investors, professional associations and regulators sets the scene for a research project of interest to a wide audience. A relative paucity of research considers how market participants view the role of CRAs and the concept of ratings quality. Perspectives from other areas of related academic work are therefore examined to provide a framework for the investigation. In particular, the project focuses on the audit quality and service quality literatures which have some potential to inform contemporary thinking on CRAs. CRAs are similar in a number of ways to auditors as suppliers of audit services. Similar to auditors, CRAs operate in a unique economic market in which the CRA is a central participant. A significant source of demand for ratings comes from third parties (especially fixed income investors) who do not directly pay for the rating service. The debt issuer (like the audit client) is often a forced participant in the market, required by investors to engage the services of a CRA to obtain a rating of the credit quality of its securities.

INTRODUCTION

The focus of the research project is ratings quality and the principal aims of the project are to: •

elicit and compare interested parties views’ regarding the determinants of the quality of ratings provided by credit rating agencies;



develop a holistic conceptualisation of ratings quality, borrowing elements from audit theory; and



compare interested parties’ attitudes regarding a broad set of items relating to ratings quality, derived from the semi-structured interviews and the audit quality and service quality literatures.

A major purpose of this investigation is that it allows the attitudes of the main interest groups in credit ratings to be systematically compared, allowing the identification of major points of agreement and disagreement to be identified. Evidence of this nature is potentially valuable to regulators, such as the International Organization of Securities Commissions (IOSCO), Committee of European Securities Regulators (CESR), Committee of European Banking Supervisors (CEBS), and the Securities and Exchange Commission (SEC), whose aim it is to encourage best practice and resolve potential conflicts of interest between groups. Questions asked The investigation is informed by two sources of data. First, from the 16 semi-structured interviews undertaken with issuers, debt investors and other interested parties, to understand the determinants of ratings quality and to ascertain the relationship CRAs have with financial markets and associated market participants. Second, from a questionnaire survey developed to elicit the importance of specific characteristics of CRAs from a wide range of market participants. These characteristics were derived from a detailed review of the academic and professional literature

7

8

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

relating to audit quality and service quality published up to the start of 2006, and informed by the findings of the 16 semi-structured interviews. In total, a comprehensive list of 71 items was created. For convenience these were grouped into 14 categories. Table 1.2 lists the main categories, together with a code and the number of items. Table 1.2

Categories of questions used in the questionnaire

Category

Code

Reputation

Rep.

Values

Value

Expertise

Exp.

Methodology

Meth.

Timeliness

Time.

Issuer orientation

Issue.

Investor Orientation

Invest.

Responsiveness

Resp.

Service portfolio

SP

Independence

Indep.

Internal process

Proc.

Transparency

Trans.

Trust

Trust

Co-operation

Coop.

Total

Description The reputation and standing of the CRA within the financial community How the CRA goes about doing business The sunk costs CRAs have incurred in producing ratings, their coverage of markets The confidence market participants have in the techniques CRAs use in their assessment of issuers The ability of ratings to accurately reflect events happening within the issuer’s organisation the general economy. The ability of CRAs to offer a high level of service to the issuers The ability of CRAs to offer a high level of service to the investors Relationship issues between CRAs and individual issuers and investors The ability of CRAs to provide services other than ratings The detachment of CRAs from issuers Processes CRAs employ to ensure ratings are of the highest quality The willingness and ability of CRAs to communicate the ratings process The level of trust between CRAs and issuers and investors The degree of communication between issuers and CRAs

No. items 3 4 4 8 4 4 3 6 3 7 6 6 7 6 71

INTRODUCTION

The questionnaire consisted of five sections. For sections one to three, respondents were asked to base their responses on their general views of credit rating agencies, rather than their experiences with a single CRA. Specifically they were asked to think about the kind of rating agency that would produce ratings of the highest quality. A five-point scale, ranging from 5 (absolutely essential) to 1 (not at all essential) was used. Section one consisted of factors relating to the ideal CRA. The second section consisted of factors relating to CRA’s lead analyst and other staff. Section three consisted of other relevant factors. Section four was an open-ended section that asked for free comments regarding the quality of credit ratings and general views on CRAs, and the ratings process. The final section asked questions related to the background and experience of the people completing the questionnaire, and their experience of using ratings information. Respondent groups and response rate Financial managers were UK corporate treasurers and finance directors drawn from the Association of Corporate Treasurer’s (ACT) membership directory. Investors were UK institutional investors drawn from the Euromoney Institutional Investors Database. Other interested parties were a relatively heterogeneous group of finance academics, consultants, financial professionals connected with finance professional bodies, financial journalists, and corporate and investment bankers – all from the UK. Of the 2,400 questionnaires distributed (with two follow up mailings), 408 responses were received. This represents a response rate of 18.2 per cent - see Table 1.4. A more detailed description of the sampling procedures and sample composition is described in chapter five.

9

10

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 1.3

Respondent groups and response rates

Group

No in sample

No. excluded from sample*

No. of responses

Response rate

Corporate treasurers

950

74

198

22.6%

Investors Other interested participants

800

33

90

11.7%

650

52

120

20.1%

2,400

159

408

18.2%

Total

Note: * these represent questionnaires returned as the recipient no longer worked with the organisation.

Structure of the report Chapter two provides some background to the environment in which CRAs operate. Specifically the chapter describes: the ratings process; identifies potential sources of demand for rating information; explains how the CRA market has become dominated by just a few players; reviews the recent financial performance of CRAs; and considers international demands for greater supervision and accountability of CRAs. Chapter three contains a literature review. It begins by a consideration of the recent professional literature that documents proposals for changes in the way CRAs conduct business. The next section then describes the audit theory literature which is used to gain an insight into the interaction between the CRAs and their stakeholders. Specifically this section considers: ways in which CRAs and auditors as suppliers of audit services are similar or dissimilar; an overview of audit theory and definition of audit quality; and the economics of audit quality. The chapter considers how each of these literatures extend to the ratings industry. Finally, empirical research considering different

INTRODUCTION

stakeholder groups of audit quality is reviewed as this has the potential to inform the present study. Chapters four, five, six, and seven present the results of the present research. Chapter four presents and discusses the findings from the 16 interviews. The list of characteristics used in the questionnaire is presented along with an identification of their source. Chapter five explains how the questionnaire survey was conducted and reports the results of specific background questions asked of the three groups. Chapter six reports the findings of the questionnaire survey for all groups. Chapter seven compares findings between the groups for characteristics and individual items. Chapter eight summarises and concludes. Implications of the research for policy makers are discussed along with suggestions for future research.

11

2

CREDIT RATING AGENCIES

This chapter provides an introduction to the work of CRAs. Specifically, five issues are addressed: how a rating of a security is conducted; the demand for ratings services; the growth in revenue enjoyed by major CRAs in the 21st century, and their related financial performance; demands for greater regulation, self regulation and oversight of the activities of CRAs from both US and international viewpoints; and finally, potential barriers to entry in the CRA market.

A brief history of the ratings industry Historically, CRAs developed from mercantile credit agencies that supported business and trade creditors. The first of these agencies was The Mercantile Agency, established in 1841, followed by R.G. Dun and Company in 1859, which became Dun and Bradstreet, Inc., the owner of Moody’s until 2001 (for a full history, see Sinclair, 2005). The first individual to focus on debt issuers was Henry Poor in his History of Railroads and Canals of the United States. Poor’s Publishing Company later merged with the Standard Statistics Bureau (founded 1906) in 1941. The only other suppliers of debt ratings information were Fitch Publishing Company, established in 1924 and Duff and Phelps Credit Rating Co., which specialised in public utilities. Duff and Phelps merged with Fitch in 2000. Fitch had previously acquired some smaller CRAs that attempted to challenge Moody’s and S&P, such as IBCA (in 1992) and Thomson BankWatch (in 2000). To summarise, historically the ratings industry has been dominated by a few suppliers. Historical analyses suggest this has been due more to a common perception that the ratings industry:

14

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

…was simply too small to justify the start-up costs incident to covering the large universe of publicly held bonds. (Coffee, 2006, p.294) Furthermore, when a threat appears to the oligopoly the industry leaders enjoy, the smaller competitor is subsumed.

Ratings processes The processes used by individual CRAs vary considerably. It is important to distinguish between two types of rating. Solicited ratings occur where the issuer has invited the CRA to rate it, usually for a fee, or has at least participated in the rating. Unsolicited ratings are undertaken by the CRA without reference to the company’s management. Such unsolicited ratings are usually carried out by CRAs to improve market coverage of their ratings, or as a marketing device to encourage the company to commission a rating. At the start of the rating, a lead analyst is appointed by the CRA. The lead analyst requests information from the issuer and collects other available sources of information to undertake an assessment of the issuer’s industry and economic environment. The lead analyst will meet with the company’s management and visit its offices and production facilities. A rating is usually based on an assessment of quantitative and qualitative indicators, which are reported to a ratings committee. Individual CRAs have their own methodologies, which they disclose to a greater or lesser degree to the public via their websites. The ratings committee usually consists of the lead analyst, senior managers at the CRA and junior analysts who work on the rating. The final decision is made by a ratings committee, not by a sole individual. A report is then prepared with a draft recommendation. Issuers are allowed to comment on the report to ensure the information collated by the lead analyst is correct. If the CRA has relied on incorrect information or is able to present new evidence, the CRA may reconsider its decision.

CREDIT RATING AGENCIES

The demand for ratings services Issuers seeking to market a debt issue commission the services of a CRA to grade their securities. This is undertaken primarily for five reasons, described below: •

To define investment eligibility, as many investment organisations will have policies which specifically prohibit investment in noninvestment grade bonds. Therefore, the value of ratings has become institutionalised into organisational practice. Without an appropriate rating, investor appetite for an issue will be thin. Consequently, some commentators (Langohr, 2006) suggest CRAs provide a gate-keeping function to enter global capital markets, where the rating provides the ticket for entry.



To reduce the information asymmetry effects (Myers and Majluf, 1984) between the issuer and investors. In undertaking an independent review of the probability of default of a debt issue, the CRA is able to provide objective and independent information to the market about the issuer’s future prospects. Furthermore, the CRA has access to the company’s management and is able to access information about the organisation’s future strategy and prospects not available to the market. Therefore, the information gap between issuers and investors is reduced. The effect of reducing this information asymmetry is to reduce the risk premium associated with a debt issue, lowering the margin payable on the issue and reducing the organisation’s overall cost of capital.



CRAs are able to provide an advisory role to management, offering feedback about the long-term effects of their approach to risk. CRAs can advise management of the effect of changes in strategy to their perceived risk profile, and the effect such a change would have on their credit rating, and consequently cost of capital.

15

16

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?



Ratings are used by regulators, for example, the Committee of European Banking Regulators in the European Union (EU) and the Securities and Exchange Commission in the United States (US) to assess capital adequacy and permitted investments for regulatory purposes. As well as providing information upon which investors can base investment decisions, ratings are used to evaluate trading partners, financial counterparties, for example, in financial instruments such as swaps, and business partners, for example, where an organisation enters into a joint venture agreement or strategic alliance.



For internal treasury purposes, such as assessing counterparty risk. Typically a corporate will have an internal policy which will set absolute and relative limits to the amount that can be invested with any one counterparty. This limit may vary according to the nature of counterparty, with higher limits applying to counterparties with higher credit ratings.

Financial performance of credit rating agencies Having considered the market structure of the ratings business, it is useful to understand what effect this may have on the financial performance of the CRAs themselves. The leading CRAs have demonstrated significant growth in revenue and on the basis of publicly available information appear to be highly profitable. The revenues of S&P, Moody’s and Fitch grew at an average of 17.6 per cent per annum from 2001 to 2005, reaching $4,826 million – see Figure 2.1.

17

CREDIT RATING AGENCIES

Figure 2.1 Revenue of three major CRAs, 2001-2005

Revenue $millions

6000 5000 4000

Fitch

3000

Moody's

2000

S&P

1000 0 2001

2002

2003

2004

2005

Moody’s represents the only CRA quoted on a stock exchange (New York Stock Exchange, NYSE). S&P and Fitch are owned by McGrawHill and Fimalac respectively, and therefore the performance of these ratings businesses is harder to observe. It is therefore only possible to examine Moody’s in any detail. Table 2.1 identifies some of the key financial performance statistics of Moody’s relative to those companies also quoted on NYSE. This demonstrates that Moody’s performance is superior to the average of its peers quoted on NYSE. Table 2.1

Financial performance of Moody’s versus NYSE, source Thomson Disclosure

P.E. ratio (2005) Net income margin % (2005)

Moody’s 35.30 29.56

NYSE mean 7.39 11.54

It is difficult to assess how comparable Moody’s financial performance is to Fitch and S&P. Moody’s exhibit the strongest average growth in revenue over the five-year period (21.5 per cent p.a.) compared to S&P (16.4 per cent p.a., source: McGraw-Hill Inc. 2006 annual report)

18

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

and Fitch’s (13.8 per cent p.a., source: Fimalac 2006 annual report). In conclusion, the demand for ratings has generated significant sales growth for the three leading agencies, and in particular the structured products market, for example, credit derivatives. In the first quarter of 2007, structured products accounted for 43 per cent of Moody’s Investor Services total revenue, an increase of 44 per cent on the first quarter of 2006 (source: Moody’s 2007 first quarter results). However, the market concentration enjoyed by the three largest CRAs operating on a global scale, together with their strong financial performance led to criticisms that the ratings market is an oligopoly and anti-competitive. It is a matter of speculation how sustainable such growth is in the longterm. The production of ratings is a labour-intensive business requiring the employment of new analysts. Therefore, the ratio of experienced ratings staff to new hires will decrease. It is plausible that, without the facilities in place to rapidly develop new analysts, the quality of ratings information will fall, which in turn will reduce an important barrier to entry, namely reputation, to new participants.

Demands for self regulation of credit rating agencies Concerns about the work and market structure of CRAs have led to a variety of responses from regulators, governments and professional bodies on an international basis. This section is structured as follows: first, the concept of the SEC’s Nationally Recognized Statistical Rating Organisations (NRSRO) is discussed. Second, recent debates in the US are summarised. Third, international perspectives are discussed including the International Organization of Securities Commissions (IOSCO) Code of Conduct Fundamentals for credit rating agencies, and an exposure draft developed by corporate treasurers in the UK, US and France. The third subsection, describes the European response to the work of CRAs including that of regulators, the European Commission and the European Parliament.

CREDIT RATING AGENCIES

Introduction and recognition of the work of CRAs A degree of regulation has existed in the ratings industry since 1975 with the establishment of Nationally Recognised Statistical Rating Agencies (NRSRO) by the SEC. At the time, the SEC had no intention of becoming the de facto regulator of CRAs (Chambers, 2006). NRSRO status was originally created by the SEC to identify those agencies that it could rely on to distinguish among grades of creditworthiness in various regulations under the federal securities laws, that is, for its own internal purposes. Specifically the SEC was motivated to formulate rules on net capital requirements for broker-dealers. To allow dealers investing in highly rated securities to hold less capital the SEC needed to designate those CRAs that were competent to judge the quality of securities in which the brokers invested. Since then the role of the NRSROs has expanded; the SEC has accelerated issuance approval for issuers that are regular users of capital and have a high enough rating with an NRSRO. At present there are five NRSROs: Moody’s, S&P, Fitch, Dominion Bond Rating Service Limited (DBRS), and A.M. Best. Although the SEC created the NRSRO label specifically for use in its own jurisdiction, NRSROs’ ratings are now used as benchmarks in federal and state legislation, financial and other regulators, foreign regulatory schemes, and private financial contracts. As the financial community began to place greater reliance on CRAs’ ratings, the NRSRO concept found greater currency. Likewise, although NRSRO designation is confined to the US, the global nature of ratings has resulted in the designation taking on the role of an accreditation standard worldwide. The SEC undertook a significant investigation into the role of CRAs, as described in the following section. As part of this exercise, the SEC conducted a consultation into the future of NRSRO designation. Arguments against NRSRO status are:

19

20

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?



that the NRSRO designation acts as a barrier to entry into the credit rating business (see discussion of Credit Rating Agency Reform Act in the following section);



the SEC’s lack of explicit regulatory authority over NRSROs; and



the SEC’s view that it would be inappropriate to impose a more comprehensive regulatory framework on CRAs (SEC, 2005).

Perspectives from the US In January 2003, the SEC produced a preliminary report concerning the relationship between CRAs and the US capital markets (SEC, 2003a). This preliminary report was in response to the provision of SarbannesOxley Act 2002 and the US Congress’s concerns about the role that CRAs may have played in the corporate failures at Enron and other corporations. Specifically, the preliminary report examined: •

the role of CRAs and their relation to the securities markets;



impediments faced by CRAs in performing that role;



measures to improve information flow to the market from CRAs;



barriers to entry into the rating business; and



conflicts of interest faced by CRAs (SEC, 2003a).

To follow up the preliminary report the SEC issued a concept release considering the role and function of CRAs in June 2003 (SEC, 2003b). Specifically, the SEC (2003b) investigated issues concerning:

CREDIT RATING AGENCIES



The utility of the NRSRO designation: including alternatives to NRSRO, the appropriateness of the existing recognition criteria, the recognition process, and the evaluation of ratings quality.



Examination and oversight of CRAs. Is more direct, ongoing oversight of CRAs required? Are there minimum standards or best practices to which CRAs should adhere? Should CRAs incorporate general standards of diligence in their analysis and the training and qualifications of CRA analysts? Should CRAs be required to (rather than recommended to) register as investment advisers?



Conflicts of interest. Should CRAs implement procedures to manage potential conflicts of interest when CRAs develop ancillary fee-based business; do firewalls exist to separate this business from the ratings business? Are issuers unduly pressured to purchase ancillary advisory services by a rating analyst at a NRSRO?



Alleged anti-competitive, abusive and unfair practices. For example, smaller CRAs have alleged that S&P and Moody’s have attempted to squeeze them out of certain structured finance markets by a practice of ‘notching’, that is lowering ratings, or refusing to rate certain asset pools unless a substantial proportion of the assets are also rated by them. Similar concerns exist with unsolicited ratings where a CRA attempts to induce an issuer to pay for a rating it has not requested, for example, by sending a bill for an unsolicited rating.



Information flow: should the ratings process be more transparent? How could the quality of information to users of credit ratings be improved? Should all ratings information be available to all market participants, regardless of subscription?

21

22

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

In 2005, as part of the ongoing review of CRAs addressed by the concept release (SEC, 2003b) the SEC issued a proposed new rule on the definition of a NRSRO. The definition has three components which the SEC believes: ...to be the most important criteria in determining whether an entity’s ratings should be relied upon for the purposes of the securities laws and Commission rules and regulation. (SEC, 2005, p.20) These criteria are summarised in Table 2.2.

CREDIT RATING AGENCIES

Table 2.2

Proposed criteria for NRSRO recognition, SEC (2005)

Component & scope

Proposed criteria

A. Publicly available, current assessments of specific securities 1.

Ratings should be made publicly available at no cost to all market participants

2.

NRSRO concept should be limited to those who rate specific securities (rather than just legal entities)

3.

Ratings reflect current opinions (actively monitored and updated)

B. Credibility to securities markets 4.

The CRA is generally accepted in financial markets

5.

CRAs which provide only limited coverage (rate only a limited sector of a debt market, or specific geographic area) could acquire NRSRO status

C. Internal processes and compliance 6.

CRA should have procedures in place to ensure analysts are competent

7.

CRAs should have in place controls to assess the integrity of the information sources they rely on

8.

CRAs organisational structures should ensure that ratings are analysed, reviewed and approved at appropriate levels

9.

CRAs take steps to identify sources of conflicts of interest, and minimise or avoid them

10.

CRAs have policies and procedures which effectively protect nonpublic information provided by issuers

11.

CRAs have sufficient financial resources to ensure they can comply with their own procedures

Note:

A, B, C = component

1-11 = scope

23

24

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Finally, the US government has debated the need to regulate CRAs. The Credit Rating Agency Reform Act (CRARA) was enacted on 29 September 2006 with the intention of improving the quality of ratings. CRARA aims to: ...improve ratings quality for the protection of investors and the public interest by fostering accountability, transparency and competition in the credit rating industry. (US Senate Committee, 2006) Specifically, the CRARA establishes a new registration process to create a clear route for a CRA wishing to gain NRSRO status (SEC, 2007). The SEC also provides oversight so that the NRSROs continue to issue credible and reliable ratings, and afford protection against conflicts of interests and the abuse of non-public information. However, the legislation has received a mixed response from market participants. The Association of Finance Professionals (AFP), an association for treasury operations professionals in the US, suggests that the CRARA will improve investor confidence in CRAs and capital markets in general (Kaitz, 2006). However, some investors suggest opening the market up to greater competition increases the number of recognised CRAs and increases the likelihood of opinion shopping (Laughlin, 2006). International perspectives

In contrast to the enthusiasm shown for greater regulation of CRAs in the US, international responses to the regulation of CRAs have favoured greater self regulation. The demand for self regulation comes from two sources. First, the IOSCO and second, from professional associations representing financial professionals who as part of their job role may issue debt securities, which might be rated, and invest in securities which also may be rated.

CREDIT RATING AGENCIES

IOSCO initially conducted an investigation into the role CRAs play in securities markets. The report (IOSCO, 2003) examined six issues: •

Whether CRAs disclose information about their decisions and rating criteria and if disclosure is adequate for making investment decisions.



How, when and to whom CRAs disseminate ratings decisions.



How CRAs protect against the misuse and unauthorised disclosure of non-public information.



The conflicts of interest CRAs face.



The types of regulatory oversight to which CRAs are subject.



The barriers to entry that exist as a result of the regulatory framework, and whether such barriers impact on the quality of CRA ratings.

As a result of the 2003 investigation, IOSCO established a code of conduct for CRAs (IOSCO, 2004). The key points of the IOSCO Code are summarised in Table 2.3.

25

26

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 2.3

Summary of key features of the IOSCO Code of Conduct

Key features Quality and integrity of the ratings process a. Quality of the ratings process, analyses in accordance with written policy, the rigour of methodologies used, the availability of sufficient resources to undertake the rating b. Monitoring and updating: regular review, and review on provision of information which could affect the rating c. Integrity of the ratings process, compliance with applicable laws, the employment of staff with highest standards of integrity CRA independence and conflicts of interest a. General b. CRA procedures and policies c. CRA analyst and employee independence CRA responsibilities to the general public a. Transparency and timeliness of ratings disclosure b. Treatment of confidential information Disclosure of the code of conduct and communication with market participants

At the same time, European regulators were giving consideration to whether CRAs should be regulated. The Committee of European Securities Regulators (CESR) was asked by the European Commission (EC) about the potential options for regulating CRAs. CESR undertook a consultation in 2004 (CESR, 2004) and delivered its response to the EC in 2005 (CESR, 2005). CESR decided not to regulate CRAs, but proposed that ‘a pragmatic approach be adopted’ (CESR, 2006a p.1) to keep under review how CRAs would implement the standards set out in the IOSCO code of conduct. CESR’s strategy has the voluntary participation of the five largest CRAs (Moody’s S&P, Fitch, DBRS and A.M.Best - see CESR, 2006a). This is evidenced by first, each participating CRA providing a letter to CESR detailing how it has

CREDIT RATING AGENCIES

complied with the IOSCO Code of Conduct, indicating any deviations. Second, an annual meeting between CESR and the CRAs is held to discuss implementation of the Code. Finally, the CRA has to provide an explanation to the national CESR member if a ‘substantial incident’ (CESR, 2006a p.1) occurs with a particular issuer in its market. CESR (2006b, p.4) reported the findings of CRAs compliance with the IOSCO code using survey responses from the four CRAs (DBRS, Fitch, Moody’s, and S&P), and concluded that the ‘CRA codes comply to a large extent with the IOSCO code’. However, it identified a number of minor deviations, and some of greater importance, leading to the opinion that ‘there is room for improvement in these areas’. Specifically, CESR (2006b) identified two concerns common to all CRAs surveyed. The first area relates to ancillary services and the requirement for the legal and operational separation of rating business with other CRA business which may create a conflict of interest. CRAs consider their rating assessment services as an integral part of the rating business, rather than an ancillary service. For example, in structured finance ratings where the debt structure reflects negotiations between the issuer and CRA. CESR (2006b), however, considers that conflicts of interest can arise from such situations, ‘especially when the service is actively marketed and proposed by the CRAs to issuers’ (p.76) rather than from specific issuer demand. The second area relates to the provision of unsolicited ratings. CRAs possess differing degrees of transparency in reporting unsolicited ratings, partly because they use different interpretations of what constitutes an unsolicited rating. CESR notes that it is difficult for investors to identify whether a particular rating was solicited or not. In particular, CRAs could include a simple flag in their databases to identify whether the issuer participated or not in the rating. The French securities regulator, Autorité des Marchés Financiers (AMF) has also taken an interest in the functioning of CRAs and has published annual reports on CRAs in 2004 (AMF, 2005) and 2005

27

28

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

(AMF, 2006). AMF’s interest in CRAs is motivated by a requirement under France’s Financial Security Act of 2003 to publish such a report, reflecting the impact ratings have on the prices of securities concerned, and a desire to establish links with counterparts in other countries to coordinate efforts to supervise and assess CRAs. AMF’s first (2004) annual report on CRAs (AMF, 2005) is a largely descriptive document, detailing: general CRA activity; the French market for financial ratings, the role and added-value of CRAs for different market participants; CRAs rules of conduct; and transparency of methods. AMF’s second annual report on CRAs activities is more analytical. It considers how four CRAs (Moody’s, S&P, Fitch, and A.M. Best) have responded to the IOSCO code of conduct. AMF restricts its assessment to an analysis of the CRAs’ codes of conduct published on their individual websites. AMF (2005) conclude that the CRAs adhere closely to the principles of the IOSCO code. However AMF (2006, p.10) suggest that: Further clarification is needed in terms of the legal commitment made by agencies to enforce their own codes and in terms of disclosures of deviations from the IOSCO code. Finally, the UK’s Association of Corporate Treasurers and its equivalents in France, the Association Française desTresories D’Enterprises (AFTE) and the US, the Association of Finance Professionals (AFP), published a the Code of Standard Practices for Participants in the Credit Ratings Process. The Code extends the idea of best practice not just to CRAs but to other market participants, such as issuers. In particular the Code outlines the need for minimal regulation of CRAs, placing reliance on industry-agreed best practices. The Code was motivated by a decision of the International Group of Treasury Associations (IGTA) in 2003 to work towards an industry agreed code. The Code is outlined in Table 2.4.

CREDIT RATING AGENCIES

Table 2.4

Code of conduct for market participants (ACT/AFP/AFTE, 2005 and O’Donovan, 2004)

Issuer responsibilities a. An obligation to disclose its business strategy, legal and management structure, and management processes b. Discussion of risks and opportunities of the business environment, assisting the CRA to understand its approach to risk management, and financial policies c. Provision of other financial information that allows the CRA to better understand its circumstances d. To inform the CRA about changes in the financial situation of the business, including new financings. Issuers should commit to hold a full review with the CRA on (at least) an annual basis e. To respond to communications from the CRA on a timely basis

CRA responsibilities a. To enhance the transparency of the ratings process, publishing and adhering to its methodologies b. Confidential information gathered by the CRA must be protected and not publicly disseminated c. To establish and document policies and procedures against potential conflicts of interest d. To clearly distinguish between solicited and unsolicited ratings and identify when the rating was last updated e. To improve communication with issuers and the market

Barriers to entry It is widely recognised by commentators, regulators, investors and issuers that the market for corporate and sovereign ratings is dominated by the two leading CRAs, S&P and Moody’s, and a third smaller one, Fitch. Although the CRA market has always been small, it is difficult to believe that this simply reflects high barriers to entry creating a lack of competition. As Coffee (2006, p.294) suggests:

29

30

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

The paucity of competitors may have been more the product of the common perception that the market was simply too small to justify the start-up costs incident to the covering the large-universe of bonds. That is, substantial growth and high profitability in the ratings industry has occurred only in the past 15 years. As described earlier within this chapter, regulators and various financial commentators see this situation as problematic. As the issue of perceived lack of competition within the CRA market is relevant to the present research it is useful to consider how such market concentration has been created. Six potential factors have been identified: •

NRSRO designation. The SEC has been slow to recognise new agencies, providing issuers with little choice but to engage Moody’s, S&P or Fitch. How strong this argument is, is difficult to gauge, as market concentration existed prior to the introduction of NRSRO in 1975. Furthermore, NRSRO classification was not undertaken with the intention of formally regulating CRAs, as described earlier, but its status became institutionalised. The CRARA (2006) theoretically reduces barriers to gaining NRSRO recognition by the SEC (SEC, 2007). The effect on the rating industry and credit markets remain to be seen.



Reputation of the CRA. Long established CRAs have a track record of rating securities which can be measured against default statistics. Therefore, the performance of ratings information can be assessed against actual performance, for example, that provided by default statistics. Consequently CRAs benefit from a learning curve on how to undertake ratings successfully, which asserts their first-mover advantages (Langohr, 2006).

CREDIT RATING AGENCIES



Networking. Langohr (2006) suggests issuers are motivated to have their debt securities rated because other market participants do, so they want investors to note the rating of their bonds to stimulate appetite for issues. In turn, bond investors look for adequate coverage of investment opportunities. As others seek a rating, it becomes necessary for an issuer to seek a rating from a credible CRA to be seen to fit in with other issuers. Therefore, CRAs benefit from having network externalities.



Sunk costs. Ratings reflect the sunk labour costs of producing them (Langohr, 2006). Once ratings have been created, the production costs are unrecoverable. Therefore, for new CRAs to provide adequate coverage of a particular debt it would require them to make a substantial investment in sunk costs, for instance, rating securities on an unsolicited basis.



Resources. A new entrant may have fewer resources, for example, staff and analytic tools, than established CRAs (IOSCO, 2003). Without these resources, a new entrant may be at a disadvantage compared to an established CRA who may be able to employ additional, and more experienced, staff to analyse a large issuer involved in complex transactions.



Conflicts of interest. Conflicts of interest threaten a CRA’s independence, arising when a single issuer’s fees comprise a large proportion of a CRAs overall revenue. To the new entrant a single fee-paying issuer may represent a significant proportion of their total income. This creates a potential conflict of interest if the new entrant perceives that an adverse rating will lose it future business from the issuer. IOSCO (2003) suggest that the large amount of time necessary to establish a new entrant may necessitate an affiliation with a larger firm. Such an affiliation could create a conflict of interest if the financial interests of the parent firm influence the rating decisions of the CRA affiliate.

31

32

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Therefore the nature of the credit rating business creates high levels of market concentration, dominated by a few suppliers of ratings information.

Summary This chapter has provided an overview of how CRAs provide ratings services, and the market for ratings information. Specifically, five sources of demand are identified: investment eligibility; reduction of information asymmetry between issuers and investors; ancillary advisory services provided to management; the use of ratings information for regulatory purposes; and internal treasury investment purposes. The CRA market is dominated by two or three CRAs: Moody’s, S&P and the smaller Fitch. From the information available, CRAs are relatively profitable enterprises, reporting strong growth in revenue over the past five years. The perceived lack of competition in the CRA market, coupled with their strong financial performance, has fuelled demands for greater regulation of their activities. The US seems to favour formal regulation, whilst the remaining international financial community has favoured self regulation. The IOSCO Code has been in force now for three years. Investigations undertaken by CESR and France’s AMF, indicate that, by-and-large, the Code is being adhered to by the CRAs. If competition within the CRA market is to be increased, barriers to entry need to be reduced. Theory suggests six potential barriers: NRSRO designation by the SEC; existing CRAs high reputation; networking externalities; sunk costs associated with the production of ratings; new entrants having inadequate resources; and conflicts of interest created by a new entrant being too reliant on a few issuers.

3

LITERATURE REVIEW

An objective of this research is to develop a model of ratings quality, provided by CRAs to issuers, investors and other interested parties. To develop a conceptualisation of ratings quality, quality is viewed through the lens of two literatures: •

Contemporary discussions considering the regulation and self regulation of CRAs and their interaction with credit markets; and



Audit theory, and definitions of audit quality and developing ideas of what constitutes quality in the provision of audit and assurance services by professional accounting firms.

Market participants’ attitudes to credit rating agencies Academic research considering CRAs tends to focus on the performance of their ratings. Studies include those of the ratings process itself (see Pottier and Sommer’s (1998) review), the response of share and bond prices to changes in debt ratings (for example, Barron et al. 1998, Nayar and Rozeff, 1994; and Pinches and Singleton, 1978) and the ability of explanatory variables such as profitability, leverage, liquidity, and growth to determine published ratings (for example, Adams et al., 2003). However, relatively little academic research considers how debt market participants perceive the work of CRAs, with a literature review locating only three studies, all conducted in the US as shown by Table 3.1. Ellis (1998) and Baker and Mansi (2002a, 2002b) assess some differences in investor and issuer views of CRAs. The general findings are that issuers prefer multiple ratings, as they attract better debt pricing,

34

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

while investors require only one, as they usually also undertake their own analysis. Table 3.1

Academic literature considering issuer and/or investor views of CRAs

Authors Ellis (1998)

Sample Survey of CFOs of investor-owned utilities and institutional fixedincome investors in US

Results Issuers obtain ratings from three or more CRAs, but investors require just one. Investors, unlike issuers, like to see ratings updated immediately to reflect all relevant information.

Baker and Mansi (2002a)

Survey of industrial issuers and institutional investors in the US

Majority of investors require just one rating, while issuers require two or more ratings. Issuers generally use more than two CRAs to: resolve a split rating; or achieve better coverage. Only 11% of issuers are the subject of unsolicited ratings, with 75% believing that such ratings should be clearly identified as such. 97% of issuers believed the current model of issuers paying for ratings does not create a conflict of interest for CRAs. Issuers and investors perceive S&P and Moody’s as more accurate than DCR and Fitch.

Baker and Mansi (2002b)

Survey of industrial issuers in the US of investment and noninvestment grade debt

Investment grade issuers commission more ratings than non-investment issuers. Investment-grade issuers see solicited ratings as indispensable compared with non-investment-grade issuers. Respondents are generally satisfied with the performance of CRAs, but more satisfied with S&P than Moody’s.

LITERATURE REVIEW

Three recent surveys (AFP, 2002, 2004; TBMA, 2006) have been undertaken by professional bodies which are described in the following section. However, it is important to note that these are not academic studies and it is difficult to assess the generalisablility of the views presented. In response to the Enron collapse, the Association for Financial Professionals (AFP) (2002) surveyed 2,700 treasury and finance professionals to assess attitudes concerning the accuracy and timeliness of the NRSRO ratings and the demand for regulation of CRAs. The majority believed their company’s rating to be accurate but not timely enough. Most respondents also believed that upgrades took longer to occur than downgrades. Ninety per cent of treasury and finance professionals believed that the SEC should take additional action to improve its oversight of CRAs and foster greater competition in the ratings market. AFP (2004) undertook a follow-up survey in 2004 sampling 2,000 AFP practitioner members. This time a slightly higher percentage of respondents believed their ratings to be inaccurate. However, marginally more respondents considered changes in their ratings to have been made on a timely basis. However, demand for greater regulation had fallen with 59 per cent (compared to 90 per cent in 2002) believing that the SEC should take a greater role in overseeing the CRAs. This difference in perception shows the general move towards a code of conduct and away from regulation. A third survey was undertaken by The Bond Market Association (TBMA) and the ACT in 2005/6 to ascertain issuers’ and investors’ attitudes to CRAs. Issuers (N=41) were generally corporates or financial institutions, with nearly half based in the UK and the rest in other EU countries or the US. Investors (N=51) were again based in the UK (41 per cent) or in another EU country or the US. Two-thirds were asset managers. Issuers perceived greater access to capital and lower costs of funding as the main advantages of a rating. Investors tended to see

35

36

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

ratings as useful indicators of internal risk management and useful for portfolio management. Eighty five per cent of issuers believed that two ratings was the optimum, compared to investors’ belief that three ratings was the ideal: a finding that contradicts Baker and Mansi’s (2002a) finding that the majority of investors require just one rating. Issuers tended to believe that the lack of competition in the ratings market was problematic, in contrast to investors who did not perceive this as a problem. Issuers tended to believe their relationship with the CRA was good. Perhaps surprisingly, a majority of both issuers and investors had not heard of the IOSCO code of conduct. In summary, the lack of research considering the behaviour of CRAs, their processes and the concept of ratings quality means there is only a limited theoretical base. Consequently, insights from other literatures are required. The next section describes the audit theory and audit quality literature, from the field of accounting, which in part draws on the service quality literature from the fields of marketing and operations management. However, before describing the audit quality literature it is useful to consider the (dis)similarities of CRAs and auditors offering audit and assurance services. In particular CRAs and auditors have a number of similarities about their mission, the structure of the market in which they operate and the types of services they offer. Table 3.2 compares the ways in which CRAs and auditors are similar (Panel A) and dissimilar (Panel B).

37

LITERATURE REVIEW

Table 3.2

Comparison of CRAs and auditors

Panel A: Features that are similar CRAs offering ratings services Market for ratings services dominated by three CRAs.

Auditors offering audit and assurance services Market for audit and assurance services dominated by four firms (Big Four).

Concept of client and ‘real’ client

Issuer pays fee for ratings, to provide information to investors and other stakeholders about probability of default on its debt security. Primary value is to investors and other stakeholders.

Client (i.e. auditee) pays the auditor to provide information about accuracies of auditee’s financial statements. Primary value of audit is to investors and stakeholder groups.

Reduction in agency costs

Rating by a credible CRA allows issuer to issue debt at a lower cost by reducing information asymmetries between issuer and investors.

An audit undertaken by a credible auditor reduces contractual conflict between owners, managers and other stakeholders and reduces agency costs.

Additional services

CRA can provide additional services to issuer in form of advice on risk strategy to maximise bond rating.

Auditors offer additional ‘business lines’, such as taxation; consulting; corporate finance; and corporate recovery.

Long-term relationship between CRA/Auditor and client

Relationship with CRA is ‘for life’ of security with CRA continuing to rate organisation on an unsolicited basis, even if issuer no longer pays for its services.

Audit tenure in UK, typically between 30 and 40 years (Ridyard and DeBolle 1992).

Key staff

CRA appoints a lead analyst to the rating who is the key point of contact between the CRA and issuer.

Auditor appoints an audit engagement partner to the audit, who is the key contact between the auditor and auditee.

Feature Market structure

38

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 3.2

Comparison of CRAs and auditors (Continued)

Panel B: Features which are dissimilar CRAs offering ratings services

Auditors offering audit and assurance services

Regulation

Little regulation of CRAs exists.

Extensive regulatory framework exists for auditors offering audit and assurance services.

Professional bodies

No profession exists for CRA analysts.

Well-developed profession for auditors exists dating back to 19th century.

No. of CRAs/Auditors client has a relationship with

Organisations will often seek more than one CRA to rate its securities.

Organisation has one principal auditor, although is likely to commission services of other auditors for other non-audit services.

Competition

CRA market relatively uncompetitive, with issuers forced to recruit at least one CRA. Issuers are price takers, with little ability to influence fee structure.

Competition in audit and assurance market (Martinov and Roebuck, 1998), characterised by pressure on fees and slow growth (Behn et al., 1997).

Feature

Audit theory and audit quality Considering the quality of ratings information, the audit literature appears to offer a definition of ratings quality. DeAngelo’s (1981) classic definition of audit quality focuses on two elements. First, the competence of auditors (i.e. their ability to identify breaches in clients’ accounting systems); and second, their independence (i.e. willingness to report the breaches). Although competence and independence are distinct constructs, the two dimensions are not completely separate. For example,

LITERATURE REVIEW

an auditor might decide not to try to identify problems (competence) which they are not motivated to disclose (independence). Technical competence is relatively easy to conceptualise. Independence is more problematic, with a distinction being made between: •

Independence in fact, which reflects DeAngelo’s conception of independence (an auditor’s willingness to report on defects in audited financial statements), which in itself is not directly observable. The Federation of European Accounting Experts (FEE) (1996, p.24) define independence in fact as ‘the state of mind which has regard to all considerations relevant to the task in hand but no other’.



Independence in appearance, which is defined by signals or other directly observable indications. When independence is mentioned in laws or professional rules, it is ‘independence in appearance’ that is usually the focus.

Therefore, ratings quality could be similarly described as reliant on two matters: the competence of the CRA (their ability to correctly rate a debt security); and their independence (willingness to downgrade an issuer’s debt security). The economics of audit quality and extension to ratings quality Much of the early thinking on audit theory applies economic principles to the study of audit quality. In the economics of audit quality, auditing is conceptualised as an economic exchange between a supplier (the auditor) and direct users (an audit client who purchases the services) and indirect users (investors and other stakeholders for whom the audited accounts are destined). In an economic exchange, sufficient resources need to be employed to prevent the parties from maximising their own

39

40

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

interests at the expense of the other parties to the exchange. This is achieved by a market mechanism. If the auditors have not complied with the contract of the exchange (e.g. by behaving opportunistically) they will attract criticism from other market participants and suffer damage to their commercial reputation. Auditors with a good commercial reputation can charge higher fees than those who contract with a poorer reputation. The additional remuneration available to the auditor with the higher reputation is termed a ‘quasi-rent’. Specifically, the quasirent is defined as: The difference between the remuneration for any productive resource in its current use and the maximum remuneration which would be received for its alternative use. (Arruňada, 1999, p.21) Audit quality is protected by both explicit and implicit contracting safeguards. In an explicit contract, if the auditor has undertaken a defective audit, this may result in litigation against the auditor. If the work is found defective by third parties, then the auditor may be required to pay compensation to those injured parties. In an implicit contract, the auditor may be punished by existing or potential clients withdrawing their trust in the auditor. This loss of reputation will result in existing or potential clients either withdrawing their business, or demanding more onerous terms to the engagement. The collapse of the major auditor Andersen, could be attributed to the breach of an implicit contract safeguard. For the reputation (quasi-rent) incentive mechanism to operate, three conditions have to exist (Arruňada, 1999): •

The sale price (audit fee) must exceed the marginal or opportunity cost, so the quasi-rent is created.



The expected value of the quasi-rents must be sufficient to discourage non-compliance i.e. the additional remuneration available to the

LITERATURE REVIEW

higher quality auditor is large enough to motivate the auditor to continue to produce high quality audits. •

Audit clients and investors/other stakeholders must be aware of the auditor’s incentives (i.e. the creation of quasi rents) so that they trust them.

Extending these ideas to ratings quality is straightforward. The CRA, with a good commercial reputation, is able to generate higher levels of fees than those with poorer reputations through the establishment of quasi rents. Ratings quality is protected by implicit contracting safeguards. An implicit safeguard exists, because if the CRA is perceived to have undertaken low quality work issuers may withdraw their business. The explicit contracting safeguard is only theoretically conceivable if there is a change in the regulatory environment for CRAs. For an explicit safeguard to exist the organisation would have to demonstrate that the CRA had undertaken defective work to have legal recourse. However, this is unlikely as CRAs argue that their rating is simply an opinion. In the US, arguably the most litigious jurisdiction, CRAs enjoy constitutional, federal and state law protections designed to mitigate exposure to third-party subpoenas and civil litigation. CRAs understandably are keen to protect their relative immunity from such explicit contracting safeguards. Furthermore they argue that: …such (legal) exposure for opinion providers is not only contrary to the protections provided by the First Amendment of the US Constitution, but it would also have the ultimate chilling effect on our ability to publish independent and potentially controversial opinions. (Moody’s, 2003) Marketing investments and the expense created by the development of commercial brands play a role in safeguarding commercial relationships

41

42

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

(Arruňada, 1999). The creation of a brand i.e. a CRA’s good name, provides a guarantee. The brand, as a marketing investment, provides an incentive to provide a service which meets the expectations of stakeholder groups. The marketing investment represents an intangible asset whose value rapidly erodes if a firm delivers defective work. An unusual feature of CRAs is their periodic publication of unsolicited ratings, where the issuer plays no part in the rating. If the unsolicited rating is of a lower quality than solicited ratings, where the CRA has access to the organisation’s management and information not in the public domain, then the unsolicited rating will damage the reputation of the firm and reduce or eliminate quasi-rents. Empirical work considering perceptions of audit quality A literature review identifies eight studies which consider different market participants views of the factors which are antecedents of audit quality. These studies have some potential to inform the present study of the type of factors which are indicative of ratings quality. Each study uses a questionnaire survey mailed to the sample identified. The results are summarised in Table 3.3.

LITERATURE REVIEW

Table 3.3

Results of behavioural audit quality research

Authors

Sample

Results

Mock and Samet (1982)

US auditors.

Identified five audit quality dimensions: planning; administration; procedures; evaluation; and conduct.

Schroeder et al. (1986)

Audit committee chairs and auditors in US.

Effect of 15 factors on audit quality. Audit team factors more important than audit firm factors.

Carcello et al. (1992)

Financial statement preparers, auditors and users in US.

Twelve quality factors identified. Most important factors identified were: experience with the client; industry expertise; responsiveness to client needs; and adherence to general standards of GAAS.

Beattie and Fearnley (1995)

Finance directors of UK companies listed on London stock exchange.

Identified five key characteristics focusing on the importance of the Audit Engagement Partner’s technical skills and people skills.

Behn et al. (1997)

Controllers in US.

Used Carcello et al.s’ (1992) 12-item survey. Identified: responsiveness to client needs, executive involvement, effective and ongoing interaction with audit committee, conduct of fieldwork, industry expertise, and prior experience of client as all positively associated with client satisfaction.

WarmingRasmussen and Jensen (2001)

Shareholders, financial journalists, auditors and managing directors.

External users tend to perceive audit quality attributes as attributes that inspire confidence in the auditor; six main quality dimensions identified (moral and ethical aspects); four groups rate quality dimensions differently.

43

44

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 3.3

Results of behavioural audit quality research (Continued)

Authors

Sample

Results

Duff (2004)

Audit engagement partners (AEPs), finance directors (FD) and fund managers (FM).

Nine factor model of audit quality developed and validated, with distinction made between technical qualities and service qualities. All three groups rated the technical quality dimensions of reputation and capability the highest. AEPs and FDs responses were broadly similar. FDs rate service quality attributes higher than AEPs FMs rated technical quality issues as being more important than AEPs or FDs.

The first study conducted of this type was undertaken by Mock and Samet (1982) who created a list of 110 factors derived from Statements on Auditing Standards, Statements on Quality Control Standards, peer review manuals, and firm quality control standards. The list of factors was reviewed by a small group of auditors, to reduce the 110 items to a questionnaire with 32 variables. Mock and Samet’s (1982) research was extended by Schroeder et al. (1986) who considered perceptions of various audit quality factors of Fortune 500 audit committee chairpersons and the (then) Big Eight accounting firm partners. The Schroeder et al. (1986) investigation required respondents to rate the importance of 15 factors to overall audit quality. Audit committee chairpersons rated audit team factors as more important than audit firm factors. For example, the attention of partners and managers to the audit scored more highly than the overall reputation of the firm. Carcello et al. (1992) sampled financial statement preparers, financial statement users and audit partners in the US, and identified

LITERATURE REVIEW

12 audit quality factors. The five most important quality factors were: team experience with client; industry expertise; an audit team that operated to high ethical standards; a partner knowledgeable about the client’s industry; and frequent communication between auditors and management. Again, overall team characteristics were rated as more important than firm characteristics. Beattie and Fearnley (1995) surveyed finance directors of 210 listed UK companies, and identified five main factors that they labelled: integrity of the firm; the technical competence of the firm; the quality of the working relationship with audit partner; the reputation of the firm; and the technical competence of the audit partner. As a development of Carcello et al.’s (1992) investigation, Behn et al. (1997) considered the relationship between audit quality attributes and client satisfaction. Controllers of Fortune 1000 companies in the US were asked to rate their current auditor on the 12 audit quality attributes identified by Carcello et al. (1992). Six (of the 12) audit quality variables had a positive relationship with client satisfaction with the audit team: responsiveness to client needs; audit firm executives actively involved in the audit; effective interaction with the audit committee; appropriate conduct of audit field work; industry expertise; and team and firm experience with client. Only one of the audit quality variables was negatively associated with client satisfaction that ‘the audit team members maintained a skeptical attitude throughout the audit engagement’ (Behn et al., 1997, p.23). However, this finding is perhaps indicative of the balancing act an auditor has to perform between satisfying the client and fulfilling professional and stakeholder expectations. Suggestions provided by respondents to improve client satisfaction included the auditor being more proactively involved in the client’s business, providing service beyond statutory compliance and making value-added suggestions. Behn et al. (1997) also report that client satisfaction is higher during the first few years after a change in auditors and some evidence that

45

46

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

satisfaction is higher when the controller has previous work experience with the auditor. Warming-Rasmussen and Jensen (2001) studied the perceptions of audit quality in Denmark, and sampled two groups of external users, shareholders and financial journalists, and managing directors (preparers) and public accountants (auditors) to create an audit quality scale. Their research was unusual in the sense that it sampled external users’ perceptions of audit quality, rather than the simple auditor/auditee samples used by previous researchers. The investigation identified six audit quality dimensions, focusing largely on moral and ethical issues, using EFA. These they labelled: personal credibility; the auditor’s independence of company management; openness in the report about matters of interest to creditors and shareholders; knowledge of the industry; loyalty about minority shareholders; auditors sceptical attitude to the auditee. The two groups of external users tended to rate all six of the audit quality dimensions higher than the auditors and managing directors. Duff (2004) in an ICAS monograph developed a model of audit quality synthesising the audit quality literature with the service quality literature from marketing. The model is shown in Table 3.4. The research instrument, AUDITQUAL, was created by borrowing items from the prior literature considering audit quality along with other items from the service quality literature. AUDITQUAL was administered to three groups: partners in the 16 largest audit firms in the UK (representing auditors); finance directors of UK listed companies (auditees); and fund managers in the UK (representing investors). All three groups rated the technical quality dimensions of reputation and capability the highest. The dimension considered the least important by all three groups was the ability to provide non-audit services. Auditors and auditees responses were broadly similar, with auditees tending to rate service quality attributes higher than auditors. Auditors generally rated technical quality dimensions higher than auditees. As expected,

LITERATURE REVIEW

investors rated technical quality issues as being more important than auditors or auditees. Investors also rated service quality issues as being less important than the other two groups, probably reflecting their minimal involvement in the audit process, and hence inability to directly observe service quality issues. Table 3.4

Constructs used in AUDITQUAL

Dimension

Description

Technical qualities Reputation

The reputation the auditor (firm, partner, staff) enjoys in the market place

Capability

The technical abilities the firm has including ethical standards of partners/staff

Expertise

The expertise the auditor has in specific markets

Independence

The degree of detachment the auditor has from the client

Non-audit services

The ability of the auditor to provide non-audit services

Experience

The experience the auditor has of undertaking the audit in the client’s firm

Service qualities Assurance

The ability of the auditor to be proactive in monitoring service levels provided to clients

Empathy

Issues relating to the ability of the firm and partner to empathise with the client and client’s finance director

Responsiveness

The ability of the auditor to deliver services

Summary This chapter has reviewed two literatures with direct relevance to this research. The first considered market participants attitudes to CRAs. However, this academic literature is confined to just three

47

48

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

studies conducted in the US. Three practitioner-led investigations voice criticisms of CRAs post-Enron from US company treasurers (AFP, 2002, 2004) and international issuers and investors views of the value of ratings information (TBMA, 2006). Overall, these investigations provide relatively little understanding of what market participants expect from CRAs. Second, to gain insights into the nature of the relationship CRAs have with issuers, investors and other interested parties the audit quality literature was considered. Audit quality is the subject of relatively vigorous academic research, with a focus of quality relating to the independence and competence of the auditor (DeAngelo, 1981). Other researchers have incorporated into this definition other technical qualities relating to the experience, expertise and integrity of the audit team and engagement partner. Service aspects such as the relationship between the auditor and auditee, the responsiveness of the auditor, and processes employed by the auditor to ensure client service are also factors relating to a more holistic definition of audit quality. The economics of audit quality and its extension to ratings quality is discussed alongside these studies. This work emphasises the advantage the high quality auditor/CRA has over inferior auditors/CRAs by virtue of ‘quasi-rents’ which allow the high quality service provider to charge a premium. Thus, quasi-rents act as a quality safeguard, as any loss of reputation damages or eradicates these quasi-rents. Audit/ratings quality is further protected by explicit (threat of litigation) and implicit (loss of trust) safeguards.

4

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

The purpose of this chapter is to describe how the ratings quality survey was developed. The chapter aims to link the literature described in chapter three with the findings of the interview research to describe the 14 key characteristics of ratings quality used in the survey of market participants. Where possible, the characteristics are illustrated with quotes from the interviews undertaken with key debt market participants.

Interviews Sixteen interviews with nine financial managers who issue credit-rated debt, two investors, and five other interested parties were undertaken to elicit a broad understanding of how different stakeholders in the market for ratings services view the role of CRAs. The other interested parties included two treasury consultants, a commercial banker specialising in project finance, and two senior officers in CRAs. All interviewees were experienced and senior individuals with experience of working with CRAs and credit markets. Specifically, the interviews were undertaken to allow refinement of the constructs developed from the audit quality and service quality literatures described in chapter two, within the contextual framework described in chapter three. Table 4.1 provides background details of each interviewee.

50

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 4.1

Interviewee and their job role Market capitalisation (£billion)

Interviewee

Sector

Description

1. Financial Manager

Food

Treasurer

1-10

2. Financial Manager

Packaging

Treasurer

1-10

3. Financial Manager

Food

Treasurer

1-10

4. Financial Manager

Transport

Treasurer

1-10

5. Financial Manager

Transport

Treasurer

>10

6. Financial Manager

Drinks/Tobacco

Treasurer

>10

7. Financial Manager

Drinks/Tobacco

Treasurer

> 10

8. Financial Manager

Transport

Treasurer

1-10

9. Financial Manager

Drinks/Tobacco

Treasurer

> 10

10. Investor

Private money market brokerage

Fund manager

NA

11. Investor

Private investment company

Fund manager

NA

12. Other interested party

Self-employed

Consultant

NA

13. Other interested party

Self-employed

Consultant

NA

14. Other interested party

European commercial bank

Banker

> 10

15. Other interested party

UK

Head of CRA

> 10

16. Other interested party

US

CRA senior ratings analyst

> 10

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

Questionnaire development The questionnaire development was informed by the prior literature considering the intersection of professional service firms with clients and investors, refined by evidence from the 16 interviews undertaken. Specifically questionnaire items were created to describe 14 characteristics of CRAs and their interaction with stakeholder groups. These 14 characteristics are grouped broadly into two categories: technical features describing those characteristics of technical qualities of ratings analysis; and service features addressing characteristics of the relationship CRAs have with issuers and investors. An overview of the questionnaire is provided in Table 4.2. Specifically, Table 4.2 describes the characteristics of ratings quality, and illustrates each characteristic with a typical questionnaire item.

51

52

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 4.2

Characteristics of the quality of services provided by CRAs

Characteristic

Description

Example questionnaire item

1. Expertise

The ability of the CRA to make competent and informed decisions about the probability of default.

The CRA undertakes research into the client’s own industry.

2. Independence

The ability of the CRA to make objective decisions about the issuer.

The CRA analysts’ promotion prospects are not dependent on their ability to generate ancillary business.

3. Internal process

Effective internal processes within the CRA.

The CRA employs wellqualified and educated staff.

4. Methodology

Those processes the CRA uses to assess the probability of default.

The CRA adjusts its ratings on a timely basis.

5. Reputation

How credible the CRA is to third parties.

The CRA operates to the highest standards of integrity.

6. Transparency

The clarity of CRA decision making.

The CRA publishes all assumptions relevant to its decisions.

7. Timeliness

The willingness of the CRA to upgrade/ downgrade a rated security.

Ratings are only adjusted after due consideration when long-term factors change materially.

8. Values

The organisational values and norms of the CRA.

The CRA analyst is actively involved throughout the ratings process.

Technical features

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

Table 4.2

Characteristics of the quality of services provided by CRAs (Continued)

Characteristic

Description

Example questionnaire item

9. Co-operation

Effective communication between the CRA and issuers and investors.

The CRA and issuer make a point to keep each other informed about what is going on.

10. Empathy

The effective relationship between the CRA and issuers and investors.

The CRA analyst has investors’ best interests at heart.

11. Investor orientation

The ability to provide high levels of service to investors.

The CRA responds to investors’ questions.

12. Issuer orientation

The ability to provide high levels of service to issuers.

The CRA regularly conducts client service review meetings.

13. Service portfolio

The ability of the CRA to provide specialised, ancillary services.

The CRA is able to provide structured finance ratings.

14. Trust

The degree of trust that exists between the CRA and issuers and investors.

The CRA publishes no unsolicited ratings.

Service features

Interviewees were asked a range of questions about the desirable characteristics of CRAs, and their expectations of those supplying ratings information. This information is incorporated into the following section which describes the development of the 14 characteristics of ratings quality. These are segmented into technical characteristics, describing features of CRAs’ operations which determine the outcome of a rating; and service characteristics, which describe how a CRA goes about doing its business.

53

54

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Technical features The first four technical features of the quality of ratings information relate to the reputation, independence, expertise and values of the CRA. These relate to the DeAngelo (1981) definition of audit quality, based on competence (reputation and expertise) and independence (objectivity and values). Reputation This characteristic relates to the competence element of the DeAngelo (1981) definition of audit quality. As competence is difficult to prove, the reputation of a CRA is a proxy for this feature. The reputation of a CRA also relates to the economics of quality, where a high quality supplier has the ability to charge a premium for its services (quasi-rents) over those providers with a lesser reputation. This definition seems particularly appropriate in the CRA industry given the duopolistic nature of competition. All interviewees, without prompting, identified that the nature of competition in the ratings industry meant that only two or three CRAs were suitable choices: There are really only three agencies. So, I suppose, it is logical that we used the one that I was happiest with. (Financial Manager) The NRSRO designation afforded to just five CRAs by the US’s SEC also created a means of the major CRAs differentiating themselves from smaller competitors: We are essentially an organisation that has come into being as a result of market needs and yet it is different in different locations in the world. So, our corporate origins are in the US where regulation in the form of SEC and NRSRO status gave us a sort of reason for

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

being, but when it is Europe, it is not actually in any way through legislation or through regulatory requirements. So, we operate on a kind of a global basis for the markets. (OIP, CRA) Independence A critical feature of the provision of ratings information concerns its independence. If market participants consider the objectivity of the CRA to be compromised, the value of that information diminishes rapidly. Encouragingly all interviewees believed the CRAs to be independent of issuers or other market participants. For example: I’ve seen clients kicking and screaming, but they [the CRA] stuck to their opinion. I think they’re independent. (OIP, consultant) Expertise Another element of the DeAngelo (1981) antecedent of quality was identified in interviews. This related to the expertise of the CRA. Generally, all interviewees believed the leading CRAs to be competent; however, most interpreted this question as relating to the CRAs’ people, rather than the CRA itself. For example: Fairly competent, have some good analysts and mediocre analysts. Need to pull up the mediocre side to an acceptable level for everyone. (Financial Manager) It depends on the people involved. You tend to find technically that they are fine. Sometimes they lack a bit of experience because the analysts are quite young. (Financial Manager)

55

56

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

We have three good analysts. Competent people. (Financial Manager) Values The values characteristic relates to those values the CRA shares with the issuer, and market participants in general. The characteristic therefore has an ethical dimension. None of the interviewees alluded to any ethical issues concerning CRAs, reflecting their satisfaction with their independence. Should be more proactive in an ongoing dialogue. I see banks several times a year, CRAs only once a year. (Financial Manager) The remaining four technical characteristics relate to the process of publishing a rating. Internal process Internal process relates to human resource management processes employed by the CRA, to ensure that ratings information is of the highest quality. This characteristic is related to expertise, reflecting that the competence of a CRA relies not just on the reputation and procedures of the CRA, but also the competence of its people, as noted by OIP, consultant: The system itself in the sense of how they do their statistical analysis, how they take information into account and how they take account of those other factors. In the sectors I have seen, which is fast moving consumer goods, and engineering and construction, it seems pretty good. If you get over the problem with an individual

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

analyst, you find that the rest of the body can understand the same. (OIP, Consultant) An analyst will never issue an opinion by himself. He has to go through certain internal protocols before he can. I am not aware of any rating agency that would allow an analyst just to issue an opinion. I don’t think that that exists. (Financial Manager) Methodology Methodology is how the CRA arrives at a rating for an issuance. It relates to the specific rating, and also associated credit reports made available to subscribers. CRAs’ methodologies have also been the subject of relatively vigorous scrutiny from academic researchers examining the efficacy of rating decisions, and the ability of various financial reporting items to predict a rating. Questionnaire items relating to this dimension reflect the importance issues connected with methodologies play in the ratings process. Examples include cyclicality, consistency, and inter-CRA comparability. This is illustrated by the following quote: Obviously, there are going to be nuances in particular sectors, but it should be the same whether an airline is a British airline or a Chinese airline or European. There should be a consistent methodology across. (Investor) Transparency This characteristic reflects the clarity of CRAs’ decision-making and relates to the original rating decision, as well as rating upgrades or downgrades in changing credit circumstances. Transparency was a key characteristic identified by the interviews:

57

58

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Well, they keep on saying they’re trying to make it more transparent and I’d wish I could actually see through it. The numerical side is made reasonably transparent. The opinion sides, they state what their views are, often one wonders where they come from. They do have their own hobby horses and you can see an agency moving like this towards one issue, then another moving towards another one. (Financial Manager) CRAs are however sceptical that more needs to be done in this direction: Given that we publish all of the process and it can extend into lots and lots of documentation. If there is a problem it may be that issuers and investors aren’t prepared to read that and understand it. So, often criticism of our activities may come on the back of people who haven’t taken the trouble to do their own work about what is involved. (OIP, CRA) However, as CRAs use information that is proprietary in nature, not all information components can be made available: The first pole of the relationship is that it is professional. And the other one is, that it is transparent within the limits of the discretion that we are able to handle any of the information that is handed to us by one party or another. (OIP, CRA) Timeliness The final technical characteristic of the CRAs activities relates to the timeliness of ratings upgrades and downgrades. Naturally, issuers will hope to see their rated bonds upgraded when prospects improve, and delay downgrading when prospects deteriorate:

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

Everybody wants to have a higher rating than necessarily they will receive from us. Investors are more sceptical, they don’t want to be told a story that they think is inappropriate, and yet at the same time they would like us to do all of the homework appropriate. So, I think, just the very nature of the business of getting a balanced and independent opinion means that there will be sometimes a gap in expectations as you defined it. (OIP, CRA) One financial manager described the situation where his company had a third, unsolicited rating. When the company was subject to takeover speculation, the third, unsolicited rating over-reacted: And I found that [CRA1] and [CRA2] have been very sensible and put us on credit watch as you would expect because the deal is not being consummated, they are sending a clear signal to the market that ratings may change. This is ok, but with [CRA3] we have an unsolicited rating. They downgraded us to junk bond status. Why? You know, we had a good chat with them to understand their rationale. They shouldn’t be jumping the gun. (Financial Manager)

Service features Service features relate to six characteristics that reflect the relationship a CRA has with its issuers (co-operation, issuer orientation, and trust), and investors (investor orientation), how reactive they are to enquiries from requests from various market participants (responsiveness) and their ability to offer more than bond ratings information (service portfolio). Each of these features are drawn from the audit quality literature described in chapter three.

59

60

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Co-­operation This characteristic encompasses the willingness of an issuer to ‘tell the CRA everything’ and keep the CRA informed about current developments: There’s an article in the treasurer’s handbook which says ‘What should you tell your rating agency’ and it says ‘everything’. You don’t refuse to answer any question. So the rating agency has potentially, and I say potentially because they don’t ask the kinds of questions that the outside world thinks they ask. They ask much more sensible questions, but potentially, they have a lot of inside information. They also have what’s in the United Kingdom called relevant information not generally available. That’s to say share price sensitive information which does not fall in the narrow definitions of the EU directive. (OIP, consultant) We have an open relationship with the rating agencies. We believe, that we should give them access to information so that they are able to come up with an informed credit view of our company. (Financial Manager) It is, however, a two-way relationship and CRAs are expected to keep issuers, and also other market participants, informed about relevant developments so there are no surprises: They need to communicate what they want from management, which information they want. They should keep us up to date with their changes in methodology. (Financial Manager)

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

Issuer orientation This characteristic describes how the CRA provides high levels of client service to issuers. It includes matters such as the CRAs willingness to disclose fee levels, provide discounts for frequent issuers and the ability to meet with a relationship manager to identify issues of concern: What am I expecting? A regular dialogue, updating press releases at least on a yearly basis to the market. This is what they should be doing. Clear, concise rationale behind why we are being rated as we are. (Financial Manager) Trust The trust characteristic includes a number of items relating to the confidence market participants have in the CRA to provide high quality ratings information. Included are items relating to the continuity of an analyst assigned to a rating, and feelings of mutuality between the issuer and the CRA, and the investor and the CRA. The characteristic also includes questions relating to the use of unsolicited ratings which are also uniformly disliked by issuers: I don’t think they’re [unsolicited ratings] helpful because management has had typically no input in them at all. And typically, they are just based on the information of equity analysts or newspapers or whatever. I don’t think they provide a chief amount of benefit to the market. (Financial Manager) Unsolicited ratings were also described as: A very bad marketing tool, trust with the issuer can break down. (Financial Manager)

61

62

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Investor orientation Investors often subscribe to ratings information services provided by CRAs, like issuers, therefore, they too represent direct clients requiring a level of service. Items within this dimension include the range of information available to investors, including websites, courses, seminars and the availability of opportunity or discussion. An investor explained: For example I may have [name of CRA] providing the trigger rating or a ratio that is particularly important for them and it has to be a particular rating band. Sometimes it’s easier now to just say – now, I’m not exactly sure how you worked that one out and they will actually send you their calculation and that’s becoming more prevalent. (Investor) Responsiveness A key characteristic common to all service companies reflects the responsiveness of the service provider to client requests. This characteristic uses items from the audit quality literature described in chapter three, to describe the empathy that exists between issuers and CRA analysts, the contactability of the analyst, and the willingness of a CRA to resolve problems. As one issuer said: What am I expecting? A regular dialogue, updating press releases at least on a yearly basis to the market. This is what they should be doing. (Financial Manager)

INTERVIEWS WITH KEY MARKET PARTICIPANTS AND THE DEVELOPMENT OF THE RATINGS QUALITY SURVEY

Service portfolio In addition to providing ratings information, the CRAs have also diversified into other more specialist information services. These include structured finance ratings associated with relatively new financial products such as credit derivatives. These services are usually termed ancillary services by the CRAs. The IOSCO Code identifies that these services should be separated from ratings information services for independence reasons. In a sense, these services are similar to the consulting services that auditors used to sell to their audit clients, prior to corporate governance changes in the early part of this century. CRAs are keen to promote the utility of such services, presumably because they are relatively lucrative. No issuers interviewed made any reference to these services. This finding suggests that these services are more widely used by the financial rather than the corporate sector. However, the questionnaire includes items relating to the ability of the CRA to provide advisory services, structured finance ratings, and ancillary business services.

Summary This chapter has described the findings from 16 interviews with key market participants in the ratings industry. The interviews were used to construct a questionnaire to assess views on aspects of the work of CRAs which lead to high quality ratings information. Specifically, 14 characteristics of CRAs and their interaction with stakeholders are identified. These are divided into technical features (expertise, independence, internal process, methodology, reputation, transparency, timeliness, and values) and service features (co-operation, empathy, investor orientation, issuer orientation, service portfolio, and trust).

63

5

CONDUCT OF SURVEY

The purpose of this chapter is to describe the administration of the survey and report how the three stakeholder groups of financial managers, other interested parties and investors have responded to those sections of the questionnaire designed to elicit their interaction with CRAs and the state of the ratings industry. Each of the three groups is examined in turn.

Financial managers The questionnaire was mailed to 950 treasurers and finance directors. They were chosen from the Association of Corporate Treasurer’s Directory 2006. Seventy four parties were eliminated from the sample because either the company did not exist any longer or the individual had left the company. One hundred and ninety eight useable responses were returned, resulting in a response rate of 22.6 per cent. Of the non-respondents, 61 replied but indicated that they did not want to participate in the survey. Thirty three of these stated that they could not participate because they had no experience with CRAs, 13 said it was company policy not to take part in any surveys, six indicated that they did not have time and one person stated that the survey had already been completed by another member of his company. Although the response rate may appear relatively low, when the questionnaire length, the somewhat unorthodox nature of the subject within the accounting field and the fact that marketing-related studies usually achieve a response rate of between 10 and 30 per cent (Green et al., 1988), the response rate is acceptable for a postal survey of this nature.

66

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 5.1 identifies that 61 per cent of respondents issue debt in public markets. Of the 121 who issue debt in public markets, the most common public debt market to issue is in the UK (85 per cent), with a majority also issuing in the EU (64 per cent), and the US (53 per cent). A much smaller number issue outside the UK/EU/US (16 per cent) – see Table 5.2. Table 5.1

Respondents issuing debt in public markets

Respondents Issue debt in public markets Do not issue debt in public markets No response Total Table 5.2

Frequency (N=)

()

121

61

75

38

2

1

198

100

Public markets in which respondents issue debt

Debt issuers in public markets

Frequency (N =)

(%)

In UK

103

85

In EU

78

64

In US

64

53

Other

19

16

Table 5.3 reports respondents’ use of CRAs. The majority (N=112) of the sample engage CRAs and issue debt. A further 18 use CRAs but do not issue debt. Presumably, these individuals subscribe to CRA ratings information services for investment purposes. Only 7 (4 per cent) respondents indicated they issued debt, but do not engage the services of a CRA. One of these 7 respondents indicated that it did not have a rating because its parent company was rated and guaranteed its issues.

67

CONDUCT OF SURVEY

Table 5.3

Respondents’ use of CRAs

Use of CRAs Engage CRA(s) and issue debt Use CRA services but do not issue debt Do not use CRA, but issue debt Do not use CRA or issue debt No response Total

Frequency

Frequency

(N =)

(%)

112

57

18

9

7

4

57

29

4

2

198

100

The engagement of CRAs by respondents is shown in Table 5.4. The most frequently used CRA was Moody’s in 112 cases (86 per cent of CRA users), followed by S&P with 105 respondents (81 per cent of CRA users). Fifty nine (45 per cent CRA users) also used Fitch as a CRA, and 14 (11 per cent CRA users) also used another CRA, which in most cases was either A.M. Best or Dominion Bond Ratings Services. The number of CRAs used by individual respondents can be summarised as follows: • • • •

16 per cent of respondents use just one CRA 48 per cent of respondents use two CRAs 43 per cent of respondents use three CRAs 3 per cent of respondents use four CRAs

Evidently, it is common for financial managers who use the services of a CRA to maintain multiple relationships, as illustrated in Table 5.4.

68

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 5.4

CRAs engaged by respondents CRA combination

Moody’s √ √ √

S&P √ √

Fitch

Frequency Other

√ √

√ √ √ √

√ Total

√ √ √

√ √ √ √

√ √

√ √

(N=) 48 35 10 8 7 6 4 4 4 2 2 130

(%) 37 27 8 6 5 4 3 3 3 2 2 100

Table 5.5 describes the fee levels paid by those respondents who use the services of a CRA. The majority, 43 respondents, pay more than £120,000 in annual fees, 70 respondents pay their fees on a per issue basis, 52 are part of a frequent issuer programme. Obviously, to some extent the fee levels paid by issuers reflects the number of CRAs employed to rate an issue, so the figures in Table 5.5 bear a relationship to Table 5.4.

69

CONDUCT OF SURVEY

Table 5.5

Annual fees payable by issuers

Level of fees per annum

Frequency (N=)

(%)

Less than £15,000

5

4

£15,000 - £30,000

14

13

£30,000 - £60,000

20

18

£60,000 -£120,000

30

27

More than £120,000

43

38

112

100

Total

Other interested parties The questionnaire was mailed to 650 other interested parties. The other interested parties were made up of 100 academics randomly drawn from the British Accounting Review Research Register 2004 (Helliar et al., 2004), 88 members of the accounting sector, 112 consultants, and 25 members of associations and professional bodies with the latter three groups all chosen from the ACT Directory (2006). Seventy six financial journalists were also chosen from the Freelance Directory, and finally 249 employees of banks in various functions, were again drawn randomly from the ACT Directory. Overall, 120 useable responses were returned, resulting in a response rate of 20.1 per cent. Fifty two parties were eliminated from the sample because either the company did not exist any longer or the individual had left the company. Sixty one respondents indicated that they did not want to participate in the survey. Fifty six of these 61 stated that they could not participate because they had no experience with CRAs, two indicated that they did not have time and one person stated that the survey had already been completed by another member of his company.

70

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Seventy seven respondents make regular use of ratings, 20 indicated that they do not use ratings in any way, and 23 did not respond to this question. Table 5.6 gives an overview of what use other interested parties make of ratings information. Table 5.6

Use of ratings information by other interested parties

Use made of ratings information Research, training, teaching Advise clients Assess creditworthiness of client/suppliers Investment decisions Analysis Internal risk management Background information Compare companies Benchmarking Regulatory purposes Total

Frequency (N=)

(%)

14 13 13 11 8 6 4 4 2 2 77

18 17 17 14 10 8 5 5 3 3 100

Investors The questionnaire was mailed to 800 UK institutional investors chosen at random from the Euromoney Institutional Investors’ Database. Ninety useable responses were returned, resulting in a response rate of 11.7 per cent. Thirty three parties were eliminated from the sample because either the company did not exist any longer or the individual had left the company. Ninety two individuals indicated that they did not wish to participate in the survey. Of the 92 non-respondents, 59 indicated they could not participate because they had no experience with CRAs, five said it was company policy not to take part in any surveys, 12

71

CONDUCT OF SURVEY

indicated that they did not have time, and 12 indicated that the survey had already been completed by another member of their company. The size of responding investors, in terms of assets under management, is shown in Table 5.7. Over half the sample managed funds of under £50 billon. Table 5.7

Size of responding investors

Assets under management

Frequency (N=)

(%)

Less than £10 billion

27

33

£10 – 50 billion

20

24

£50 – 100 billion

10

12

£100 – 200 billion

12

15

£200 – 300 billion

10

12

3

4

Total

77

100

Undisclosed

13

More than £300 billion

The use of ratings information by responding investors is shown in Table 5.8. The majority of respondents (60 per cent) indicated that the company did not have an internal policy requiring a rating when investing in debt securities. Although the majority of responding investors did not have an internal policy that required ratings, 16 respondents indicated that they usually expected to see more than one rating for an issue. Only 12 said that one rating was enough. Thirty one individuals found it useful to see if two CRAs come to the same rating and 39 respondents indicated that they like to see more than one rating because each CRA uses a different methodology.

72

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 5.8

Use of ratings information by investors

Policy regarding use of ratings

Frequency (N=)

(%)

- Requires at least one rating

12

13

- Requires at least two ratings

11

12

5

6

Total

28

31

No internal policy of requiring ratings for internal purposes

54

60

8

9

90

100

Internal policy requires a rating for internal investment purposes

- Requires three or more ratings

Unknown Total

When asked how important the rating was for their investment decisions, ratings were generally seen as valuable, but over half the sample indicated that the external rating was only one input of many. Only nine respondents stated that the rating had no influence on their investment decisions (Table 5.9). Table 5.9

Importance of rating for decision making purposes

Importance of ratings The rating is very important The rating provides useful information about the company’s financial situation Ratings are mainly used as a control function Ratings are used as a means of pre-selection The external rating is just one input of many The rating has no influence on investment decisions Total

Frequency (N=) (%) 31 34 39 24 22 49 9 174

43 27 24 54 10 192

CONDUCT OF SURVEY

Summary Over 60 per cent of financial managers responding to the survey issue debt in public markets, with the most common markets being the UK (85 per cent of issuers), the EU (78 per cent of issuers), and the US (53 per cent of issuers). Of those issuing in public markets, the vast majority (93 per cent) engage at least one CRA to rate the debt. Relationships with multiple CRAs are commonplace for issuers, with 84 per cent of issuers engaging two or more CRAs. As expected, the CRAs used most frequently are Moody’s (86 per cent) and S&P (81 per cent). However, 63 per cent of participants indicated that they engage Fitch or another CRA alongside, or to substitute for, Moody’s and/or S&P. Therefore, although it may appear Moody’s and S&P enjoy a duopoly market position, it seems the services of smaller CRAs are also employed by UK issuers. Responding other interested parties use ratings information for a diverse range of purposes. The most common include: research, training and teaching; advising clients; assessing the creditworthiness of suppliers; and for investment decisions. Surprisingly, the majority of investors did not have in place an internal policy requiring a rating when making investment decisions. However, nearly half indicated that they expected to see at least one rating when making an investment decision. These results suggest that ratings are now so common in public debt markets that investors do not see the need to stipulate the availability of ratings information.

73

6

RESULTS OF THE QUESTIONNAIRE SURVEY: ALL GROUPS

This chapter presents an overview of the questionnaire findings. The first section considers the rankings of the 14 characteristics of ratings quality for the whole sample. The second section reports the rankings of the set of 71 questionnaire items for the whole sample. Comparisons between the three groups are presented in the next chapter.

Characteristics Table 6.1 presents the rankings for the whole sample for the 14 characteristics of ratings quality. Characteristics were ranked on a scale of 1 to 5, with 5 being very important and 1 being not at all important. Table 6.1 also shows the mean scores and the dispersion. The higher the mean score the more important the item is judged to be. The dispersion (disp.) column indicates very low (VL), low (L), moderate (M), high (H) or very high (VH) standard deviation. Very low standard deviation, for example, would mean that there was a high level of consensus within the sample. Very high standard deviation, on the other hand, would indicate a high degree of disagreement between the respondents. The highest ranking characteristic was reputation, closely followed by issues relating to trust. It is perhaps not surprising that reputation was regarded as the most important characteristic as CRAs are reputational intermediaries. Ratings would have little value unless the organisation issuing it was a respected party with high credibility in the market and the involved parties were able to trust it. One financial manager said: Fundamentally, the whole market is based on trust. If there isn’t trust between what the rating agencies are doing with respect to

76

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

the investor base and the issuers then the whole thing will break down. (Financial Manager) The values of the CRA firm were also regarded as essential in a high quality CRA. That is, market participants expected the CRA and its employees to operate to the highest standards of integrity, and be fully committed to ensuring its ratings accurately reflected the credit quality of the security being rated. Table 6.1

Ranking of characteristics

Characteristic

Code

Mean

Disp

Reputation

Rep.

4.65

VL

Trust

Trust

4.56

L

Values

Value

4.39

L

Transparency

Trans.

4.10

L

Timeliness

Time.

4.07

H

Expertise

Exp.

3.90

VL

Investor orientation

Invest.

3.82

VL

Methodology

Meth.

3.78

L

Co-operation

Coop.

3.72

VL

Independence

Ind.

3.65

VL

Issuer orientation

Issue.

3.58

L

Internal process

Process

3.56

VL

Responsiveness

Resp

3.49

VL

Service portfolio

SP

2.92

VL

Transparency was rated highly, and records a low dispersion indicating significant consensus among respondents. All parties wish to understand the CRA’s ratings methodology, why the CRAs may upgrade

RESULTS OF THE QUESTIONNAIRE SURVEY: ALL GROUPS

or downgrade an issue, and wish to be informed about all relevant ratings’ developments. Transparency was of particular concern to investors and other interested parties, with a general belief that the CRAs could improve their communication. Financial managers seem to be more informed and knowledgeable than the other parties, which might be expected as they are required to work closely with the CRAs. Considering the least important characteristics, service portfolio ranks the least important. It is interesting to note that none of the parties placed much importance on any other services the CRAs market, other than the provision of ratings information. A surprising finding is the lack of weight given to characteristics of independence. All interviewed parties stated that while independence certainly was crucial for CRAs, they never had any experience of the CRAs’ independence being compromised. As one financial manager explained: I think that worrying about their independence is probably the least of our worries. There is no evidence to show that they are particularly influenced by any of their stakeholders. There is none of the kind of issues that they are unduly influenced by anything in the way they run things in the media, for example. (Financial Manager) The dispersion for all characteristics, except timeliness, was low or very low which shows that there is a high level of agreement within the three samples. It is no surprise that the dispersion for the characteristic of timeliness is high, because timeliness clearly means different things for financial managers and investors/other parties. Investors/other parties would like to see ratings updated immediately whereas financial managers would like ratings updated only when relevant characteristics change materially. Moreover, financial managers would like to have the opportunity to comment on and react to any ratings actions before they are published.

77

78

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Individual items Table 6.2 presents rankings for the 71 questionnaire items. Note that, in some cases, the description of the items has had to be abbreviated due to space constraints – the full description appearing in the questionnaire can be found in the Appendix. The ratings quality characteristic the individual item belongs to is shown in the Characteristic column (Char). Table 6.2 also shows the mean scores and the dispersion (Disp.). The higher the mean score (on a scale of 1-5) the more important the item is judged to be. The dispersion column indicates very low (VL), low (L), moderate (M), high (H) or very high (VH) standard deviation. Very low standard deviation, for example, would mean that there was a high degree of agreement within the sample. Very high standard deviation, on the other hand, would indicate a high degree of disagreement within the sample.

79

RESULTS OF THE QUESTIONNAIRE SURVEY: ALL GROUPS

Table 6.2

Ranking of questionnaire items

Questionnaire item

Char.

Mean

Disp.

CRA operates to highest standards of integrity

Rep.

4.89

VL

CRA reports are factually correct

Meth.

4.83

VL

CRA is highly competent

Exp.

4.79

VL

CRA operates to high ethical standards

Value

4.76

VL

CRA is credible to third parties

Rep.

4.68

VL

CRA is open and honest with issuers

Trust

4.67

VL

Ratings are accurate

Meth.

4.65

L

CRA deploys staff with adequate experience

Proc.

4.65

VL

CRA’s methodology is robust

Meth.

4.65

VL

CRA employs well-qualified and educated staff

Proc.

4.63

VL

CRA has no financial interest in an issuer

Ind.

4.61

L

CRA is open and honest with investors

Trust

4.57

L

CRA’s methodology is consistent across sector

Meth.

4.54

L

Analyst is knowledgeable about issuer’s industry

Exp.

4.50

L

CRA allows issuers a right of appeal

Trans.

4.48

L

Analyst is actively involved in ratings process

Value

4.43

L

CRA is conscientious

Rep.

4.39

L

CRA adjusts its ratings on a timely basis

Time.

4.38

L

CRA take concerns of investors seriously

Invest .

4.30

M

Issuers understand the CRA’s methodology

Trans.

4.27

L

CRA keeps parties informed about developments

Coop.

4.26

M

Issuers keep CRA informed about developments

Coop.

4.23

L

Analyst keen to understand issuer’s organisation

Value

4.22

L

Analysts’ promotion prospects not dependent on anc. bus.

Ind.

4.22

H

80

CREDIT RATING AGENCIES: MEETING THE NEEDS OF THE MARKET?

Table 6.2

Ranking of questionnaire items (Continued)

Questionnaire item

Char.

Mean

Disp.

Issuer and CRA keep each other informed

Coop.

4.21

M

CRA regularly evaluates performance of its staff

Proc.

4.21

L

Analyst regularly meets with issuer’s key personnel

Issue.

4.21

L

Analyst’s salary not linked to issuer fees

Ind.

4.18

H

CRA is proactive in its ratings

Time.

4.17

M

CRA’s main responsibility if to market as a whole

Value

4.16

L

CRA publishes all assumptions relevant to its decisions

Trans.

4.16

L

High level of trust between issuer and CRA

Trust

4.11

H

CRA tells issuers about its targets for their performance

Trans.

4.07

M

CRA tries to resolve problems with the issuer

Resp.

4.05

M

High level of trust between investor and CRA

Trust

4.03

M

Ratings take cyclicality into account

Meth.

4.02

M

Investors understand CRA’s methodology

Trans.

4.00

M

Issuer’s relationship with CRA is based on trust

Trust

3.99

M

Analyst should not change repeatedly

Trust

3.98

M

CRA staff have appropriate workload

Proc.

3.97

M

Issuers have time to respond to any CRA actions

Time.

3.95

M

Analyst easily contactable by issuers

Resp.

3.93

M

CRA provides issuers with individual attention

Resp.

3.92

M

CRA staff undertake regular professional development

Proc.

3.89

M

CRA adjusts ratings if targets for issuers are not me

Trans.

3.86

M

Ratings of different CRAs are comparable

Meth.

3.84

H

Frequent communication between issuer and CRA

Trans.

3.83

M

CRA tells investors its targets for an issuer’s performance

Trans.

3.83

M

81

RESULTS OF THE QUESTIONNAIRE SURVEY: ALL GROUPS

Table 6.2

Ranking of questionnaire items (Continued)

Questionnaire item

Char.

Mean

Disp.

Ratings only adjusted when long-term factors change

Time.

3.77

H

CRA provides structured finance rating

SP

3.76

H

CRA adapts methodology for specific circumst. of issuers

Meth.

3.74

H

CRA conducts client service review meetings

Issue.

3.74

M

CRA discloses detailed fee scheme

Issue.

3.73

H

CRA identifies unsolicited ratings

Trust

3.71

H

CRA provides info. about its method to inv. and op.

Invest.

3.66

H

Issuers provide CRA with non-public information

Coop.

3.62

H

Analyst is contactable by investors

Resp.

3.46

H

CRA enables discussions between analysts and investors

Invest.

3.39

H

CRA is not owned by a financial institution

Ind.

3.24

VH

CRA does not provide ancillary business services

Ind.

3.20

H

CRA has other clients in the issuer’s industry

Exp.

3.19

H

CRA has some discretion in negotiating rating

Ind.

3.13

H

CRA provides ratings advisory services

SP

3.13

H

CRA rates a wide range of issuers

Exp.

3.13

H

CRA provides investors with individual attention

Resp.

3.00

H

CRA’s revenues from ancillary services