Accounting for Financial Instruments:

An Investigation of Preparer and User

Preference for Fair Value Accounting

Chyi Woan (Rebecca) TAN

Doctor of Philosophy
(Accounting)
Murdoch University
Perth, Western Australia.

This thesis is presented for the degree of Doctor of

Philosophy of Murdoch University 2005
 Declaration

I declare that this thesis is my own account of my research and contains work
which has not previously been submitted for a degree at any tertiary education
institution.

......................................................
(Chyi Woan Rebecca TAN)
Dedicated to My Friends & Family,
Both Past & Present,

For all the support through

this PhD journey
&
For giving meaning to my life
 ACKNOWLEDGMENTS
This thesis would not have taken shape if not for the dedicated guidance, assistance
and support of a myriad of individuals. First and foremost, my heartfelt thanks and
gratitude is extended to all three of my supervisors, Professor Greg Tower, Associate
Professor Phil Hancock and Dr Ross Taplin, without whose selfless and dedicated
supervision and encouragement, I would not have made it. Their expertise has
greatly improved this thesis, their sense of humour helped in trying times and their
words of encouragement are blessings from heaven. It has been a privilege and a
great honour to work under three such knowledgeable, intuitive and understanding
supervisors.

My sincerest appreciation is also extended to the academic staff of both Murdoch

Business School and the School of Business and Information Management at the
ANU who are always forthcoming with words of encouragement, expert advice and
critical comments. Special thanks to Professor Monroe for being so helpful and
understanding and for making it possible for me to find time to finish the thesis. I
would also like to thank Professor Gregor and Professor Izan for research advice and
constant encouragement. Thanks Jo Kestel, Kuan Lim and Kerri Carr for helping me
in the most trying of times and for listening to my complaints. I am also indebted to
the computer technicians for their help and technical support whenever it was
required; especially Michael and Robert at the School of Business and Information
Management.

I am also grateful for my peers at Murdoch; Abdul, Lynn, Joyce, Emita, Stacey and
Alistair for their enthusiasm, encouragement, support, critical feedback and mostly
for offering to proof-read my thesis. I thank them wholeheartedly. I would also like
to thank fellow PhD students at the School of Business and Information Management
at the Australian National University, especially Leanne, Teh, Alex and Sigi for
sharing the burden of trying to finish a thesis with the help of really strong coffees.
Also, sincerest thanks goes out to all respondents and participants in this study,
conference and seminar participants for their critical comments and advise that
played a big part in shaping this thesis.

To all my friends whom I had the pleasure of knowing over the years without whose
support and friendship I could not have gone this far in life. I extend my sincerest
thanks especially to Brendon (for adding spice to my life), Sigi (PhD buddy), Alex
(helpful coffee drinking mate), Royston (for words of advice and drinking mate) and
Teh (for his critical thoughts) and Meredith and Toni (for being so supportive even
thought they have just met me). I am also deeply grateful to Chris, Alistair, Juliana
and Mike for proof-reading my chapters.

Last but not least, I am forever indebted to my mum and dad whose love, sacrifice
and support allowed me to complete my education and for making it possible for my
achievements today. Thank you for giving me this opportunity to extend my
horizons and especially for letting me study. To my sisters, Rosemary and Tian
Huey, and to my brother Tian Shyong, thanks for putting up with a bossy elder sister.

i
 ABSTRACT
This research study, motivated by the difference in opinion between the Joint
Working Group of Standard Setters (JWGSS) and the Joint Working Group of
Banking Associations (JWGBA), generates empirical evidence on preparer and user
preferences for fair value accounting for all financial instruments. Australian and
Singaporean respondents’ perceptions on the measurement of financial instruments at
fair value and the recognition of changes in fair value as gains or losses in the
Income Statement are obtained. This study provides better understanding of the
perceptions on the international proposals for change by examining possible
explanatory factors for respondents’ views.

There is ongoing controversy regarding the appropriate accounting for financial

instruments. Perceived shortcomings of the mixed measurement model resulted in an
all inclusive fair value accounting standard proposed by the JWGSS (2000). This was
met with strong criticism from the JWGBA, established in response to this proposal
(1999a). An understanding of actual preparer and user views is thus important for
standard-setters to determine the most appropriate and acceptable accounting
standard for financial instruments, particularly for the complex financial institutions
industry in these two prominent financial markets of the Asian Pacific region.

A positivist-objectivist approach is chosen as the theoretical perspective of this

research study because of its ability to help explain real world phenomena. Both
qualitative (interviews) and quantitative (surveys) methods are used in tandem to
derive evidence on user and preparer perceptions. This integration of methods is
important to achieve a better understanding of the issues at hand. Evidence collected
from the interviews and the preparer and user surveys are analysed with univariate
and multivariate statistical tools to determine the level of support (or opposition) for
fair value accounting for all financial instruments and to identify factors that explain
user and preparer views.

Results show that on average, preparers neither strongly support nor strongly oppose
the fair value accounting proposal, while users are slightly more supportive.
However, respondent users and preparers tend to have similar perceptions on most of
the contentious issues raised in this international debate, sometimes giving credence
to the JWGBA position while other times agreeing with the JWGSS. On average,
users and preparer responses are similar but there is substantial variation within each
group. Findings indicate higher support for fair value accounting when the trading
and banking books are perceived to be not different, fair values for non-traded
financial instruments are reliable and when there is comparability across entities.

This thesis generates empirical evidence on the highly topical issue of accounting for
financial instruments in the midst of international accounting standard setting
movements toward fair value accounting. The lack of variation between users and
preparers affirms the robustness of the qualitative characteristics espoused by the
IASB framework for financial reporting. Results show that achieving these
qualitative characteristics is far more important than trying to fulfill the different
needs of various groups. This slight support for fair value accounting is part of a
bigger tapestry of a slow but steady movement towards fair value measurement in
financial accounting and reporting.

ii
 TABLE OF CONTENTS
CHAPTER ONE: OVERVIEW OF RESEARCH PROJECT
1.1 INTRODUCTION............................................................................................ 1
1.2 RESEARCH QUESTIONS.............................................................................. 5
1.3 SIGNIFICANCE OF PROJECT ...................................................................... 6
1.4 RESEARCH PROJECT DESIGN ................................................................... 8
1.5 RESEARCH PROJECT CONSTRAINTS..................................................... 10
1.6 OUTLINE OF THE THESIS ......................................................................... 13
1.7 SUMMARY ................................................................................................... 15

CHAPTER TWO: LITERATURE REVIEW

2.1 INTRODUCTION.......................................................................................... 16
2.2 THE CONTROVERSY.................................................................................. 16
2.2.1 The Standard Setters’ View................................................................ 17
2.2.2 The Bankers’ View............................................................................. 19
2.2.3 Key Themes Raised in the Controversy ............................................. 21
2.3 ACCOUNTING FOR FINANCIAL INSTRUMENTS ................................. 22
2.3.1 Definitions and Background............................................................... 22
2.3.2 Current Accounting Practice .............................................................. 25
2.3.3 The Proposed Fair Value Accounting Model..................................... 26
2.3.4 Current Developments........................................................................ 30
2.3.4.1 International Accounting Developments............................................ 30
2.3.4.2 Australian Changes ............................................................................ 34
2.3.4.3 Singaporean Situation ........................................................................ 34
2.4 THE CASE OF FAIR VALUE ACCOUNTING........................................... 35
2.5 SURVEYS ON FAIR VALUE ACCOUNTING........................................... 38
2.6 RECENT STUDIES....................................................................................... 41
2.6.1 Fargher (2001) on an AFMA (2000) Survey...................................... 41
2.6.1.1 Background to Fargher (2001) ............................................. 41
2.6.1.2 Fargher (2001) Results ......................................................... 42
2.6.2 Bradbury’s Reliability Study (2001) .................................................. 43
2.6.3 Links to this Project............................................................................ 45

iii
2.7 USER INFORMATION NEEDS................................................................... 46
2.7.1 Objectives of Financial Reporting...................................................... 46
2.7.2 Different User Groups, Different User Needs.................................... 51
2.7.3 Qualitative Characteristics of Useful Information ............................. 53
2.7.3.1 Relevance ............................................................................. 53
2.7.3.2 Reliability ............................................................................. 54
2.7.3.3 Understandability ................................................................. 55
2.7.3.4 Comparability....................................................................... 56
2.7.3.5 Timeliness ............................................................................ 57
2.7.3.6 Cost versus Benefit............................................................... 57
2.7.4 An Expectations Gap.......................................................................... 58
2.8 SUMMARY ................................................................................................... 59

CHAPTER THREE: RESEARCH APPROACH

3.1 INTRODUCTION.......................................................................................... 60
3.2 THE RESEARCH PROCESS........................................................................ 61
3.3 EPISTEMOLOGY - OBJECTIVISM............................................................ 62
3.4 THEORETICAL PERSPECTIVE – POSITIVISM....................................... 64
3.5 RESEARCH METHODOLOGY ................................................................... 65
3.6 RESEARCH METHODS – QUALITATIVE PHASE .................................. 68
3.6.1 Key Interviews.................................................................................... 70
3.6.2 Phone Surveys .................................................................................... 71
3.6.3 Statistical Analyses............................................................................. 72
3.7 RESEARCH METHODS – QUANTITATIVE PHASE ............................... 73
3.7.1 Mail Surveys....................................................................................... 73
3.7.1.1 The Preparer Survey............................................................. 75
3.7.1.2 The User Survey................................................................... 76
3.7.2 Statistical Analyses............................................................................. 77
3.8 DATA CLEANING ....................................................................................... 79
3.8.1 Accuracy of Data................................................................................ 80
3.8.2 Missing Values Analysis .................................................................... 80
3.8.3 Multicollinearity ................................................................................. 81
3.8.4 Fulfilment of Statistical Assumptions ................................................ 81
3.8.5 Identification of Outliers .................................................................... 82
3.9 SUMMARY ................................................................................................... 83

iv
CHAPTER FOUR: QUALITATIVE PHASE
4.1 INTRODUCTION.......................................................................................... 84
4.2 INSIGHTS FROM KEY INTERVIEWS....................................................... 84
4.2.1 Preparer Views ................................................................................... 85
4.2.2 Standard Setter Thoughts ................................................................... 88
4.2.3 Sophisticated Users’ Perceptions ....................................................... 89
4.2.4 Academic Views................................................................................. 91
4.2.5 Summary of Interviews ...................................................................... 92
4.2.6 Key Themes Identified ....................................................................... 93
4.3 PHONE SURVEY RESULTS ....................................................................... 95
4.3.1 The First Phone Survey ...................................................................... 95
4.3.1.1 Demographics....................................................................... 95
4.3.1.2 Questions and Responses ..................................................... 96
4.3.2 The Second Phone Survey................................................................ 101
4.3.2.1 Demographics..................................................................... 102
4.3.2.2 Questions and Responses ................................................... 102
4.4 PRELIMINARY CONCLUSIONS.............................................................. 106
4.5 SUMMARY ................................................................................................. 108

CHAPTER FIVE: QUANTITATIVE PHASE – PREPARER VIEWS

5.1 INTRODUCTION........................................................................................ 109
5.2 SURVEY PARAMETERS .......................................................................... 109
5.3 TESTING NON-RESPONSE BIAS ............................................................ 111
5.4 PREPARER DESCRIPTIVES..................................................................... 112
5.4.1 Type of Financial Institution by Country ......................................... 112
5.4.2 Size of Financial Institution by Country .......................................... 113
5.4.3 Level of Experience by Country....................................................... 114
5.4.4 Level of Experience by Size............................................................. 115
5.4.5 Level of Experience by Type of Financial Institution...................... 116
5.4.6 Size by Type of Financial Institution ............................................... 117
5.5 CURRENT USE OF FAIR VALUE ACCOUNTING ................................ 117
5.6 KEY THEMES AND POSSIBLE INFLUENCES ...................................... 119
5.7 DISTINCTION BETWEEN TRADING AND BANKING ........................ 120
5.7.1 Regression Results............................................................................ 122
5.7.1.1 Country By Type of Financial Institution Effect ............... 125

v
5.8 QUALITATIVE CHARACTERISTICS...................................................... 127
5.8.1 Regressions for Relevance of Fair Value Accounting ..................... 129
5.8.1.1 Country Effect .................................................................... 131
5.8.1.2 Type of Financial Institution Effect ................................... 131
5.8.2 Regressions for Reliability of Fair Values ....................................... 133
5.8.2.1 Traded Financial Instruments............................................. 133
5.8.2.1.1 Type of Financial Institution Effect................... 135
5.8.2.2 Non-Traded Financial Instruments..................................... 136
5.8.2.2.1 Type of Financial Institution Effect................... 137
5.8.2.2.2 Size Effect.......................................................... 138
5.8.3 Regressions for Comparability of Fair Value Information............... 139
5.8.3.1 Type of Financial Institution Effect ................................... 141
5.9 CONCERNS REGARDING VOLATILITY ............................................... 142
5.9.1 Regressions on Concerns Regarding Volatility................................ 142
5.10 PREFERENCE FOR FAIR VALUE ACCOUNTING................................ 144
5.10.1 Trading and Banking Books............................................................. 145
5.10.2 Relationship between the Four Questions: Additional Analysis...... 147
5.10.2.1 Cross-Tabulations............................................................... 147
5.10.2.2 Four Becomes One ............................................................. 150
5.10.3 Overall Preference for Fair Value Accounting................................. 152
5.10.3.1 Regression Results ............................................................. 153
5.10.3.2 Country by Type of Financial Institution Effect ................ 157
5.10.3.3 Distinction between Banking and Trading Book Effect .... 159
5.10.3.4 Reliable Non-traded Financial Instrument FV Effect ........ 159
5.10.3.5 Support for Comparability Effect....................................... 160
5.11 PREPARER VIEWS ON KEY THEMES: SUMMARY ............................ 161
5.12 PREPARER VIEWS ON OTHER ISSUES................................................. 163
5.13 SUMMARY OF PREPARER VIEWS ........................................................ 165

CHAPTER SIX: QUANTITATIVE PHASE – USER VIEWS

6.1 INTRODUCTION........................................................................................ 167
6.2 USER SURVEY PARAMETERS ............................................................... 167
6.3 TESTING NON-RESPONSE BIAS ............................................................ 168
6.4 USER DESCRIPTIVES............................................................................... 169
6.4.1 Type of User by Country.................................................................. 169
6.4.2 Level of Experience by Country....................................................... 170
6.4.3 Type of User by Level of Experience............................................... 171

vi
6.5 USAGE OF FAIR VALUE INFORMATION............................................. 171
6.6 KEY THEMES AND POSSIBLE INFLUENCES ...................................... 172
6.7 DISTINCTION BETWEEN TRADING AND BANKING ........................ 174
6.7.1 Regression Results............................................................................ 175
6.8 QUALITATIVE CHARACTERISTICS...................................................... 176
6.8.1 Regressions for Relevance of Fair Value Accounting ..................... 179
6.8.1.1 Type of User Effect ............................................................ 180
6.8.2 Regressions for Reliability of Fair Values ....................................... 181
6.8.2.1 Traded Financial Instruments............................................. 181
6.8.2.1.1 Type of User Effect ........................................... 183
6.8.2.2 Non-traded Financial Instruments ...................................... 184
6.8.2.2.1 Country by Type of User Effect ........................ 185
6.8.3 Regressions for Comparability of Fair Value Information............... 186
6.9 CONCERNS REGARDING VOLATILITY ............................................... 188
6.9.1 Regressions on Concerns Regarding Volatility................................ 189
6.9.1.1 Country by Type of User Effect ......................................... 190
6.10 SUPPORT FOR FAIR VALUE ACCOUNTING ....................................... 191
6.10.1 Regression Results............................................................................ 192
6.10.1.1 Level of Experience Effect................................................. 196
6.10.1.2 Distinction between Banking and Trading Book Effect .... 196
6.10.1.3 Reliability of Non-traded Financial Instruments Effect ..... 196
6.10.1.4 Volatility Will Not be Misunderstood Effect ..................... 197
6.11 USER VIEWS ON KEY THEMES: SUMMARY ...................................... 197
6.12 USER VIEWS ON OTHER ISSUES........................................................... 199
6.12.1 General Usefulness of Fair Value Accounting................................. 200
6.12.2 More on Relevance........................................................................... 201
6.12.2.1 Fair Value Accounting for Different Categories................ 201
6.12.2.2 Materiality .......................................................................... 202
6.12.3 More on Reliability........................................................................... 203
6.12.4 Understandability and Comparability Over Time ............................ 205
6.12.5 Usefulness of Mixed Measurement .................................................. 206
6.12.6 User Views on Other Issues: Summary............................................ 207
6.13 MORE ON QUALITATIVE CHARACTERISTICS .................................. 208
6.13.1 Relevance ......................................................................................... 208
6.13.2 Reliability ......................................................................................... 209
6.13.3 Qualitative Characteristics Equals Useful Information? .................. 210
6.14 USER VIEWS: SUMMARY ....................................................................... 212

vii
CHAPTER SEVEN: USER VERSUS PREPARER VIEWS
7.1 INTRODUCTION........................................................................................ 214
7.2 PREPARER VERSUS USER EXPERIENCE............................................. 214
7.3 COMPARISON OF PREPARER AND USER VIEWS.............................. 215
7.3.1 Distinction Between Trading and Banking ...................................... 215
7.3.2 Qualitative Characteristics................................................................ 217
7.3.2.1 Relevance of Fair Value Accounting ................................. 217
7.3.2.2 Reliability of Fair Values of Financial Instruments ........... 218
7.3.2.2.1 Traded Financial Instruments ............................ 218
7.3.2.2.2 Non-Traded Financial Instruments .................... 219
7.3.2.3 Comparability of Fair Value Accounting........................... 220
7.3.3 Concerns Regarding Volatility ......................................................... 221
7.3.4 Preference for fair value accounting................................................. 222
7.3.5 Comparison of Preparer and User Views: Summary ....................... 224
7.4 SUMMARY ................................................................................................. 225

CHAPTER EIGHT: IMPLICATIONS & CONCLUSIONS

8.1 OVERVIEW................................................................................................. 226
8.2 SUMMARY OF FINDINGS ....................................................................... 226
8.2.1 Preference for Fair Value Accounting.............................................. 227
8.2.2 Alignment of Views ......................................................................... 228
8.2.2.1 Alignment of User Views................................................... 230
8.2.2.2 Alignment of Preparer Views............................................. 232
8.3 IMPLICATIONS OF FINDINGS................................................................ 234
8.3.1 Accounting for Financial Instruments: Political Process?................ 234
8.3.2 Classic Concerns on Volatility, Reliability and Comparability ....... 240
8.3.3 Complexity of Topic and Need for Education ................................. 244
8.3.4 Other Implications ............................................................................ 246
8.4 IMPORTANCE OF THIS RESEARCH ...................................................... 247
8.5 FUTURE SUGGESTIONS AND RESEARCH IDEAS.............................. 249
8.6 FAIR VALUE ACCOUNTING - FINAL WORDS .................................... 251

REFERENCES......................................................................................................... 254

APPENDICES ..................................................................................................... A1-J6

viii
 LIST OF APPENDICES

APPENDIX A AUSTRALIAN PREPARER SURVEY .............................. A1-A4

APPENDIX B AUSTRALIAN USER SURVEY ........................................ B1-B4

APPENDIX C SINGAPOREAN PREPARER SURVEY............................ C1-C4

APPENDIX D SINGAPOREAN USER SURVEY...................................... D1-D4

APPENDIX E PROTOTYPE PREPARER SURVEY.................................. E1-E7

APPENDIX F PILOT-TESTED PREPARER SURVEY ............................. F1-F6

APPENDIX G LAST PHONE SURVEY QUESTION...................................... G1

APPENDIX H PILOT STUDY .................................................................. H1-H13

APPENDIX I TESTING FOR NON-RESPONSE BIAS .............................. I1-I4

APPENDIX J TRADING VERSUS BANKING BOOK.............................J1-J10

ix
 LIST OF TABLES
Table 3.1: Different Stances of the Research Process ........................................... 62
Table 3.2: Differences between Objectivism and Non-objectivism Research....... 63
Table 4.1: Preparers’, Standard Setters’, Sophisticated Users’ and Academics’
Views: Interview Summary .................................................................. 92
Table 4.2: Summary of Preparer, User, Standard Setters’ and Academics’
Interviews.............................................................................................. 94
Table 4.3: Characteristics of Phone Survey Respondents...................................... 96
Table 4.4: Relevance of Fair Value Accounting for all Financial Instruments ..... 96
Table 4.5: Qualitative Characteristics of Fair Value Accounting.......................... 99
Table 4.6: Users’ Perception of the Need for the Separation of Fair Value
Figures from Historical Cost Numbers............................................... 100
Table 4.7: Characteristics of Respondents of the Second Phone Survey............. 102
Table 4.8: Phone Surveyed Users’ Views on the Relevance of Fair Value
Accounting for all Financial Instruments ........................................... 102
Table 4.9: Phone Surveyed Users’ Views on the Qualitative Characteristics of
Fair Value Accounting for all Financial Instruments ......................... 104
Table 4.10: Phone Surveyed Users’ Views on the Usefulness of the Mixed
Measurement Model ........................................................................... 105
Table 5.1: Breakdown of Type of Financial Institutions ..................................... 112
Table 5.2: Size of Respondent Financial Institutions .......................................... 114
Table 5.3: The Level of Experience of Preparer Respondents ............................ 115
Table 5.4: Breakdown of Level of Experience by Type of Financial Institution 116
Table 5.5: The Extent Fair Value Accounting is Currently Used by Preparers
for Different Categories of Financial Instruments .............................. 117
Table 5.6: Level of Preparers’ Support for the Statement on the Fundamental
Difference between the Trading and Banking Books ......................... 121
Table 5.7: Regression Results for Preparers’ Perception of the Distinction
between the Trading and Banking Book............................................. 122
Table 5.8: Preparers’ Perception of the Distinction between the Trading and
Banking Book by Country by Type of Financial Institution .............. 125
Table 5.9: Level of Preparers’ Perception of the Relevance and Comparability
of Fair Value Accounting ................................................................... 127
Table 5.10: Level of Preparers’ Perception of the Reliability of Traded and Non-
traded Financial Instruments............................................................... 128
Table 5.11: Regression Results for Preparers’ Perception of the Relevance of
Fair Value Accounting........................................................................ 129
Table 5.12: Preparers’ Perception of the Relevance of Fair Value Accounting by
Country ............................................................................................... 131

x
Table 5.13: Preparers’ Perception of the Relevance of Fair Value Accounting by
Type of Financial Institution............................................................... 131
Table 5.14: Regression Results for Preparers’ Perception of the Reliability of
Fair Values for Traded Financial Instruments .................................... 133
Table 5.15: Preparers’ Perception of the Reliability of Fair Values for Traded
Financial Instruments by Type of Financial Institution...................... 135
Table 5.16: Regression Results for Preparers’ Perception of the Reliability of
Fair Values for Non-Traded Financial Instruments............................ 136
Table 5.17: Preparers’ Perception of the Reliability of Fair Values for Non-
Traded Financial Instruments by Type of Financial Institution ......... 138
Table 5.18: Regression Results for Preparers’ Perception of the Comparability of
Fair Value Accounting........................................................................ 139
Table 5.19: Preparers’ Perception of the Comparability of Fair Value Accounting
by Type of Financial Institution.......................................................... 141
Table 5.20: Level of Preparers’ Perception of the Potential Problem with
Volatility in Reported Earnings Resulting from Fair Value
Accounting.......................................................................................... 142
Table 5.21: Regression Results for Preparers’ Concerns on Volatility in Reported
Earnings from the Use of Fair Value Accounting .............................. 143
Table 5.22: Preparers’ Level of Support for Fair Value Accounting in the
Trading and Banking Book................................................................. 146
Table 5.23: Cross Tabulation of Preparers’ Responses to the Two Questions on
Fair Value Accounting for the Trading Book ..................................... 147
Table 5.24: Cross Tabulation of Preparers’ Responses to the Two Questions on
Fair Value Accounting for the Banking Book .................................... 148
Table 5.25: Correlation between the Four Questions ............................................ 149
Table 5.26: Cross Tabulation of Preparers’ Minimum Responses to Q1a and Q1b
and Q2a and Q2b ................................................................................ 150
Table 5.27: Cross Tabulation of Preparers’ Minimum Responses and Average
Responses to the Four Questions ........................................................ 151
Table 5.28: Preparers’ Overall Support for Fair Value Accounting ...................... 153
Table 5.29: Regression Results for Overall Preparers’ Preference for Fair Value
Accounting.......................................................................................... 154
Table 5.30: Preparers’ Preference for Fair Value Accounting by Country by
Type of Financial Institution............................................................... 158
Table 5.31: Cross Tabulation of Preparers’ Responses on Comparability and
their Overall Preference for Fair Value Accounting........................... 161
Table 5.32: Summary of Preparers’ Perceptions and Predictors on the Four Key
Themes................................................................................................ 162
Table 5.33: Level of Preparers’ Agreement on Other Issues Raised in the Fair
Value Accounting Debate ................................................................... 163

xi
Table 5.34: The Percentage of Financial Assets and Liabilities which Reliable
Fair Values can be Determined by Preparers...................................... 165
Table 6.1: Breakdown of Type of Respondent Users .......................................... 169
Table 6.2: The Level of Experience of User Respondents .................................. 170
Table 6.3: The Extent that Users Use Fair Value Information where Available . 171
Table 6.4: Usage of Fair Value Information for the Banking and Trading Book
by Type of User .................................................................................. 172
Table 6.5: Level of Users’ Support for the Statement on the Fundamental
Difference between the Trading and Banking Books ......................... 174
Table 6.6: Regression Results for Level of Users’ Agreement with the
Purported Fundamental Difference between the Trading and Banking
Books .................................................................................................. 175
Table 6.7: Level of Users’ Perception of the Relevance and Comparability of
Fair Value Accounting........................................................................ 177
Table 6.8: Level of Users’ Perception of the Reliability of Traded and Non-
traded Financial Instruments............................................................... 178
Table 6.9: Regression Results for Users’ Perception of the Relevance of Fair
Value Accounting for all Financial Instruments................................. 179
Table 6.10: Users’ Perception of the Relevance of Fair Value Accounting by
Type of User ....................................................................................... 180
Table 6.11: Regression Results for Users’ Perception of the Reliability of Fair
Values for Traded Financial Instruments............................................ 182
Table 6.12: Users’ Perception of the Reliability of Fair Values for Traded
Financial Instruments by Type of User............................................... 183
Table 6.13: Regression Results for Users’ Perception of the Reliability of Fair
Values for Non-traded Financial Instruments..................................... 184
Table 6.14: Users’ Perception of the Reliability of Fair Values for Non-traded
Financial Instruments by Country by Type of User ........................... 185
Table 6.15: Regression Results for Users’ Perception of the Comparability of
Fair Value Accounting........................................................................ 186
Table 6.16: Level of Users’ Perception of the Potential Problem with Volatility
in Reported Earnings .......................................................................... 188
Table 6.17: Regression Results for Users’ Perception of Volatility in Reported
Earnings from the Use of Fair Value Accounting .............................. 189
Table 6.18: Users’ Perception that the Volatility in Earnings Introduced by Fair
Value Accounting will be Misunderstood by Users by Country by
Type of User ....................................................................................... 191
Table 6.19: Users’ Support for Fair Value Accounting for All Financial
Instruments.......................................................................................... 192
Table 6.20: Regression Results for Users’ Preference for Fair Value Accounting
for all Financial Instruments ............................................................... 193
Table 6.21: Summary of Users’ Perceptions of the Four Key Themes ................. 198

xii
Table 6.22: Users’ Level of Agreement with General Usefulness of Fair Value
Accounting for All Financial Instruments .......................................... 200
Table 6.23: The Extent Users Regard Fair Value Accounting as Relevant for
Different Categories of Financial Instruments.................................... 202
Table 6.24: Level of Users’ Perception of the Materiality of Information
Resulting from Fair Value Accounting............................................... 203
Table 6.25: Level of Users’ Perception of the Reliability of Fair Value
Accounting for All Financial Instruments .......................................... 204
Table 6.26: Level of Users’ Perception of the Understandability and
Comparability Over Time of Fair Value Accounting in Financial
Statements........................................................................................... 205
Table 6.27: Users’ Level of Agreement with Usefulness of Current Mixed
Measurement of All Financial Instruments ........................................ 206
Table 6.28: Summary of Users’ Perceptions of the Other Issues .......................... 207
Table 6.29: Correlation between the Two Materiality Criteria and Relevance
from a Sophisticated User Perspective ............................................... 209
Table 6.30: Correlations between the Various Attributes of Usefulness and
Overall Usefulness of Accounting Information: Based on User
Responses............................................................................................ 211
Table 7.1: Comparison of Preparer and User Respondents’ Experience............. 214
Table 7.2: Comparison of Preparer and User Perceptions of the Trading Versus
Banking Distinction ............................................................................ 216
Table 7.3: Comparison of Preparer and User Perceptions of the Relevance of
Fair Value Accounting........................................................................ 217
Table 7.4: Comparison of Preparer and User Perceptions of the Reliability of
Fair Values for Traded Financial Instruments .................................... 218
Table 7.5: Comparison of Preparer and User Perceptions of the Reliability of
Fair Values for Non-Traded Financial Instruments............................ 219
Table 7.6: Comparison of Preparer and User Perceptions of the Comparability
of Fair Value Accounting Financial Statements Across Entities........ 221
Table 7.7: Comparison of Preparer and User Perceptions of the Earnings
Volatility Resulting from Fair Value Accounting .............................. 222
Table 7.8: Comparison of Preparer and User Preference for Fair Value
Accounting.......................................................................................... 223
Table 7.9: Summary of the Comparison of Preparer and User Views................. 224
Table 8.1: Summary of Findings for Both Research Questions .......................... 227
Table 8.2: Comparison of User Interviews, User Surveys and JWGSS and
JWGBA Assertions............................................................................. 231
Table 8.3: Comparison of Preparer Interviews, Preparer Surveys and JWGSS
and JWGBA Assertions ...................................................................... 233

xiii
 LIST OF FIGURES
Figure 1.1: Flowchart of Research Project................................................................ 9
Figure 1.2: The Two Phases of Research Evidence ................................................ 10
Figure 3.1: The Relationship between the Components of the Research Process .. 61
Figure 3.2: Preferred Methodologies for Different Epistemologies........................ 66
Figure 3.3: Research Methodology Adopted .......................................................... 67
Figure 3.4: A Knowledge Accrual Triangle............................................................ 69
Figure 3.5: Graphical Representation of Chapter Coverage ................................... 83
Figure 5.1: Possible Predictors of Preparer Views on Key Themes ..................... 119
Figure 5.2: Australian and Singaporean Preparers’ on the Distinction between
the Trading and Banking Book ........................................................... 126
Figure 5.3: Preparers’ Perception of the Relevance of Fair Value Accounting by
Type of Financial Institution............................................................... 132
Figure 5.4: Preparers’ Perception of the Reliability of Fair Values of Traded
Financial Instruments by Type of Financial Institution...................... 135
Figure 5.5: Preparers’ Perception of the Reliability of Fair Values of Non-traded
Financial Instruments by Type of Financial Institution...................... 138
Figure 5.6: Preparers’ Perception of the Comparability of Fair Value Accounting
by Type of Financial Institution.......................................................... 141
Figure 5.7: Australian and Singaporean Preparers’ on Overall Preference for
Fair Value Accounting........................................................................ 158
Figure 6.1: Diagram of Possible Predictors of User Views on Key Themes ........ 173
Figure 6.2: Users’ Perception of the Relevance of Fair Value Accounting by
Type of User ....................................................................................... 181
Figure 6.3: Users’ Perception of the Reliability of Fair Values of Traded
Financial Instruments by Type of User............................................... 183
Figure 6.4: Users’ Perception of the Reliability of Fair Values of Non-traded
Financial Instruments by Country by Type of User ........................... 185
Figure 6.5: Users’ Perception that the Volatility in Earnings Introduced by Fair
Value Accounting will be Misunderstood by Users by Country by
Type of User ....................................................................................... 190
Figure 8.1: Spectrum of JWGBA versus JWGSS Views: Compared to Thesis
Evidence from Users and Preparers.................................................... 229
Figure 8.2: Recent Developments of IAS 39 and the Influences that Played a
Part ...................................................................................................... 236

xiv
 LIST OF APPENDICES TABLES & FIGURES
Table G1: Phone Surveyed Users’ Views on the Reasons for Low Response Rate
............................................................................................................... G-1
Table H.1: The Extent that Financial Instruments are Measured at Fair Value in
the Balance Sheet. ................................................................................. H-2
Table H.2: The Extent that Changes in Fair Value of each of the Financial
Instruments is Taken to the Income Statement ..................................... H-3
Table H.3: Level of Support for Fair Value Accounting. ....................................... H-5
Table H.4: Independent Samples T-test Results on the Significance of Size on
Respondents’ Support for Fair Value Accounting. ............................... H-6
Table H.5: Level of Support for the JWGBA Reasons for the Fundamental
Difference between the Trading and Banking Books ........................... H-7
Table H.6: Level of Support for the JWGBA Views on the Reasons Against Fair
Value Accounting.................................................................................. H-9
Table H.7: Independent Samples T-test Results on the Significance of Size on
Respondents’ Support for JWGBA Views.......................................... H-10
Table I.1: Regression Results for Overall Preparer Preference for Fair Value
Accounting to Test for Non-response Bias .............................................I-2
Table I.2: Regression Results for User Preference for Fair Value Accounting for
all Financial Instruments to Test for Non-response Bias ........................I-3
Table I.2: Regression Results for User Preference for Fair Value Accounting for
all Financial Instruments to Test for Non-response Bias ........................I-4
Table J.1: Cross Tabulation of Preparer Preference for Fair Valuing Financial
Instruments on the Balance Sheet .......................................................... J-1
Table J.2: Cross Tabulation of Preparer Preference for Fair Value Accounting in
the Banking Book................................................................................... J-2
Table J.3: Cross Tabulation of Preparer Preference for Fair Value Accounting in
the Trading Book.................................................................................... J-3
Table J.4: Regression Results for Preparer Preference for Fair Value Accounting
in the Banking Book............................................................................... J-4
Table J.5: Preparer Preference for Fair Value Accounting by Country by Type of
Financial Institution ............................................................................... J-6
Table J.6: Regression Results for Preparer Preference for Fair Value Accounting
in the Trading Book ............................................................................... J-8
Table J.7: Comparison of Significant Predictors of Preparers' Overall Preference
for Fair Value Accounting, and in the Banking and Trading Books.... J-10

Figure J.1: Australian and Singaporean Preparers Perception of Preference for

Fair Value Accounting in the Banking Book ........................................ J-5

xv
 LIST OF ACRONYMS

ACRONYM DEFINITION

ABA Australian Banking Association

AFMA Australian Financial Markets Association survey

AICPA American Institute of Certified Public Accountants

AIMR Association for Investment Management and Research

APRA Australian Prudential Regulatory Authority

ARCB Association of Reserve City Bankers

BIS Bank for International Settlements

CCDG Council of Corporate Disclosure and Governance

CICA Canadian Institute of Chartered Accountants

DP Discussion Paper

ECB European Central Bank

ED Exposure Draft

EFRAG European Financial Reporting Advisory Group

EMH Efficient Market Hypothesis

FASB Financial Accounting Standards Board

FRC Financial Reporting Council (Australia)

FRS Financial Reporting Standard (Singapore)

FV Fair Value

GLM General Linear Models

IAS International Accounting Standards

IASB International Accounting Standards Board (2001-)

IASC International Accounting Standards Committee and

predecessor of the IASB (period 1973-2001)

ICPAS Institute of Certified Public Accountants of Singapore

xvi
ACRONYM DEFINITION

IFRS International Financial Reporting Standard

IMF International Monetary Fund

INT FRS Interpretations of the Financial Reporting Standards

JWGBA Joint Working Group of Banking Associations

JWGSS Joint Working Group of Standard Setters

JWG is the acronym used on the 2000 Draft Standard but

JWGSS is used in this thesis consistent with the five-letter
JWGBA acronym.

MAS Monetary Authority of Singapore

NAB National Australia Bank

SAS Singapore Accounting Standard

SFAS Statement of Financial Accounting Standard

UK United Kingdom

US United States of America

xvii
 GLOSSARY OF TERMS

TERM DEFINITION
Balance Sheet To avoid confusion, the term ‘Balance Sheet’ is used
throughout the thesis to refer to Statement of Financial
Position

Banking activities Activities consists of raising funds and investing them in

assets in order to make a profit from the margin between
the amount received on interest bearing assets and the
amount paid on interest bearing liabilities.

Derivative financial A financial contract whose value changes in response to

instrument the change in an underlying financial asset such as
shares, property, foreign currency and other tangible and
intangible assets

Equity instrument Any contract that evidences a residual interest in the

assets of an enterprise after deducting all of its liabilities

Fair Value Fair value is an estimate of the price an enterprise would

have received if it had sold the asset or paid if it had
been relieved of the liability on the measurement date in
an arm’s-length exchange motivated by normal business
considerations (JWGSS, 2000).

Fair Value Accounting Fair value accounting is used in this research to refer to
the measurement of all financial instruments at fair
value and the recognition of changes in fair value as
revenues or expenses in the Income Statement in the
period in which they arise (ie. the JWGSS proposed
measurement model)

Fair value option The option that permits entities to designate

(irrevocably) any financial asset (or liability) as one to
be measured at fair value with gains and losses
recognized in the Income Statement

Financial asset Any asset that is cash, a contractual right to receive cash
or another financial asset from another enterprise, a
contractual right to exchange financial instruments with
another enterprise under conditions that are potentially
favourable or an equity instrument of another enterprise

xviii
TERM DEFINITION
Financial instruments Cash; an equity instruments; or contracts that give rise
to both a financial asset of one enterprise and a financial
liability or equity instrument of another enterprise

Financial liability Any liability that is a contractual obligation to deliver

cash or another financial asset to another enterprise or to
exchange financial instruments with another enterprise
under conditions that are potentially unfavourable

Income Statement Similarly, the term ‘Income Statement’ is used

throughout the thesis to refer to Statement of Financial
Performance or otherwise also known as the Profit and
Loss Statement

Preparers In this study, preparers refer to individuals that have to

prepare financial statements for financial institutions.

Trading activities Transactions undertaken with the objective to profit

from short-term fluctuations in market prices

Triangulation “the use of more than one method to gain a more

complete understanding of the level of support for fair
value accounting” by utilising both qualitative and
quantitative approaches in a complementary way
(Ticehurst and Veal, 2000, p. 50).

Users Present and potential investors and creditors who makes

rational investment and credit decisions (FASB, 1978).
In this study, users are represented by auditors, analysts,
investment bankers and fund managers.

xix
 RELATED THESIS PUBLICATIONS

The material in Chapter 4 had been published in:

Tan, C. W. R., Tower, G., Hancock, P. & Taplin, R. (2003). The Unspoken Voice:
What do Users Think of Fair Value Accounting for all Financial Instruments?,
presented at the annual European Applied Business Research Conference, Venice,
Italy (June) and at the Accounting and Finance Association of Australia and New
Zealand (AFAANZ) Annual Conference, Brisbane, Queensland (July).

Tan, C. W. R., Tower, G., Hancock, P. & Taplin, R. (2003). Fair Value Accounting
for all Financial Instruments – User Views, presented at the International Conference
on Quality Financial Reporting and Corporate Governance - Building Public Trust,
Integrity and Accountability, Kuala Lumpur, Malaysia (July).

The material in Chapter 5 had been published in:

Tan, C. W. R., Hancock, P., Taplin, R. and Tower, G. (2004). Fair Value Accounting
for Financial Instruments: Australian Versus Singaporean Preparers’ Perspectives,
presented at the 16th Annual Asian-Pacific Conference on International Accounting
Issues in Seoul, Korea, November 7-10, 2004.

Tan, C. W. R., Hancock, P., Taplin, R. and Tower, G. (2005). The Use of Fair Value
Accounting for Financial Instruments: The Controversy Continues, presented at the
6th Annual Asian-Pacific Journal of Accounting & Economics Symposium in
Guangzhou January 7-8, 2005.

The pilot study mentioned in Chapter 5 as detailed in Appendix H had been

published in:
Tan, C. W. R., Hancock, P., Taplin, R. and Tower, G. (2005). Fair Value Accounting
for all Financial Instruments: Perceptions from Managers of Australian Financial
Institutions, Australian Accounting Review, In press.

The material in Chapters 6 to 8 is contained the following conference submissions:

Tan, C. W. R., Hancock, P., Taplin, R. and Tower, G. (2005). Multiple Views on
Accounting Measurement for Financial Instruments, submitted to the Accounting
and Finance Association of Australia and New Zealand (AFAANZ) Annual
Conference, Melbourne, Victoria (July).

Tan, C. W. R., Tower, G., Hancock, P. & Taplin, R. (2005). Accounting for
Financial Instruments in Financial Institutions: Views from Australian and
Singaporean Users on Fair Value Accounting, submitted to the Accounting and
Finance Association of Australia and New Zealand (AFAANZ) Annual Conference,
Melbourne, Victoria (July).

xx
 CHAPTER ONE:
OVERVIEW OF RESEARCH PROJECT
For the love of money is the root of all Evil – 1 Tim 6:10 (Holy Bible)

1.1 INTRODUCTION

A difference in opinion between two global professional groups forms the crux of

this study. The Joint Working Group of Banking Associations (JWGBA) strongly

disagrees with the views of the Joint Working Group of Standard Setters (JWGSS)

that proposed fair value accounting should be used for all financial instruments (i.e.

the measurement of financial instruments at fair value and the recognition of changes

in fair value as revenues or expenses in the Income Statement)1. To help better

understand this issue, this research project examines the level of support for the use

of fair value accounting for all financial instruments from both a user and preparer

perspective. Possible determinants of their views are also explored.

The growth in the use of financial instruments since the 1980’s has been phenomenal

as spurred by the deregulation of financial markets and financial engineering

innovations (Hancock, 1999). This deregulation caused companies to be exposed to

higher levels of foreign exchange and interest rate risks. Consequently, the financial

markets developed and expanded the range of financial products (for example

forward foreign exchange contracts, interest rate swaps and financial futures) to

assist firms in the management of these (increasing) risks (Daugaard and Valentine,

1
As stated in the glossary of terms, ‘Income Statement’ is used throughout this thesis as per the
IASB framework. However, ‘Statement of Financial Performance’ and ‘Profit and Loss Statement’
are used to refer to the ‘Income Statement’ in the Australian and Singaporean surveys respectively.
As shown in the list of acronyms, JWGSS is used rather than the JWG acronym in the 2000 Draft
Standard for comparative consistency with the JWGBA.

1
1995). Subsequently, there was a large explosion of new products and new markets

created to further accommodate these new instruments (Bianco, 1985; Stewart &

Neuhausen, 1986). However, accounting standards and guidelines for them were

substantially lagging behind (Kay, 1985; Sanborn and Atchison, 1985).

Accounting for financial instruments has recently attracted tremendous attention due

to the enormous growth of the exchange-traded derivative financial instruments

market that enjoyed a turnover for financial futures and options contracts in the third

quarter of 2002 of over 190 trillion US dollars, more than thirty times its level ten

years ago (Bank for International Settlements, 2002). Its sheer size and the

prominence of the market players (large, financial institutions central to the world’s

financial system) mean that the stakes in the derivatives game are very high (Hu,

1993). This rapid growth also brought increased concerns and spectacular losses in

organisations around the world (Guerrera, Parker and Pretzlik, 2003). In Australia,

AWA Ltd suffered losses of A$49.8 million in forward foreign exchange contracts.

Barings PLC is probably the most publicised case involving huge derivative losses

where the company lost in excess of US$1 billion and faced receivership. More

recently in Australia, concerns have been raised over possible billion dollar losses

incurred by the Federal Treasury as a result of losses on certain swap contracts

(Davidson, 2002). In addition, the National Australia Bank (NAB) reported losses

totalling A$360 million in unauthorised foreign exchange options trading (Maiden,

2004; Kemp, 2004).

These events raise important questions on the role of financial reporting and whether

the current accounting for financial instruments and related disclosure allows

investors to make a proper assessment of a company’s risk exposure from its

derivative financial instruments (Matolcsy and Petty, 2001; Crawford, Wilson and

2
Bryan, 1997; Young, 1996; Hancock, 1994; Walker, 1993). Therefore, the then

International Accounting Standards Committee (IASC) (now reconstituted as the

International Accountings Standards Board (IASB)) and the Canadian Institute of

Chartered Accountants (CICA) began a joint project to develop a comprehensive

standard on financial instruments (Bradbury, 2003). The result of this collaboration

was IAS 32 Financial Instruments: Disclosure and Presentation in March 1997 and a

discussion paper (DP) Accounting for Financial Assets and Financial Liabilities.

Comments received on the DP contributed to an interim IAS 39 standard in

December 1998 and a longer term project in collaboration with national standard

setters (Bradbury, 2003). This long term project is in the form of a Joint Working

Group (i.e. JWGSS) to examine the issues associated with accounting for financial

instruments. The JWGSS2 is made up of powerful key accounting standard regulators

(US, UK, Canada, Australia, Germany, France, Japan, New Zealand, five Nordic

countries and the International Accounting Standards Board3). The banking industry

responded by establishing a Joint Working Group of Banking Associations

(JWGBA) made up of representatives from the banking associations of the US,

Australia, Canada, Japan and the European Union.

Following extensive deliberations, the JWGSS, in a draft standard, proposed the use

of fair value accounting to overcome the shortcomings of current measurement

practices (JWGSS, 2000). Under this draft standard, financial instruments (except

certain private equity investments) would be measured at fair value when recognised

and be re-measured at fair value in periods subsequent to initial recognition. Fair

value is defined as the price that arm’s-length market participants would pay or

2
The JWGSS as well as the JWGBA have since been disbanded. However, the diverse opinions that
they advocated are the main focus of this thesis.
3
The new abbreviation of IASB will be used hereafter to represent the International Accounting
Standards Committee.

3
receive in a routine transaction under the market conditions at the date at which it is

to be measured for accounting purposes (JWGSS, 2000). The draft standard further

proposes that the resultant changes in fair value be taken to the Income Statement.

The international financial institutions’ community (as represented by the JWGBA’s

official views) strongly disagrees with any move towards a fair value accounting

model for all financial instruments (JWGBA, 1999a). Of major concern to the

JWGBA is the application of the fair value accounting model to financial instruments

in the banking book. The JWGBA argues that the principal difference between the

banking industry and other sectors is that a financial institution’s financial

instruments earn revenue from both trading and traditional banking activities,

which (they feel) are two entirely different areas (JWGBA, 1999b). Fair value

accounting is used for financial instruments from trading activities and the JWGBA

stressed that it is the appropriate measurement method. However, the JWGBA feels

that the fundamental characteristics of banking activities make the use of fair value

accounting for the banking financial instruments inappropriate (JWGBA, 1999a).

In contrast, the JWGSS favours the adoption of a fair value accounting model for all

financial instruments for inclusion in the general purpose financial statements

because it provides more relevant information for investors and creditors. Fair

values better reflect the effects of current economic conditions and thus, are better

predictions of future conditions (see for example Bernard, Merton and Palepu, 1995;

Barth, Landsman and Wahlen, 1995; Barth, 1994).

This difference of opinion between the JWGSS and the JWGBA highlights two

crucial issues that need to be resolved. The first relates to the appropriateness of the

proposed fair value accounting model for the financial institutions’ industry. In other

4
words, does the industry agree with the views of the JWGBA? This project gathers

evidence on Australian and Singaporean financial institutions’ views regarding the

fair value accounting model in relation to the specific issues and arguments

forwarded by the two global bodies.

The other crucial issue relates to the usefulness of the proposed fair value accounting

model to users of financial institutions’ financial reports (given the purpose of this

report, as per the IASB framework (1989), is to provide information useful for

decision making purposes). This thesis provides crucial information regarding the

rarely heard ‘user voice’ by gathering evidence on the actual views of sophisticated

users4 in Australia and Singapore on fair value accounting for financial instruments.

1.2 RESEARCH QUESTIONS

Based on the discussion above, the specific research questions this study addresses,

in relation to the measurement of financial instruments in financial institutions are:

1) To what extent do preparers and users support the use of fair value
accounting for all financial instruments?

2) What factors help to explain preparers and users’ support for (or opposition
of) the use of fair value accounting for all financial instruments?

The following sub-questions are also explored in relation to the first research

question to enrich the study.

1 a) What is the level of congruence between the assertions of the JWGBA and
the JWGSS with preparers and users?

1 b) To what extent do users and preparers perceive fair value accounting to

possess the qualitative characteristics of useful information?

1 c) How comparable are the preparer viewpoints and user perceptions?

4
In this study, the sophisticated users are represented by auditors and financial analysts.

5
1.3 SIGNIFICANCE OF PROJECT

This research study is important for several reasons. First and foremost, there are

two fundamentally contrasting views as to the most appropriate way of measuring

and accounting for financial instruments. The JWGBA strongly disagrees with any

move by standard setters towards fair value accounting model. These two

contrasting viewpoints form the crux of this study.

Secondly, this study empirically examines preparer and user perceptions on the

proposed and current measurement model. An understanding of preparer and user

views and needs will facilitate the propagation of the most appropriate and most

acceptable accounting standard on financial instruments (especially in the financial

institutions’ industry). In addition, this study identifies potential areas of controversy

and difficulties with fair value accounting for all financial instruments that will

enhance the standard setting process.

Thirdly, the financial institutions industry is arguably one of the most complex

industries in terms of their accounting practices (see Ryan, 2002 for extensive

examples and real-life cases on the difficulties in financial institutions’ statements).

Empirical research into their accounting practices with regards to financial

instruments provides invaluable insights and facilitates a better understanding of the

complex accounting world of financial institutions. Even a decade after the

introduction of IAS 39, it has the reputation for being the most complex standard to

date (Kawaller, 2004). Furthermore, the losses (to the tune of billions of dollars)

incurred by companies trading in derivative financial instruments also highlights the

importance of research into the accounting for financial instruments to enhance the

stewardship and accountability function of accounting.

6
Fourthly, this study employs both qualitative and quantitative methods (also known

as a form of triangulation5) to enable a detailed exploration of the possible

explanation and determinants of the level of support (or opposition) for fair value

accounting. This result in a broader and more complete understanding of the issues

being investigated (Ticehurst and Veal, 2000). Thus, this project provides

multifaceted explanations for the positions and views of both preparers and users.

Fifthly, this study examines preparers and users from two key financial markets in

the Asia-Pacific region, Australia and Singapore. Singapore has a large diversified

financial institutions industry (accounting for 11.6% of Gross Domestic Product) and

ranks among the top ten most sophisticated financial markets in the world (Monetary

Authority of Singapore, 2003). In addition to being a key player in the accounting

standard setting debate internationally, Australia has a strong and independent

domestic accounting standard setting body and thus, is a major influence in the Asia-

Pacific region. Singapore has traditionally placed great efforts in closely following

the IASB lead. The examination of two countries further enhances the understanding

of preparer and user views and facilitates the examination of country effect on

preferences for accounting measurement policies. Singapore was also not a member

of the JWGBA. This allows comparison with views of preparers in Australia, which

was represented on the JWGBA to check for alignment between official and

constituents’ views.

5
There are many definitions (and uses) of the term triangulation. For the purposes of this research
study, triangulation method uses more than one research approach to gain better understanding of
the issues being investigated (Ticehurst and Veal, 2000).

7
Finally, this study also attempts to bridge the gap in the accounting literature in

relation to user information needs. The lack of a user focus in the standard setting

process is referred to as a ‘systemic problem’ by Jonas and Young (1998) who urged

more research into this area. Therefore, this study seeks to better understand the

information that sophisticated users need in the process of making economic

decisions in relation to financial instruments in financial institutions.

Therefore, this research aims to offer users, preparers and the accounting regulatory

bodies a wider picture of the acceptance and decision-usefulness of the fair value

accounting model for all financial instruments within the financial institutions

industry.

1.4 RESEARCH PROJECT DESIGN

Figure 1.1 shows the graphical representation of the whole research project in terms

of the two contrasting views that makes up the perception of the major players in this

international debate on the accounting for financial instruments. These official views

are compared with preparers and users perception of fair value accounting. Actual

preparer and user views are also contrasted and possible determinants of their views

explored.

8
Figure 1.1: Flowchart of Research Project

Contrasting Views Joint Working

Joint Working
Group of Banking Group of
Associations Standard Setters
(JWGBA) (JWGSS)

Perceptions of the major players

PREPARERS USERS
⇒ To what extent do ⇒ To what extent do users
Level of
preparers support fair support fair value
alignment of
value accounting for all accounting for all
financial instruments? user and financial instruments?
⇒ Why do they think that preparer views ⇒ Why do they think that
way? way?

This study assesses the level of support for the use of fair value accounting for all

financial instruments by financial institutions (preparers) and sophisticated users

alike and explores possible reasons that explain the level of support or opposition. In

this capacity, as previously mentioned, this study adopts the triangulation method

defined as “the use of more than one method to gain a more complete understanding

of the level of support for fair value accounting” by utilising both qualitative and

quantitative approaches in a complementary way (Ticehurst and Veal, 2000, p. 50).

In this project, interviews are conducted with various preparers and users to gather

deeper and more detailed insights into the views of preparers and users about the

usefulness of fair value accounting for all financial instruments (i.e. Qualitative

Phase). This qualitative phase also serves to explore the intricate issues involved in

9
this controversy and to define and develop the parameters for the following

quantitative phase. Subsequently, the quantitative phase, adopted to complement the

interviews conducted, utilised the survey method to explore reasons why preparers

and users think the way they do. This process is summarised in Figure 1.2.

Figure 1.2: The Two Phases of Research Evidence

Qualitative Phase
Key Interviews and Phone Surveys
Development of in-depth understanding of the key issues.
Development of a focal point for the quantitative phase.
Identification of key research questions.
Acquisition of preliminary perceptions.

Quantitative Phase
Surveys
Development of user and preparer surveys.
Acquisition of preparer and user perceptions.
Statistical analyses of responses and possible explanations.

Data collected from the survey is subject to empirical testing to develop the profiles

of the sample. Descriptive statistics are employed to provide a general overview of

the demographics of the survey sample and to indicate the distribution of responses

to the survey. In addition, statistical techniques such as T-tests and regressions are

used to test for statistical significance and possible associations in the survey

responses.

1.5 RESEARCH PROJECT CONSTRAINTS

There are three major constraints in this research project: the scope of this study,

underlying assumptions and inherent limitations of such a research project.

10
The accounting standard-setting process is dynamic in nature. Therefore, the scope

of the project needs to be explicit. A myriad of important changes, domestically and

internationally, comes into play over time6. In addition, various initiatives and

research are also being conducted contemporaneously into various aspects of this

proposed fair value accounting measurement model. Due to completion deadlines,

an end cut-off date is necessary for the literature component, which contains

materials through 30 October 2004. In addition, this study only examines preparers

and users from Australia and Singapore.

One of the underlying assumptions of this thesis is that respondents are being truthful

and not using game-playing strategies when responding to interviews and surveys.

There is inherent faith that the results of the interview process is trustworthy and

accurate without unduly bias (Silverman, 1993). It is also assumed that views of

users and preparers alike are important to this whole debate surrounding the

accounting for financial instruments.

This research project deals with highly complex issues and although terminology

used are explained in the surveys, it is also assumed that the preparers (especially the

credit unions in Australia) and sophisticated users that are sought to fill them in have

adequate knowledge of fair value accounting and financial instruments. In addition,

the Likert scale used in the surveys is usually assumed to be (and treated as) an

interval measure.

Although the survey method has been criticised in past research (Kerlinger, 1986), it

is the most appropriate method that could be utilised in this research project to obtain

6
The update of international developments on the accounting for financial instruments is provided
in Chapters Two and Eight.

11
evidence about the actual information needs of users. In addition, the relatively low

response rate (see Chapters Five and Six) further adds to the limitation of this

research. However, the use of key interviews to complement this method serves to

overcome many of the problems associated with surveys. Furthermore, Baruch

(1999) maintains that response rates for academic studies have generally declined in

recent years and Griffis, Goldsby and Cooper (2003) asserted that achieving the

traditional benchmark of 20% usable responses from a mail survey is less and less

common today than ever before. The results of this thesis provide crucial insights

into the controversial issues surrounding the accounting for financial instruments.

A further evidence phase, the use of a user survey, is open to criticism (Marsh, 1982;

De Vaus, 1995), and subject to questioning for its contribution to accounting

research (Young, 1996). One of the main problems eliciting information about users'

perceptions of annual report measurement and disclosure issues is their lack of

understanding of financial information. To overcome this concern, a pilot study is

conducted and a simplified non-technically worded survey instrument is evolved.

The target sample of chief financial institutions and sophisticated users (although

most appropriate) represents another inherent limitation. As very busy professionals

in the financial institutions industry, eliciting responses from both the sophisticated

user and preparer groups is extremely difficult. It is also difficult to identify the so-

called experts due to the complexity of this topic. All these limitations lead to

difficulty in obtaining reasonable response rates. However, the data collected is

checked and non-response bias is not statistically present.

12
This study also focuses on Singaporean and Australian data and thus is only partially

generalisable to the rest of the world. Care is taken to ensure that appropriate

terminology is used for each group and definitions provided for the main variables of

interest in this research.

Despite these constraints, this project offers important insights into actual preparer

and user views on the still controversial and constantly evolving (there are four

exposure drafts issued in 2004 to amend IAS 39 Financial Instruments: Recognition

and Measurement) accounting for financial instruments.

1.6 OUTLINE OF THE THESIS

The thesis is structured as follows:

Chapter One Overview of the research project: The two global

contrasting views that form the crux of this study are
reviewed. Research questions relevant to this study are
proposed to explore the preparer and user views. The scope,
assumptions and limitations of this research are also outlined.

Chapter Two Literature review: The controversy surrounding the two

global joint working groups is summarised in conjunction
with the international developments on this issue. The
literature on the financial institutions industry in both
Australia and Singapore is critiqued. Current research into
financial instruments, financial institutions and fair value
accounting is then reviewed. In addition, the IASB
Framework definition of the qualitative characteristics of
useful financial information is summarised.

13
Chapter Three Research Approach: This study employs an objective-
positivist methodology with qualitative methods used in
conjunction with quantitative/empirical tools. Such a mixed
methodology has tremendous potential in enriching the
understanding of quantitative results. Research methods
adopted in this study include interviews, survey method,
univariate and multivariate statistical tools. Measurement
methods and related documentation are also explained.

Chapter Four Qualitative Phase: Information collected from interviews

with prominent players in the industry from all aspects of the
controversy is detailed. The evolution of the key
themes/issues from the interviews conducted with preparers,
standard setters and users is then provided. Evidence from the
phone surveys conducted with users is analysed. This data is
further examined for possible relationships and statistical
significance.

Chapter Five Quantitative Phase – Preparer Views: This chapter

explores the opinions of chief financial officers from
Australian and Singaporean financial institutions on the key
themes as gleaned from the surveys. Descriptive statistics for
the survey data collected are provided for preparers in both
Australia and Singapore and significant trends identified.

Chapter Six Quantitative Phase – User Views: This chapter details the
results obtained form the major user study and provides the
opinions of users from Australia and Singapore on the key
themes identified from the qualitative phase. Descriptive
statistics for the survey data collected are also provided.

14
Chapter Seven User versus Preparer Views: This chapter examines the
perceptions of both preparers and users. Univariate and
multivariate statistical tools are used to profile the responses
and test for any relationships between the perceptions of both
groups and statistical significance of explanatory factors.

Chapter Eight Implications and Conclusion: The entire project is reviewed

and major findings highlighted. Findings of the qualitative
and quantitative phases are scrutinised and the implications of
the results detailed in the context of past literature and current
developments. Major conclusions are then discussed followed
by proposed future research directions.

1.7 SUMMARY

This chapter sets out the scene of the current controversy between the global standard

setters and the global financial institutions’ community that forms the crux of this

research project. Triangulation, where qualitative means are used to enrich

qualitative results is employed. Key interviews coupled with user and preparer

surveys are adopted to gather views of financial institutions and sophisticated users

in both Australia and Singapore on the decision usefulness of fair value accounting

for all financial instruments by financial institutions. The outline of the thesis

concludes this chapter. The next chapter provides an overview of the past literature

relevant to this research project.

15
 CHAPTER TWO:
LITERATURE REVIEW

2.1 INTRODUCTION

As outlined in Chapter One, the difference in global opinions on fair value

accounting for all financial instruments is the core of this study and raises two crucial

issues in this debate. The first issue relates to the appropriateness of a proposed fair

value accounting model for all financial instruments and the level of alignment with

JWGSS and JWGBA viewpoints. The second issue concerns the usefulness of

accounting information resulting from a proposed fair value accounting model. In

relation to both issues, factors that could possibly explain their views are explored.

The following sections in this chapter review the relevant literature to October 2004.

This includes accounting for financial instruments, developments on the accounting

standard for the recognition and measurement of financial instruments, literature on

user information needs and a review of the qualitative characteristics that accounting

information should possess outlined in the IASB Framework.

2.2 THE CONTROVERSY

The official position of the international financial institutions community (as

represented by the JWGBA) is in strong disagreement with the JWGSS proposal for

a fair value accounting model. The latter calls for all financial instruments to be

measured at fair values in the primary financial statements. The two contrasting

views as to the most appropriate way of measuring financial instruments are: the fair

value accounting model (proposed by the JWGSS) and the preparers’ current mixed

16
measurement model used by many financial institutions. There have been a myriad

of developments and things have changed since this clash.

2.2.1 The Standard Setters’ View

The Joint Working Group of Standard Setters (JWGSS) was a partnership of

standard setters established in 1997 to develop a proposed comprehensive standard

on accounting for financial instruments. This group included powerful major

accounting standard setters (US, UK, Canada, Australia, Germany, France, Japan,

New Zealand, five Nordic countries and the (then) International Accounting

Standards Committee). They favoured the adoption of a fair value accounting model

for all financial instruments for inclusion in general purpose financial statements.

The JWGSS argued that the fair value approach provides more relevant information

for investors and creditors as fair values better reflect the effects of current economic

conditions and are thus better predictions of future conditions (see Willis (1998) for a

summary of arguments for fair value). In addition, they asserted that the widely used

mixed measurement model allows for too much subjectivity and opportunities to

manage earnings (JWGSS, 1999). Furthermore, the JWGSS (1999) claimed that

some banks do fair value financial instruments in the banking book for internal

management purposes. These assertions have little past empirical support.

The accounting system currently used in financial institutions utilises a mixed

measurement basis. Under this system some financial instruments are measured at

fair value while others at cost or amortised costs based essentially on management

intent. This approach is currently being applied in many jurisdictions, including

Australia and Singapore. The JWGSS argued that the mixed measurement model has

significant deficiencies and is not sustainable in the future. Furthermore, the

17
proposed fair value accounting system is argued to be superior in relevance and

therefore, also superior in the usefulness of the information that can be derived from

it. However, the JWGSS added that the effective implementation requires the

integration of certain finance and capital markets concepts and practices with

financial accounting objectives and the IASB Framework. This essentially calls for a

set of different skills from that required by the traditional historical-cost-based

accounting (Stevenson, 2004b).

Although the proposed fair value accounting model may seem to contradict

conventional accounting for non-financial assets and liabilities, the JWGSS is of the

opinion that there exist real differences between financial activities and non-financial

operating activities that ultimately gives rise to different but valid accounting

considerations (JWGSS, 1999). Furthermore, Stevenson (2004a) spoke of the need

to reexamine the thought process behind the conceptual framework due to lessons

learnt from the accounting for financial instruments.

Financial instruments represent contractual rights or obligations while non-financial

assets and liabilities have non-contractual relationships to future cash flows and,

therefore, are at an earlier stage in the development of economic returns. In addition,

non-financial assets are inputs to a productive process and its value depends on how

effectively it is used in the revenue-generating process. Under traditional

accounting, these assets are usually carried at historic cost (or lower of cost and

market) until the point that they are used up or are realised into revenues consisting

of cash or claims to cash (i.e. when they become financial assets) (IASC, 1997). The

IASC Discussion Paper (1997) asserted that value of financial assets does not depend

on this realisation process but is determined by its contractual rights and, thus, the

accounting for these assets should reflect these contractual rights and benefits.

18
To test the appropriateness of fair value accounting, surveys are sent to chief

financial officers and sophisticated users in Australia and Singapore to gather actual

views on the proposed fair value accounting model.

2.2.2 The Bankers’ View

In contrast, the Joint Working Group of Banking Associations (JWGBA) disputes the

application of fair value accounting to financial instruments in the banking book.

They argue that the fundamental characteristics of the banking book are substantially

different and thus warrant the use of a different accounting measurement attribute.

This purported crucial distinction is addressed in this study in both the preparer and

user surveys to determine their perceptions in relation to the difference between the

trading and banking books.

Banking activities consists of raising funds and investing them in assets in order to

make a profit from the margin between the amount received on interest bearing

assets and the amount paid on interest bearing liabilities. In addition, the long-term

customer relationship entered into by the bank is considered central to this function

(JWGBA, 1999a). Following this, financial institutions argue that they are not

concerned with the current value of such financial instruments, as it does not reflect

the nature of the banking transactions. According to the JWGBA (1999a), the

proposed fair value accounting model would fail to represent the substance behind

banking activities as these transactions are part of the long-term customer

relationship under which income accrues over time and there is no intention to

dispose of the financial instruments in question. The JWGBA (1999b) further

stresses that fair value is not used in the management of the banking book because it

lacks economic relevance and objectivity. The subjectivity and difficulty in

19
determining fair values of instruments not traded on established markets is also

acknowledged by Chalmers and Godfrey (2000).

Trading or non-banking activities are transactions undertaken with the objective to

profit from short-term fluctuations in market prices (JWGBA, 1999a). Here, active

decisions are taken to hold or dispose of financial instruments making the use of fair

value accounting more appropriate in their view, as it better represents these trading

transactions and management performance. Financial institutions thus argue that the

use of fair value accounting for all financial instruments will undermine this

fundamental difference in banking and trading activities.

The JWGBA (1999b) argued that the JWGSS incorrectly assumed fair value

accounting as the most relevant measurement basis for all financial instruments

without compelling evidence of preparer and user support. They asserted that “a full

fair value system for the measurement of financial instruments in the primary

financial statements of the banking industry would not be an improvement on current

practice…” and “where such information is provided in the notes to the financial

statements, it is not perceived by users as possessing attributes that contribute greatly

to the predictive process” (JWGBA, 1999a, p. 11). Furthermore, the JWGBA stated

that there “is no evidence that users such as investors and practicing analysts would

prefer fair value accounting” (p. 7) and pointed out that users have not complained

about the current mixed measurement approach. In the bankers’ opinion, the mixed

measurement model best reflect the underlying economic substance of banking

activities. The banking community also argued that a bank’s financial instruments

earn revenue from both trading and traditional banking activities and a fair value

accounting model undermines the fundamental difference between the two activities.

20
Although this is the official view, most of the banking associations represented by

the JWGBA do not include all the financial institutions in their respective countries.

For example, although the Australian Banking Association (ABA) represents banks

in Australia who are their members, views of non-members, usually the smaller

financial institutions in Australia are not represented. Therefore, in this study,

preparer surveys are sent to chief financial officers in financial institutions, both

small and big to provide evidence on preparers’ support for the proposed fair value

accounting model as compared to the JWGBA’s official views.

2.2.3 Key Themes Raised in the Controversy

From the discussion in Sections 2.2.1 and 2.2.2, four important issues can be

identified from this controversy between these two global groups. At the centre of it

all is the purported preference (and opposition) toward fair value accounting for all

financial instruments and the very specific dispute over the distinction between the

trading and banking books. The banking association stresses the fundamental

difference between the two but the JWGSS disagrees. A third important issue

concerns the relevance and reliability (otherwise referred to as qualitative

characteristics possessed by useful information by the IASB Framework) of

information resulting from fair value accounting for all financial instruments. The

last, but not least, crucial issue is the concern over volatility in reported earnings due

to fair value accounting. The banks argue this is problematic and can cause

confusion and uncertainty among investors in their decision making (JWGBA,

1999a).

The distinction between the trading and banking book is the first key theme to be

examined in this research. Views on the relevance and reliability of accounting

21
information resulting from the use of fair value accounting for all financial

instruments and the supposedly problematic volatility in reported earnings make up

the other two key themes. In addition, overall preference for fair value accounting

makes up the final all inclusive key theme of this study. Answers to these four key

themes serve to address the research questions postulated in this study.

2.3 ACCOUNTING FOR FINANCIAL INSTRUMENTS

The increasingly widespread use of financial instruments and the dynamic nature of

financial markets have resulted in the need for a comprehensive standard on

accounting for such instruments (Hernandez, 2003). However, the accounting for

and reporting of financial instruments is an area of accounting, for which existing

guidance in countries like Australia is inadequate (Alfredson, 2000). This lack of a

clear industry standard has a significant impact on the relevance of financial reports

as the use of financial instruments has increased exponentially in recent times.

International (and domestic) experience shows that the issues related to the

recognition and measurement of financial instruments are highly controversial

(Fargher, 2001; Alfredson, 2000).

2.3.1 Definitions and Background

Financial instruments are defined as contracts that give rise to both a financial asset

of one entity and a financial liability or equity instrument of another entity in IAS 39:

Financial Instruments: Recognition and Measurement (IASB, 2003b, para. 9). A

financial asset is any asset that is cash, an equity instrument of another entity, a

contractual right to receive cash or another financial asset from another entity, a

contractual right to exchange financial instruments with another entity under

22
conditions that are potentially favourable or an equity instrument of another entity.

Conversely, a financial liability is any liability that is a contractual obligation to

deliver cash or another financial asset to another entity or to exchange financial

instruments with another entity under conditions that are potentially unfavourable

while an equity instrument is any contract that evidences a residual interest in the

assets of an entity after deducting all of its liabilities (IAS 32, para. 11). Although

the definitions are recursive, they are not circular and the chain of contractual rights

or obligations established must ultimately end with the receipt or payment of cash or

with the acquisition or issuance of an equity instrument.

According to the IASC (1997) discussion paper, all financial instruments are defined

by contracts or the rights or obligations derived from the contractual provisions that

underlie them. Usually, such agreements involve two or more parties that have clear

economic consequences that the parties have little or no discretion to avoid as the

agreement is enforceable at law. Following this, financial instruments encompass a

broad range of assets and liabilities including both primary (e.g. cash, receivables,

debt, and equity securities) and derivative (e.g. options, swaps and futures) financial

instruments.

A derivative financial instrument is a financial contract whose value changes in

response to the change in an underlying financial asset such as shares, property,

foreign currency and other tangible and intangible assets (IASB, 2003b, para. 9). It

provides the holder of such instruments with the right (or obligation) to receive (or

pay) cash or another financial instrument in an amount determined by the price

changes in the underlying asset.

23
A particular financial instrument is defined by its risk exposures and reward

possibilities and thus, the recognition and measurement of financial instruments

needs to begin with an understanding of financial risks. Risk can be loosely defined

as the possibility of adverse events (IASC, 1997). In the area of finance, risk is

commonly analysed in terms of the variability of outcomes or earnings. As a general

rule, the undertaking of higher risk gives the potential to earn higher earnings (i.e.

income above the risk-free rate). Financial instruments will be exposed to one or

more ‘financial risks’ of the five definitions of financial risk: (1) currency risk, (2)

cash flow interest rate risk, (3) market risk, (4) credit risk and (5) liquidity risk

(IASC, 1997).

Currency risk refers to the risk that the value of a financial instrument fluctuates with

changes in foreign exchange rates. The risk that the value of a financial instrument is

affected by changes in interest rates is interest rate risk. When the value of a

financial instrument is affected by factors specific to the particular instruments or by

general market conditions, it is exposed to market risk. In most cases, this results

from changes in current market price and economic conditions. Credit risk is the risk

to one party to a financial instrument when the other party fails to discharge their

contractual obligation under the instrument. Finally, liquidity risk is the risk that

results in a loss because a position cannot be eliminated quickly. This may result

from an inability to sell a financial asset quickly so as to realise its fair value or the

inability to raise funds quickly enough to meet commitments of a financial

instrument (IASC, 1997).

These five risk exposures gives rise to ‘fair value exposure’ in terms of current

financial asset and liability positions as well as ‘cash flow exposure’ in the form of

changes in expected future cash flows. The financial risks mentioned above are

24
reflected in the fair values of the financial instruments that are exposed to them. For

example, the fair value of a fixed-rate financial asset reflects the current interest rates

and issuer’s credit quality and this fair value will change with any changes in interest

rates or credit quality. Therefore, fair value exposure refers to the exposure of the

fair value of financial instruments to changes in the financial risks that underlie the

contractual rights or obligations. Exposure to variations in future cash flows as a

result of possible changes in the financial risks is termed cash flow exposure. An

example is an anticipated purchase of Malaysian Ringgit in the future, which will be

affected by a change in the price the currency. The significance of this exposure in

accounting has to do with the appropriate time period for the recognition of such

exposures (IASC, 1997).

Consequently, fair value accounting for all financial instruments, in theory, seems

more relevant in that the risks contained in financial instruments are best7 reflected in

the fair value of the instruments at a point in time. This thesis examines user and

preparer preference for fair value accounting and their perceptions of the usefulness

of the accounting information resulting from its use.

2.3.2 Current Accounting Practice

Currently, financial institutions in many countries use a mixed measurement model

for financial instruments based on the classification of those instruments into two

categories: the banking book and the trading book (JWGBA, 1999a). Those

instruments grouped into the banking book are generally measured at cost or

amortised cost while those in the trading book are normally fair valued. Questions

arise as to the appropriateness of grouping certain assets and liabilities (namely

7
This assumes that the measure of fair value is reliable and from an active secondary market.

25
commodities held as inventories, physical assets held as investments, taxes,

contractual commitments to buy or sell non-financial assets, pensions and insurance

obligations, equity instruments and the crucial debt or equity classification) into

different groups of financial instruments for accounting purposes. The JWGBA

(1999a) believes the current mixed measurement model (cost or amortised cost in the

banking book and fair value accounting in the trading book) is most appropriate and

best established, especially for financial instruments in the banking book. Their

Singaporean counterparts take the same stance. However, The JWGSS (1999) are of

the opinion that the current mixed measurement model impedes comparability and

allows far too much latitude by financial statement preparers to manage earnings.

2.3.3 The Proposed Fair Value Accounting Model

Advances in financial risk management and information technology together with

globalisation of capital markets around the world resulted in the accelerated use of

sophisticated financial instruments. As a result of this fundamental shift in the

business and investment environment, traditional accounting concepts need to be

rethought especially in the case of financial instruments (JWGSS, 2000). Stevenson

(2004b) extended this by asserting that the developments in accounting for financial

instruments are potentially driving a “paradigm shift” in the conceptual framework

towards fair value accounting. This is also echoed by the European Central Bank

(ECB) who see the move to fair value accounting as a “paradigm shift since

backward-looking accounting measures based on the concepts of prudence and

reliability to measures based on prevailing economic values” (ECB, 2004b, p. 79).

In recognition of the world-wide importance of accounting for financial instruments,

accounting standard setting bodies around the world initially responded by requiring

26
more extensive disclosures about financial instruments (for example IAS 32, SFAS

105, 109). This was followed by standards for recognition and measurement of

financial instruments that utilised a mixed measurement approach (for example IAS

39, SFAS 133). However, none have proposed such a radical shift from traditional

accounting concepts as the JWGSS proposal for the adoption of a fair value

accounting model for all financial instruments consistent with accepted capital

markets practices and finance concepts for pricing these instruments.

An Invitation to Comment on the JWGSS Draft Standard "Accounting for Financial

Instruments and Similar Items" was released in December 2000. The main proposals

in the Draft Standard were, broadly, virtually all financial instruments (including

hedging instruments) should be measured at fair value at each reporting date and

virtually all changes in fair value should be recognised as income or expenses in the

Income Statement in the period in which they arise. Financial instruments for which

there exist accounting standards (for example investments in subsidiaries) need not

be measured at fair value. The single exception to the rule of taking changes in fair

value to the Income Statement is exchange translation gains and losses relating to

certain foreign operations (which are to be separately presented in accordance with

existing foreign currency translation standards). The draft standard also calls for

expanded disclosures on financial instruments, financial risk positions (information

about gains and losses by general classes of financial risks) and Income Statement

effects sufficient to enable the evaluation of risk positions and performance in

respect of each of an entity’s significant financial risks. Required disclosures include

a description of each significant financial risk and the objectives and policies for

managing those risks, information about the risk positions and financial performance

27
effects for each significant risk and the methods and key assumptions used to

estimate the fair value of financial instruments.

The JWGSS (2000) argued that fair value is the most relevant measurement attribute

for all financial instruments and holds that sufficiently reliable estimates of fair

values of financial instruments are obtainable for financial reporting purposes. The

draft standard set out principles for estimating the fair value of financial instruments

within a hierarchy. As shown in Figure 2.1, the JWGSS believed that reliable fair

value estimates could be derived using valuation techniques as developed in the

finance disciplines.

Figure 2.1: Hierarchical principles for estimating fair value of financial

instruments

Observable market exit prices for identical instruments

Available Not Available

- Use those prices

Observable market exit prices for a similar instrument

Available Not Available

- Use those prices

- Fair value estimated using a valuation

technique consistent with accepted
economic pricing methodologies (with
adequate disclosures, of course)

Adapted from JWGSS (2000, p.46).

28
The Draft Standard (IASC, 1997) also addresses circumstances requiring special

consideration in using observed market prices to determine fair value. Such

circumstances include:

ƒ observable market prices may not be determined by normal market interactions,

ƒ infrequent availability of market prices,

ƒ prices for a financial instrument is available in more that one market,

ƒ applicability of prices of small blocks of financial instruments to large block

holdings, and

ƒ where the market price includes a value that is not directly attributable to the

financial instrument.

Basic standards for selecting valuation techniques (present value concepts are central

to the development of valuation models) and for the use of estimates and

assumptions are also included. Most importantly, the JWGSS stresses that the

establishment of fair value measurement policies and procedures that are appropriate

to an entity’s financial activities is the foundation for ensuring reliable and internally

consistent fair value estimates and assumptions.

The stance taken in the both the discussion paper and draft standard is strongly

disputed by the JWGBA (1999a; 1999b). The surveys developed in this study gather

evidence on the actual perceptions of both preparers and users on the relevance and

usefulness of fair value accounting for financial instruments, which inherently leads

to the preference (or opposition) for the fair value accounting proposal.

29
2.3.4 Current Developments

Since the draft standard (in December, 2000) various developments in relation to

accounting for financial instruments have occurred both nationally and

internationally.

2.3.4.1 International Accounting Developments

In the international arena, IAS 39 was developed by the IASB in 1999 to require the

use of fair values subsequent to initial recognition, for certain financial instruments,

including most derivatives. This initial standard was very much based on the SFAS

133 (see Pacter, 1999). Subsequently, the JWGSS proposed the use of fair value

accounting for all financial instruments in December of 2000. The proposal for the

use of fair value accounting for all financial instruments in both the banking and

trading books was met with criticisms from the JWGBA (1999a; 1999b) and many

other financial institutions around the world (as per the comment letters received in

response to the 1997 IASC Discussion Paper).

More than 280 comment letters in response to the Draft Standard were received by

the IASB; two-thirds were from preparers and the balance from various other

constituents including accounting firms and professional bodies, regulators,

academics and members of the financial analyst community. According to the IASB

(2002) most of the preparers strongly oppose the use of fair value measurement for

all financial instruments and the presentation of the resulting gains and losses to the

Income Statement. One fourth of the respondents support full fair value accounting

on the condition that the perceived problems with reliability and volatility of

earnings can be resolved. Stevenson (2004a) observed that support (although most

had practical concerns) for fair value accounting comes from preparers who are

30
already using fair value accounting for financial reporting purposes (such as Dutch

banks and the Reserve Bank of Australia) whereas opposition originates from those

who are not using fair value accounting. However, many respondents commented

that the extensive coverage of the standard is overwhelming and difficult to fully

comprehend and address in depth. The main source of conceptual opposition is that

fair value measurement does not accurately reflect management intention and

internal management practices and thus, is irrelevant for financial instruments in the

banking book. There were also calls for rigourous field-testing and education,

especially in relation to practical concerns on the reliability of fair value

measurement and the resultant volatility in earnings (Bradbury, 2002; Hague, 2002).

The draft standard was criticised and deemed prematurely finalised (ECB, 2004b).

An executive of France’s second largest bank says that there will be banks “where no

one will understand the accounts…” and “… management are not going to be able to

manage the business” (Parker and Pretzlik, 2004, p. 17). The ECB also perceived

that fair value accounting could exacerbate boom and bust economic cycles.

Subsequently, the opposition in Europe (especially the concerns expressed by the

ECB) to various aspects of IAS 39 drove the IASB to issue a number of EDs

proposing a multitude of amendments to IAS 39 eight months later, in August 2001.

Following those amendments, in June 2002, the IASB issued an Exposure Draft with

proposed revisions to IAS 32 and IAS 39 and subsequently issued a revised version

of both standards in December 2003. The European Financial Reporting Advisory

Group (EFRAG) voiced support (and understanding) for the attempts by the IASB to

accommodate the concerns raised by prudential supervisors, regulators and other

various parties while maintaining the thrust of the intention of the original standard

(EFRAG, 2004).

31
More than 170 comment letters were received and the IASB then extended the

consultation process and conducted (in March 2003) a series of nine roundtable

discussions in which over 100 organisations and individuals took part (IASB, 2003a).

One of the main issues raised in the comment letters and roundtable discussions is

the concern that fair value hedge accounting for a portfolio hedge of interest rate risk

(referred to as macro hedging) is very difficult to achieve under IAS 39

requirements.

Another revision introduced in the revised IAS 39 (December 2003) was the option

that permits entities to designate (irrevocably) any financial asset (or liability) as one

to be measured at fair value with gains and losses recognized in the Income

Statement (termed the ‘fair value option’) (IASB, 2004c). The option8, according to

the IASB (2004c) was to simplify the practical application of IAS 39 (without the

option) especially in situations where the mixed measurement model could result in

an entity reporting volatility on positions that are economically matched.

Among the respondents whose comment letters discussed the fair value option, the

majority agreed with it. However, some prudential supervisors of banks, securities

companies and insurers expressed concerns that this option might be used

inappropriately and might result in increased (and unnecessary) volatility in earnings

that could ultimately lead to the demise of the entities (Parker and Pretzlik, 2004;

IASB, 2004c, Stevenson 2004a).

8
Some observers saw this move as a political response (Stevenson, 2004a).

32
The Basle Committee on Banking Supervision (2004) also voiced concern that such

an option may reduce the comparability of financial statements. Specifically, these

constituents are concerned that:

ƒ Entities apply the fair value option to financial assets or liabilities whose fair
value is not verifiable, thus allowing entities to determine their fair values in a
way that inappropriately affects profit or loss.

ƒ It might result in an entity recognising gains or losses in the Income Statement

for changes in its own creditworthiness if it applied the fair value option to
financial liabilities (see ECB, 2004a; EFRAG 2004).

Therefore, the IASB issued an exposure draft in April 2004 proposing that the fair

value option be restricted in its use by limiting the types of financial assets and

liabilities to which the option may be applied to five specified categories of: (1)

financial assets and liabilities that contain embedded derivatives, (2) financial

liabilities whose cash flows are contractually linked to the performance of assets that

are measured at fair value, (3) cases where the exposure to changes in the fair values

of a financial instrument are substantially offset by the exposure to the changes in the

fair value of another financial instrument, (4) financial assets other than loans and

receivables, and (5) items that other Standards allow or require to be designated at

fair value through profit or loss with the added condition that the option can only be

applied to financial instruments whose fair value is verifiable. With the exception of

regulators, the majority of the 115 comment letters received did not agree with this

restriction on the fair value option (IASB, 2004c).

Although many developments have taken place, the controversy about the use of fair

value accounting for all financial instruments still resonates. The recent

developments in respect of the use of fair value for all financial instruments reinforce

the need for the type of research conducted in this thesis.

33
2.3.4.2 Australian Changes

Australia has always been highly independent in its own accounting standard-setting

as well as actively participating in the international standard-setting arena (see

Haswell and McKinnon, 2003; Howieson and Langfield-Smith, 2003; Collett,

Godfrey and Hrasky, 2001). More recently, the Financial Reporting Council9 (FRC),

in a major initiative, announced the adoption of International Accounting Standards

(from 1 January 2002 to be known as International Financial Reporting Standards

(IFRS)) by Australia for financial periods beginning on or after 1 January 2005

(FRC, 2002). After this date, Australian companies need to fully comply with IFRSs

and their audit reports will attest to this compliance (FRC, 2002). Consequently, the

Australian Accounting Standards Board (AASB) is obligated to work towards the

full implementation of (and convergence to) IFRSs (Alfredson, 2002). This de jure

convergence was effectively accomplished in July 2004 when the Australian

equivalents of IFRSs were released.

Ravlic (2002) argued that the proposal was bewildering and experts in financial

reporting are only beginning to comprehend the consequences of this transition and

some question the viability of the deadline. The implication of this decision is that

IAS 39 will be fully applicable in Australia for reporting periods ending on or after 1

January 2005.

2.3.4.3 Singaporean Situation

Singaporean accounting standard setting has been much less independent than

Australia’s. The norm in Singapore is to adapt IFRSs to local Singaporean

9
The FRC is established under the Australian Securities and Investments Commission Act 2001.
One of its key functions is to advise the Australian Government on the accounting standard setting
process and the development of international accounting standards.

34
conditions. In the second half of 2000, Singaporean Accounting Standard SAS 33

Financial Instruments: Recognition and Measurement, an accounting standard similar

to IAS 39 was issued to be effective for periods beginning 1 July 2001 for entities

employing International Accounting Standards in their reporting framework (Chua,

2002).

Subsequently, in 2002, the Singaporean government replaced the national accounting

standard setter, the Institute of Certified Public Accountants of Singapore (ICPAS)

with the Council of Corporate Disclosure and Governance (CCDG). In response to

international developments towards a single set of accounting standards, the CCGD

has issued a number of Financial Reporting Standards (FRS) and Interpretations of

the Financial Reporting Standards (INT FRS) that are almost identical to the IFRSs.

Previously, the SASs were already closely aligned to IFRSs, the changes necessary

to bring all Singaporean accounting standards in line with the new IFRSs were

minimal (Deloitte Touche Tohmatsu, 2003). This further reinforces the high

alignment between Singaporean accounting standards with the IFRSs.

2.4 THE CASE OF FAIR VALUE ACCOUNTING

There are a myriad of empirical studies on fair values and fair value accounting. The

majority of such studies are based on capital market principles which adopt a

different approach to that adopted in this study. These studies form the basis of

arguments both for and against the use of fair values. The main capital market

studies on fair values are summarised in the following paragraphs.

Over the last decade, there were numerous papers investigating the empirical

relationship between share prices (or changes in share prices) and particular

35
accounting numbers for the purposes of assessing the relevance of an accounting

standard (Holthausen and Watts, 2001). This type of research is also referred to as

value relevance literature which assumes that high associations with share values are

the desirable attribute for accounting earnings (Holthausen and Watts, 2001). The

following are some examples of the fair value accounting literature.

Ahmed and Takeda (1995) studied a sample of 152 bank holding companies listed on

the American stock exchanges and concluded that unrealised gains and losses had a

significant positive effect on bank stock returns. In the Barth, et. al. (1995) study,

fair value based earnings was found to be more volatile but share prices did not

reflect that incremental volatility. Barth, et. al. (1995) also argued that banks

violated regulatory capital requirements more frequently under fair value regimes

and subsequently assisted the prediction of regulatory capital violations.

Barth (1994) also established that fair value estimates of investment securities in

American banks provide significant explanatory power beyond those provided by

historical costs. More significantly, in her study, historical cost was concluded to

have no significant explanatory power incremental to fair values. It was also posited

that fair values had less measurement error as compared to historical costs vis-à-vis

the amount reflected in share prices resulting in the conclusion that fair values are

more value relevant than historical costs.

In addition, Barth, Beaver and Stinson (1991), Barth, Beaver and Landsman (1996),

Beaver, Eger, Ryan and Wolfson (1989) and Eccher, Ramesh and Thiagarajan (1996)

all concluded that disclosures on fair values (and related risks) provide incremental

explanatory powers and are relevant to the valuation of share prices. Barth and

36
Landsman (1995) also concluded that a fair value accounting-based Balance Sheet

reflects all value-relevant information.

However, there are also studies that revealed evidence against the value-relevance of

fair values. Barth, Beaver and Wolfson (1990) examined two components of the

earnings of 150 banks: income before realised gains and the amounts of realised

gains and losses. This study discovered that realised securities gains and losses

behave in an earnings-smoothing manner and concluded that investors apparently

believe that reported gains and losses are timed by management in order to offset

gains (or losses) elsewhere.

Interestingly, the final conclusion of Barth (1994) questioned the appropriateness of

including less reliable fair value estimates in earnings while Eccher, et. al. (1996)

concluded that their results suggested that historical costs provide more value

relevant information as compared to fair value disclosures in both an absolute and

incremental sense.

These contradictory study findings add to the controversy surrounding the usefulness

of the proposed fair value accounting model. The survey instruments developed in

this study gathers actual information on the perceptions of the relevance and

usefulness of fair value accounting for all financial instruments.

The above studies are but a small sample of the value-relevance literature. These

studies focus on share price changes to determine the relevance of fair value

accounting and disclosures. In contrast, this research project does not use this

approach but examines the issue of fair value accounting from a ‘usefulness of

information’ perspective and adopts a different methodology from traditional capital

37
market research. Moreover, Holthausen and Watts (2001) question some of the

assumptions that form the fundamental foundation of this group of studies.

Specifically, most of the value-relevance models assume away the existence of

economic rent and growth options. The underlying theory used in this literature is

not capable of explaining and predicting valuation, accounting or standard setting

(Holthausen and Watts, 2001). In addition, Francis and Schipper (1999) suggested

different and new approaches to increase the understanding of the usefulness of

financial statements to investors. It is crucial to note that this study approaches the

question on the level of preference for fair value accounting for all financial

instruments and its usefulness from actual stakeholders’ point of view.

2.5 SURVEYS ON FAIR VALUE ACCOUNTING

Both the JWGSS (2000) and JWGBA (1999b) alluded to various surveys in their

deliberations of fair value accounting. The following paragraphs highlight the main

findings in four such surveys.

A survey of financial analysts and other users was conducted by Sirota Consulting

(an independent consultant) on behalf of the Association for Investment Management

and Research10 (AIMR), FASB and the Canadian Institute of Chartered Accountants

(CICA). Results of the survey are mixed, consistent with the concerns that many

users struggle with the complexity of fair value accounting. Even though the survey

targeted users who are “knowledgeable and informed about fair value accounting for

financial instruments”, respondents were evenly divided on whether to measure

financial instruments at fair value or cost. The only conclusive finding of this study

10
The AIMR supports fair value measurement of financial instruments (AIMR, 2001).

38
was the respondents’ advocating the need for more and better information regarding

fair value accounting.

The American Institute of Certified Public Accountants (AICPA) (1994) conducted a

comprehensive study on 107 professional investors and creditors and their advisors

(e.g. analysts, brokers and accountants) via various formal and informal interviews

and questionnaires between October 1992 through April 1993 to ascertain the

information needs of users to identify the types of information most useful in

predicting earnings for decision making purposes. The study finds that users do not

want fair value measurement, preferring the current mixed measurement model due

to the stability and consistent benchmark that it provides (JWGBA, 1999b). Users

also oppose fair value accounting because it is not relevant to how companies are

evaluated, how earnings or cash flows are predicted and it would introduce

unacceptable level of volatility which is not useful in making decisions (AICPA,

1994). The study also finds that users are of the opinion that fair values should only

be used in footnotes or in supplementary disclosures, not at primary measurement

because “users are willing to accept less reliability in the context of supplementary

disclosures than in the context of measurement in the Balance Sheet or the Income

Statement” (AICPA, 1994, p. 95). Respondents also believe that external financial

reporting should be better aligned with internal management reporting. The study

acknowledges the considerable body of research on the effects of financial

information and its impact in capital markets but states that the research did not

provide sufficient knowledge about users’ information needs and that “high-quality

research on users’ needs for information has been limited” (AICPA, 1994, p. 114). It

also suggests that users believe that they can handle differences in accounting among

companies, making comparability less important than the JWGSS asserts. Finally,

39
the study calls for increased focus on actual user needs rather than speculative ideas

not backed up by empirical data or direct observations.

Another relevant study on users’ information needs in relation to bank holding

company annual reports for year 1992 was conducted by KPMG. The study finds no

case for fair value disclosures with “… a significant drop-off in views towards the

usefulness of actual 1992 fair value disclosures when compared with the anticipated

usefulness…” (KPMG, 1993, p. 16). As much as 82% of users said that nothing new

had been learnt from fair value disclosures and 88% of users believe that the

historical cost accounting model should be kept. In addition, approximately 92%

believe that adjustments to fair values in the financial statements would not provide a

more accurate presentation of an institution’s financial position and results than

financial statements based on historical cost. A majority of the users also believe

that fair value accounting is not appropriate for assets held for the foreseeable future

and will not provide a more accurate measure of a financial institution’s capital.

In addition, the KPMG also conducted a study for the Association of Reserve City

Bankers (ARCB) testing the opinions of institutions towards the use of fair value for

financial instruments, also in 1992. 70% of preparers felt that financial statements

adjusted to reflect fair value is not useful and 90% replied that fair value accounting

should not be the measurement basis for an institution’s financial statements,

stressing that fair value is too volatile and judgmental and lacked usefulness and

accuracy (JWGBA, 1999b). Overall, respondents opposed replacing historical cost

with fair value accounting due to concerns on its unreliability and inconsistency.

These mixed opinions between users and preparers found in these decade-old studies

further confirm the importance of this thesis. Evidence collected will provide

40
necessary insights into actual user information needs in terms of accounting for

financial instruments and will serve as a check on the JWGBA (and JWGSS)

assertions on fair value accounting for all financial instruments.

2.6 RECENT STUDIES

2.6.1 Fargher (2001) on an AFMA (2000) Survey

Fargher (2001) analysed data from an Australian Financial Markets Association

(AFMA) survey (2000) and commented that, contrary to official JWGBA views,

“The evidence suggests that managers at financial institutions are not unanimously

opposed to the use of fair-value accounting for all financial instruments, as might be

suggested by positions taken by bank trade associations” (p. 70).

2.6.1.1 Background to Fargher (2001)

Evidence in Fargher (2001) was derived from Australian financial institutions, large

corporations and treasuries who are members of AFMA, even though the survey used

was not designed for the purpose of his research11 and this limitation was recognised.

However, the results provide responses from participants in the over-the-counter

market who are also big players in the area of accounting for financial instruments.

The two hypotheses tested were support for fair value accounting is weaker for

managers who perceive that: (1) fair-value accounting will result in transient changes

in reported earnings not related to underlying economic activity, and (2) there is a

higher proportion of assets for which a reliable fair value cannot be determined.

11
The survey was conducted by the AFMA to gather information on current accounting practices of
its members involved in over-the-counter trading. It was not specifically conducted by Fargher
(2001) to address the issues discussed in his paper.

41
The AFMA survey was sent to 117 members with a 54% response rate. Respondents

include the four largest domestic banks, a significant number of second tier

Australian financial institutions, corporate treasuries, power companies and

government treasury agencies. It was likely that any response bias is towards larger

and more sophisticated organisations.

2.6.1.2 Fargher (2001) Results

Fargher (2001) found significant support for fair value accounting12 and that the

majority of respondents perceived that reliable fair values could be determined for

(approximately) 90% of financial instruments contrary to the assertions of the

JWGBA. 44% of respondents supported the use of fair value accounting for all

financial instruments. In addition, he noted the respondents’ views that volatility in

reported earning caused only transient problems.

Subsequently, he tested for differences in characteristics between organisations that

supported or opposed fair value accounting, checked for correlations between the

variables used and employed logistic regressions to examine support for fair value

accounting and perceptions regarding consequences of fair value accounting.

Organisations supporting the use of fair value accounting tend to be larger. Smaller

firms are likely to face higher costs in implementing fair value accounting, thus are

more likely to oppose the practice. His results consistently support the proposition

that preference for fair value accounting is negatively associated with the perception

that fair value accounting will increase the volatility of earnings.

12
The actual AFMA survey question was “Do you support the concept of marking all financial assets
and liabilities, including derivatives and non-trading transactions in banking books to fair value on
the balance sheet?” It does not include the more controversial second requirement of fair value
accounting which calls for the taking of changes in fair values to the profit and loss statement.

42
Fargher (2001) also found support for fair value accounting to be negatively

associated with the level of assets for which fair value estimates cannot be reliably

estimated. Respondents who advocated fair value accounting argued that it ensured

that reported profits adequately reflect changes in economic conditions and improved

the relevance of the information in the accounts.

Respondents of the AFMA survey who opposed the use of fair value accounting

tended to agree that the resulting increased volatility in reported earnings would

result in transient changes in reported earnings, create increased difficulty in

projecting and meeting earnings forecasts and were likely to cause problems due to

the reaction of unsophisticated investors. This sub-group also asserted that fair

values of financial instruments could not be determined reliably and that fair value

accounting is not appropriate for financial assets and financial liabilities intended to

be held to maturity.

In summary, Fargher (2001) evidence suggests that AFMA members are not

unanimously opposed to the use of fair value accounting as might be suggested by

the positions taken by bank trade associations and finds no strong evidence that the

support for fair value accounting increases with the perception that reliable fair

values can be estimated.

2.6.2 Bradbury’s Reliability Study (2001)

A study was conducted by Bradbury (2001) to gather views on the alleged problems

with obtaining reliable fair value estimates. Semi-structured interviews were

conducted with New Zealand firms to define the circumstances where fair values

may not be sufficiently reliable for financial statement measurement and to explore

43
reasons for such circumstances. Bradbury (2001) picked New Zealand due to their

extensive experience in fair value disclosures under New Zealand’s Financial

Reporting Standard, FRS 31 Disclosure of Information about Financial Instruments.

Six entities from industries (including government bodies, insurance, financial

services, the energy sector and a diversified multinational company) participated in

the interview. The turnover for the six interviewed organisations averaged NZ$1,896

million with mean net income of NZ$141 million. Average total assets and equity

approximates NZ$13,547 million and NZ$3,568 million respectively. The Bradbury

(2001) interviews focused on the reliability of fair values for non-traded financial

instruments as defined as financial instruments where no market prices are readily

available to value these instruments13.

Interviewees indicated that various methods were applied in measuring the non-

traded financial instruments discussed above ranging from fair values through to

non-measurement in the Balance Sheet. In determining the reliability of such non-

traded financial instruments, Bradbury (2001) grouped them into four categories: (1)

financial asset, (2) financial liability, (3) financial derivative and (4) commodity

derivative.

On average, interviewees indicated that 36% of the time, the fair values of non-

traded financial asset, non-traded financial liability, non-traded financial derivative

and non-traded commodity derivative were not reliable. The four main reasons

postulated for the lack of reliability of financial instruments included constant

13
The focus was on trade debtors, non-traded equity investment, other financial instruments with a
pre-payment option, short-term borrowings and any other instruments with potential reliability
problems.

44
variation in value, the cost incurred to obtain fair values, irrelevance of fair values as

well as disclosure of competitive information.

Some of the interviewees also point out that independent valuation to obtain fair

values is an expensive process and the inclusion of internally generated profits

impedes the quality of profit. According to the interviewees, another big impediment

to reliability is the dynamic nature of financial markets which renders conditions

underlying the assumptions in fair value calculations obsolete as soon as the fair

values are calculated.

In conclusion, Bradbury’s (2001) interviews further confirm the various

measurement methods used for non-traded financial instruments and the lack of

guidance on the matter. Yet, his findings do suggest that 64% of respondents believe

that non-traded financial instruments can be reliably measured. Although the

interviews provide more insight into the reliability of the fair values of non-traded

financial instruments, it only collected views from six entities.

2.6.3 Links to this Project

This thesis asks the crucial questions, including those raised in both the Fargher

(2001) and Bradbury (2001) studies, concerning perceptions of the proposed fair

value accounting model for all financial instruments. Surveys are sent to both

preparer and sophisticated users in both Australia and Singapore to gather empirical

evidence from both perspectives regarding the key issues in this debate two years

after the AFMA (2000) survey. Moreover, the surveys also ask the crucial questions

on the perceived reliability of both traded and non-traded financial instruments,

including specific sub-categories.

45
2.7 USER INFORMATION NEEDS

There are varying views on information needs of user groups. Discussion on such

needs inherently starts with the fundamental aim of financial reporting – to provide

information that is useful to present and potential investors and creditors in making

rational investment and credit decisions (FASB, 1978). The following subsections

details the objectives of financial reporting, the different stated needs of various

users, qualitative characteristics of useful information and the assonance and

dissonance between information provided to and needed by users.

2.7.1 Objectives of Financial Reporting

There has always been considerable debate and controversy on the objective and

meaning of financial reporting since the publication of the Trueblood Committee

report in 1973 (Chang and Most, 1985). In October 1973, the Trueblood Committee

report stated that the objective of financial reporting is to provide information useful

for making economic decisions. Where the investor group is recognised as the major

consumer of financial statements, it then follows that a major objective of accounting

is to provide quantitative economic information that will be useful for making

economic decisions (Staubus, 1961). Similar propositions were then used by the

American Accounting Association and the Financial Accounting Standards Board

which endorsed the concept that financial statements should provide information that

is useful to present and potential investors and creditors in making rational

investment and credit decisions (FASB, 1978).

Nearly all constituents of the financial community (preparers, standard setters and

users alike) agree that the primary purpose of external financial reporting is to

“provide information to users that is useful for making and evaluating decisions

46
about the allocation of scarce resources” (Statement of Accounting Concept 2,

AASB, 2001, para 26; Pakrul, 1977). In this capacity, the American Accounting

Association (AAA) in their Statement of Basic Accounting Theory had defined

accounting as “the process of identifying measuring and communicating economic

information to permit informed judgments and decisions by users of the information”

(AAA, 1966, p. 1). The common notion encompassed by conceptual frameworks

around the world and established accounting theories is the emphasis of the

usefulness objective of accounting. Furthermore, the AAA Subcommittee of

Establishing Materiality (1976) stressed that where this notion of usefulness is not

comparable between the accountants and the users, financial reporting falls short of

its objective to serve user needs (see also Chastain, 1974).

Even back in the days of the debate about whether accounting is a trade or a science,

advocates of the latter suggested the teaching of the accounting model of

measurements to give prospective students conceptual insights into how conventional

accounting attempts to meet the objective of serving users’ needs (Burton, 1982).

In addition, Mautz (1963, p. 319) stated that:

Accounting deals with enterprises, which are certainly social groups; it

is concerned with transactions and other economic events which have a
social consequences and influence social relationships; it produces
knowledge that is useful and meaningful to human beings engaged in
activities having social implications; it is primarily mental in nature.

The development of conceptual frameworks for financial reporting that evolves

around the major aim of producing information that will satisfy the needs of a variety

of different user groups also emphasises this need for useful information (Lee, 1988).

In this capacity, FASB has consistently sought user input and to involve users in the

standard-setting process to ensure that the information required by resulting

47
standards are useful and thus, improve capital allocation decisions (Jonas and Young,

1998). Unfortunately, evidence suggests that users have yet to play a major role in

the standard-setting process, whether at the initial or the final stages of the process

(Jonas and Young, 1998). Therefore, the quality of a standard should be measured

based on the usefulness of the information required in the standard from the users’

perspective (Jonas and Young, 1998). Following this, there is a need to determine

the information that users really need and to check if this information is currently

being provided.

Even while accounting bodies and authorities around the world have accepted this

objective of financial reporting, some groups argue that financial statements are not

always useful. The majority of these groups are proponents of the Efficient Market

Hypothesis (EMH) who assert that information contained in financial statements are

impounded in the price of a security so rapidly that investors cannot make abnormal

returns in the market on the basis of this information 9benston, 1974).

Benston (1974, p. 35) goes further to say that “the value of reporting financial data

for investors’ decision depends upon its usefulness and timeliness. The SEC’s

adherence to historically based conservative accounting procedures reduces the value

of the data… this evidence [market research studies] also supports the conclusion

that the accounting statements are not useful, or timely, or both.”

If the objective of financial statements is to provide useful information for making

economic decisions, the question that follows would be: ‘Useful to whom?’ There

are many groups that use financial statements including present and potential

investors, creditors, management, government agencies, and employees.

Management has easy access to corporate information and can prescribe the form

48
and content of financial statements. Both government agencies and employees will

make decisions that impact upon an entity. However, these groups of users are only

indirectly affected as they do not directly receive the benefits or incur the sacrifices

(Chang and Most, 1985). This means that investors and creditors are the two major

groups of financial statement users. Some actually argue that the needs of these two

groups are compatible because their economic decisions are similar (Study Group on

the Objectives of Financial Statements, 1973). Rice (1973) also noted that 90% of

companies surveyed by the Financial Executives Institute gave priority in financial

reporting to existing shareholders. Hence, no other user group is regarded nearly as

important.

Therefore, as financial statements are to be useful for decision making and since

investors and creditors are the major users, financial statements should primarily

cater to the information needs of these two groups. This is reflected in the IASB’s

Framework for the preparation and presentation of financial statements, which

promulgate the objective of financial statements to be the provision of information

about the financial position, performance and changes in financial position that is

useful to a wide range of users in making economic decisions (para. 12). The IASB

framework further notes that financial statements prepared for the aforementioned

purposes meets the common needs of most user groups and provides stewardship (or

accountability) of management.

Therefore, the crucial question is: ‘What information do they need, and what are their

decision processes?’ Although agreement was reached as to who are the major user

groups and therefore where attention should be focused, the next logical step, which

is to ascertain these individuals’ information needs, has received little attention from

researchers. Research in this area must examine questions as to the needs of

49
different constituents that make up the investor group: the individual investor,

institutional investor and the financial analysts (Chang and Most, 1985). Which one

of these groups should be the ultimate target user group? The issue whether financial

statements should be prepared primarily for the average investor or the

knowledgeable professional has been widely debated. There are arguments both

ways. Some feel that financial statements should use simpler language and contain

less complex information so that individual investors can understand them without

feeling intimidated. Others argue that the exclusion of technical aspects of the

statements will render such reports useless in the deliverance of their message (see

for example Littleton, 1953). In addition, it is reasonable to expect the average

investor to seek professional advice to comprehend financial statements. Suffice to

say that the majority of writers prefer professional financial analysts as the target

user group on the basis that “most accounting information is not comprehensible to

the average investor and cannot be made so without forgoing important elements of

the accounting message” (Buzby, 1975, p. 46). Furthermore, the IASB Framework

focuses on the needs of sophisticated users (IASB, 1989).

This study further addresses the contention that information resulting from fair value

accounting is useful for the decision making purposes of users by examining the

qualitative characteristics possessed by such information. Although accounting

information is to provide information useful to various users, it is also important to

determine the actual information needs of the users of financial statements.

50
2.7.2 Different User Groups, Different User Needs

In the development of a conceptual framework for business reporting, seven groups

of users are identified: (1) the equity investor group, (2) the loan creditor group, (3)

the employee group, (4) the analyst-advisor group, (5) the business contact group, (6)

the government and (7) the general public (McMonnies, 1988; Jones, 1995). The

objective of financial reporting is to satisfy, as much as possible, the various

information needs of these user groups (Benjamin and Stanga, 1977). In general,

these user groups have interests in the following four areas14: (1) the context and

environment in which the entity operates; (2) the entity’s overall objectives; (3) the

strategy and tactics to achieve those objectives; and (4) the overall results. Corporate

reporting is concerned with the communication of information for management to

discharge their accountability to external user groups. Although management and

the government have the capacity to demand information, the other user groups who

are not directly involved in the entity are generally left to satisfy their needs with the

external reports of the entity.

From this list of potential users, three of them were identified to be most

commercially important: the equity investor group, the loan creditor group and the

analyst-advisor group (McMonnies, 1988). Members of the financial community

have called for research on the fundamental issue of information needs of financial

statement users and how they relate to the decision process (McMonnies, 1988,

Walker, 2003). Here, two key questions have to be resolved for each group: (1) Who

are these users and what are their characteristics? and (2) What are their information

needs?

14
Not all users necessarily have interests, or the same degree of interest, in all the areas.

51
The equity investor group is constituted primarily of individual and corporate

shareholders. In 1991, Epstein and Pava (1993) found that individual shareholders

are more sophisticated than they were in the 1970s and they are demanding increased

quality, quantity and information content in understandable language. Corporate

shareholders are the more sophisticated users in this group and usually make full use

of the financial report in their investment decisions. Although they have

considerable understanding of the financial report, there is still much that can be

done to improve the overall understandability. Both these users (corporate and

individual shareholders) are after information that will assist in their assessment of

risk and return.

Lenders have always played a big part in organisations and in order to protect their

investment, they need good quality and relevant information. This loan-creditor

group is concerned with a company’s ability to pay (i.e. resource and financial

structure) and they can demand information only if they are significant lenders of the

company.

Most individual investors rely on the third user group (analyst-advisor group) to help

them make the most of their investments (Lee and Tweedie, 1981). This group

comprises by far the most sophisticated users and they use data from financial reports

for further analysis and intercompany comparisons in their investment process

(Schipper, 1991; Anderson, 1988; McMonnies, 1988).

Due to the complex and ever-changing nature of the financial institutions industry,

sophisticated users of the equity-investor and analyst-advisor groups, as represented

by auditors and financial analysts (in general) are identified as the subject group to

gather evidence on user perceptions of the proposed fair value accounting model for

52
all financial instruments. Although it could be argued that the assumption that these

groups of sophisticated users have reasonable knowledge of the accounting for

financial instruments is valid, this project does not make this assumption. In fact,

this study also gathers evidence through interviews and phone surveys as to the

validity of this underlying assumption.

2.7.3 Qualitative Characteristics of Useful Information

According to the IASB “Framework for the Preparation and Presentation of Financial

Statements” (1989), the four principal qualitative characteristics of relevance,

reliability, understandability and comparability are the attributes that ensures the

usefulness of accounting information provided in financial statements. In addition,

the timeliness of information reporting was identified as the constraint on relevant

and reliable information, making it another important attribute of useful information.

These five characteristics are now explained.

2.7.3.1 Relevance

As per the IASB Framework, information must be relevant to the decision making

needs of users in order to be useful. This qualitative characteristic is said to be

possessed by information which influences the economic decisions of users by

helping them evaluate past, present or future events. Such information also has the

ability to confirm or correct past evaluations. As information regarding financial

position and past financial performance is frequently used as the basis for predicting

future trends, financial information needs to be relevant.

As indicated in the IASB Framework, relevance of information is affected by its

materiality and nature. In some cases, the nature of the information alone may be

53
significant enough to be relevant, while in most circumstances, both the nature and

materiality of the information makes it relevant. Therefore, information is relevant

(by nature and materiality) if its omission or misstatement could influence the

economic decision taken by users based on financial statements.

As previously mentioned, the JWGSS (2000) believe that the use of fair value

accounting results in relevant information. This belief is not shared by the JWGBA

and comment letters received by the IASB suggest that preparers do not believe that

fair value accounting is relevant for financial instruments in the banking book

(JWGBA, 1999b; IASB, 2002). The surveys gather preparer and user views on the

relevance of fair value accounting for all financial instruments.

2.7.3.2 Reliability

To be useful, information must also be reliable according to the IASB Framework.

Reliable information is free from material error and bias (i.e. neutral) and can be

depended upon by users to faithfully represent the item it purports (or could

reasonably be expected) to represent (for example, the Balance Sheet faithfully

represents the transactions that result in assets, liabilities, and equity of the entity at

reporting date). There are situations where information may be relevant but so

unreliable in nature that its recognition may be potentially misleading. In such

circumstances, the information should only be disclosed but not included in the

financial statements.

The IASB Framework acknowledges that most financial information is subject to

some risk of being less than a completely faithful representation due to inherent

difficulties in identifying underlying transactions or in applying measurement

54
techniques. In addition, the Framework stresses that reliable information is

accounted for and presented in accordance with the substance and economic reality

and not merely legal form. However, prudence, a degree of caution in the exercise of

judgment, needs to be exercised by preparers of financial statements when dealing

with the uncertainties that inevitably surround some situations. This ensures that

assets or income are not overstated and liabilities or expenses are not understated.

All of these will add to the reliability of the resulting accounting information.

The issue of reliability of fair values is a crucial issue in the fair value debate

between the JWGSS and the JWGBA. Preparers are highly concerned about the lack

of reliability of fair values for financial instruments, especially non-traded financial

instruments. On the other hand, the JWGSS believe that only certain private equity

investments might not be practicable of reliable fair value measurement and the Draft

Standard (2000) provides an exemption for these (see paragraph 122).

2.7.3.3 Understandability

Another essential quality of useful information is that it be readily understandable by

users. In this instance, users are assumed (in the IASB Framework) to have a

reasonable knowledge of business and economic activities and accounting and have

the willingness to study the information with reasonable diligence (IASB

Framework, 1989, para. 25). It is also states that information about complex matters

that are relevant to the decision making process should not be excluded merely on

the grounds that it may be too difficult to understand. This assumption is questioned

by the evidence in this study (see Chapter Six).

55
In the fair value accounting debate, preparers are concerned that users may not

understand the complexities introduced by fair value accounting. The JWGBA argue

that users are satisfied with (and understand) the current mixed measurement model

and moves to fair value accounting will impair the understandability of financial

statements and ultimately, users will misunderstand the resulting volatility in

earnings introduced by fair value accounting. Conversely, the JWGSS (2000)

believes that the goal of understandability can be accomplished by presenting

relevant information.

2.7.3.4 Comparability

Users of financial information must be able to compare financial statements of an

entity over time and with other entities. This facilitates the identification of trends in

financial position and performance as well as evaluation of the entity’s performance

and changes in financial position relative to other entities. Thus, the measurement

and presentation of like transactions or events must be consistent throughout an

entity and over time and be consistent for different entities. However, the need for

comparability is not merely uniformity and should not impede the introduction of

improved accounting standards at the expense of comparability.

The JWGBA (1999a) argues that the existing mixed measurement system is well

understood and provides a satisfactory degree of comparability and moves to fair

value accounting will reduce comparability. Conversely, in their basis for

conclusions, the JWGSS (2000) believes a higher level of comparability will be

achieved when a common accounting measurement policy (i.e. fair value in this

case) is adopted. Preparer and user perceptions of the comparability of information

resulting from fair value accounting are also examined in this thesis.

56
2.7.3.5 Timeliness

Delays in reporting financial information may result in those financial statements

losing their relevance. Therefore, entities need to balance the relative merits of

timely reporting and the provision of reliable information. In order to provide

information on a timely basis, it may be necessary to report before all aspects of a

transaction are known, impairing reliability. Conversely, if reporting is delayed until

all aspects are known, the highly reliable information is irrelevant and of little use to

users who had to make decisions in the interim. Ultimately, there needs to be

balance between relevance and reliability with the overriding consideration being

‘the best to satisfy the economic decision making needs of users’. Both the JWGBA

and the JWGSS did not highlight this qualitative characteristic in their discussions.

2.7.3.6 Cost versus Benefit

The IASB Framework also recognises the need to balance between benefit and cost,

a pervasive constraint on the quality of information. Essentially, the benefits of

resulting accounting information should exceed the cost of providing it. However,

the accurate determination of either costs or benefits is substantially a judgment call.

Although it is difficult to apply the cost benefit test, preparers and users need to be

aware of this constraint. The IASC (1997) only briefly mention the need for any

proposal on the accounting for financial instruments to be subject to feasibility and

cost-benefit checks but acknowledged the difficulty of such a qualitative evaluation.

As expected, the JWGBA (1999b) implied that the cost of implementing a fair value

accounting system outweighs the unproven assertions on the perceived benefits users

are to gain from its use. The JWGSS (2000, para. 1.4) was not explicit in their

assessment of the cost benefit of fair value accounting but acknowledged “the need

57
to strike a reasonable balance between conceptual ideals and practical cost-benefit

considerations”.

2.7.4 An Expectations Gap

As previously mentioned, there is considerable dispute between the JWGBA and

JWGSS on the actual information needs of users of financial institutions’ financial

statements. Both groups voice their opinions on the user groups’ information needs

without sufficient empirical evidence. The main reason for this is the lack of

empirical research evidence in this area amongst past accounting literature (Walker

and Jones, 2003).

This lack of user focus in the standard setting process is even referred to as a

‘systemic problem’ by Jonas and Young (1998). They urged academics to play a key

role in bridging the gap between users and standard setters in order to promote the

efficient allocation of capital through more relevant standards. Tower (1993) also

called for the bridging of the expectation gap between the producers and

stakeholders. However, this user focus was not intended to mean that users should

get all the information that they want and all other constituents, such as preparers,

auditors and regulators, get ignored in the process.

On that note, where a user focus is adopted, none of the other constituents will lose

out as the information provided by business reporting is ‘truly’ valuable and relevant

(Jonas and Young, 1998; Puxty and Laughlin, 1983). These authors argue that the

standard-setting process need to emphasise how information is used and not on how

information is currently accounted for. In addition, there needs to be more reliance

on facts rather than speculation about the decision usefulness of information (Jonas

58
and Young, 1998). This calls for a mechanism to institutionalise a direct pipeline to

users and also the redirection of research energies towards studying user information

needs in the international controversy on the accounting for financial instruments.

2.8 SUMMARY

This chapter reviewed the relevant literature on the controversy surrounding fair

value accounting and user information needs. The proposed fair value accounting

model for financial instruments and the arguments put forward by both the JWGBA

and the JWGSS forms the core of the literature examined. Literature on the value

relevance of fair value accounting and key studies relevant to this research project

are also critiqued. The following chapter details the research approach taken in this

study from the broad epistemology position to the specific research methods used.

59
 CHAPTER THREE:
RESEARCH APPROACH

3.1 INTRODUCTION

This chapter discusses the accounting research process broken into its component

parts of epistemology, theoretical perspective, research methodology and research

method. Justification for the research approach adopted (as well as the theoretical

position taken) in this thesis to address the two key research questions is also

provided.

Ancient arguments about the nature of knowledge and how it is acquired (rationalism

and empiricism) are still controversial (Ryan, Scapens and Theobald, 1992).

Rationalism comes from the Greeks (namely Plato) and proves tenacious in Western

culture mainly among those who spent considerable time throughout their education

in improving their powers of reason. This idea emphasises the power of logic and

mathematics in deciding the truth of competing theoretical arguments and asserts that

“real truths about the world cannot be discerned by observation alone but by reason”

(Ryan, et. al., 1992, p. 5).

However, in the eighteenth and nineteenth centuries a different philosophical

tradition suspicious of the realism idea emerged. This idea became known as

empiricism and views logic and mathematics as mere tools for exploring the

implications of observed knowledge (Ryan, et. al., 1992). This is the stance taken in

this project.

60
3.2 THE RESEARCH PROCESS

Sproull (1995) describes the research process as specific planned and controlled steps

to empirically investigate a problem. The view that accounting is a social science is

widely held with its distinct subject matter with underlying regularities conducive to

empirical relationships, concepts, authoritative generalisations, principles, laws and

theories (Riahi-Belkaoui, 1996). This position is further evidenced by the quality

and quantity of its research output and its researchers in the establishment of a

political, social and economic order and in the process led to a myriad of

perspectives in the conduct and practice of research (Riahi-Belkaoui, 1997). These

perspectives comprise different epistemologies, theoretical positions, research

methodologies and research methods, which in aggregate, forms the research process

(Crotty, 1998; Chua, 1986). Figure 3.1 shows the relationships between the different

components in the research process.

Figure 3.1: The Relationship between the Components of the Research Process

Epistemology Theoretical Research Research

Perspective Methodology Methods
The philosophical The concept Plan of action Specific techniques
grounding of reality used to used to discover
discover reality reality

As shown in Figure 3.1, the research process is broken down into four basic

elements: (1) the epistemology that drives the theoretical perspective, (2) the

theoretical perspectives behind the research methodology, (3) the methodology that

governs the choice of research methods, and (4) the methods to be used in the

research. Addressing these elements ensures that the theoretical assumptions and

knowledge underlying this research project are understood and the research

61
methodology justified, and thus, the research findings warranted. This research

project adopts a predominantly objectivist epistemological and subsequently, a

positivist theoretical perspective is used to derive the research methodology of user

and preparer surveys and statistical analyses as the research method. Each of these

elements is addressed in turn in the following sections.

3.3 EPISTEMOLOGY - OBJECTIVISM

Maynard (1994, p. 10) states that epistemology provides “a philosophical grounding

for deciding what kinds of knowledge are possible and how we can ensure that they

are both adequate and legitimate” and thus, is a way of explaining and understanding

how we know what we know. The three main epistemological approaches available

to researchers and examples (by no mean exhaustive) of the following three elements

driven by each approach are outlined in Table 3.1.

Table 3.1: Different Stances of the Research Process

Epistemology Theoretical Research Research Methods

Perspective Methodology

Objectivism ⇒ Positivism Verification of ⇒ Statistical analyses

⇒ Postpositivism hypotheses using ⇒ Measurement and
(findings are true)
mainly quantitative scaling
methods
⇒ Sampling
⇒ Survey research
⇒ Data reduction
⇒ Experiments

Constructivism ⇒ Interpretivism ⇒ Ethnography ⇒ Participant and non-

⇒ Symbolic ⇒ Grounded theory participant observation
(value mediated
findings) interactionism ⇒ Heuristic inquiry ⇒ Visual ethnographic
methods
⇒ Interpretive methods

Subjectivism ⇒ Postmodernism ⇒ Action research ⇒ Conversational

⇒ Critical inquiry ⇒ Discourse analysis analysis
(created findings)
⇒ Feminism ⇒ Feminist standpoint ⇒ Life history
research ⇒ Narrative

Adapted from Crotty (1998) and Perry, Alizadeh and Riege (1997).

62
This research project adopts a predominantly objectivist epistemology which posits

that meaningful reality exists independently of any consciousness such that the world

is external and objective (Crotty, 1998). A key tenet of objectivism is that the

researcher should focus on facts, look for causality, formulate hypotheses and then

test them. Table 3.2 details the main differences between objectivism and non-

objectivism epistemologies.

Table 3.2: Differences between Objectivism and Non-objectivism Research

Objectivism Non-objectivism*
Research objective To quantify the data and To gain an understanding of
generalise the results from the underlying reasons and
the sample to the population motivations
of interest
Sample Larger numbers of Smaller number of non-
representative cases representative cases
Data collection Structured Unstructured
Data analyses Statistical Non-statistical
Outcome Recommend a final course of Develop an initial
action understanding
* For example the constructivism and subjectivism epistemologies.
Adapted from Malhotra (1993)

Adoption of the objectivism as the epistemology does result in an opportunity cost in

that it is not always possible (or desirable) to use fully structured statistical analyses

to obtain information and putting too much faith on the assumption that science is

value-free (as opposed to human interest driven) (Malhotra, 1993). Therefore, this

research project balances the final objective-positivist approach with an initial

interpretive or qualitative approach, which has the potential to enrich the

understanding of quantitative results (Pernice, 1996 and Teagarden et. al., 1995).

63
3.4 THEORETICAL PERSPECTIVE – POSITIVISM

The theoretical perspective describes the philosophical stance behind a certain

chosen methodology and provides a context of the logic and criteria employed

(Crotty, 1998). Guba and Lincoln (1994, p. 105) define it as the overall conceptual

framework or the “basic belief system or worldwide view that guides the

investigation”. Theoretical perspective is also synonymously known as paradigm

(Sarantakos, 1993).

Positivism is chosen as the theoretical perspective of this research project because of

its ability to help explain real world phenomena. This approach views social science

as an organised method that combines deductive logic with empirical observations of

individual behaviour to discover and confirm a set of probable causal laws that can

be used to predict general patterns of human activity (Neuman, 2000).

The main strengths of positivism is that it provides a scientific explanation of

research, the use of precise quantitative data, surveys and statistics and a rigorous

and objective measure for testing hypotheses (Neuman, 2000). According to Couch

(1987), it also assumes that empirical facts exist apart from ideas and is free of

political, religious or personal values with an aura of disinterestedness and alleged

objectivity. This theoretical perspective is widely used (Crotty, 1998) and has been

internalised as a norm of the scientific community (Neuman, 2000).

However, Husserl (1931) suggests that a positivist scientific world results in an

abstraction and is not the world that people experience. Critics argue that

disinterested observation is not possible as observation with a theory’s context is

always shaped by theory within a set of selective expectations (Popper, 1972). In

addition, Kuhn (1977) states that there will be a time when this positivist paradigm is

64
inadequate because it cannot explain something within context while Feyerabend

(1987) argues that it is historically conditioned, sloppy and not as objective as it

purports. Hirschman (1985) claims that human attitudes, ideologies and values affect

science because science itself is created by people. Thus, these critics feel that

science is more of a normative field of knowledge as it is more person-centred rather

than phenomenon-centred.

Despite the above problems, positivism has much to offer. It offers both a

quantitative and qualitative framework from which to evaluate information gleaned

from the research methodology. Kuhn (1977) recognises positivism as a flexible

theoretical framework while remaining homogenous in a multitude of contexts and

this is the view this thesis takes. Therefore, positivism is adopted to explain the

extent of support for fair value accounting for all financial instruments from both

user and preparer perspectives, additional interpretive methods are also used to

overcome some of its criticisms.

3.5 RESEARCH METHODOLOGY

Crotty (1998) defines this element as a plan of action or design behind the choice and

use of particular methods and linking them to desired outcomes. Different research

methodologies are used depending on the underlying epistemology stance taken in a

research project. As previously mentioned, this study balances the objective-

positivist approach with a qualitative approach (interviews and phone

surveys/interviews). Figure 3.2 shows the various research methodologies relevant

to the different epistemological stances.

65
Figure 3.2: Preferred Methodologies for Different Epistemologies
Inductive Theory
Building

Grounded Theory
Constructivism
QUALITATIVE METHODS

Action Research
(Analytical generalisations)

Participant Observations

Focus Groups

Convergent Interviews Subjectivism

Interviews
Case Studies

Surveys
Positivism
Experiments

Deductive Theory
QUANTITATIVE METHODS Testing
(Statistical generalisations)

Adapted from Hastings (2000) and Lachlan (1995)

There has been considerable interest in the application of both quantitative and

qualitative methodologies and this idea is gaining support (Edmondson, 1996; Hayne

and Pollard, 2000; Judge, Thoresen, Bono and Patton, 2001; Shaffer and Harrison,

2001; Trevelyan, 2001). The inclusion of qualitative research components in an

otherwise quantitative methodology has tremendous potential in enhancing the

understanding of the results obtained from statistical analyses (Hodgkinson and

Payne, 1998; Wilk, 2001). Aram, Salipante and Knauf (1987) as well as Sutton and

Rafaeli (1988) go further to argue that qualitative data provide rich information that

may not be captured in quantitative techniques.

66
In this project, the integration of both qualitative and quantitative (sometimes known

as triangulation) techniques is important to achieve a better understanding of the

issues at hand. Qualitative approaches are used in an exploratory manner to first

gather deeper insights into the critical issues relevant to the two identified subject

groups (namely preparers and users) as well as to gather information about

significant concepts and terms (Ticehurst and Veal, 2000). Here, it is used to gather

richer information from a small number of key subjects using informal and in-depth

interviews as well as phone surveys. Together, evidence gathered from the

interviews and phone surveys adds to the foundation (and the relevance and validity)

of the preparer and user surveys used in the quantitative phase. Figure 3.3 depicts

the research methodology of this project.

Figure 3.3: Research Methodology Adopted

Qualitative Phase
ƒ Key Interviews
ƒ Phone Surveys

Quantitative Phase

Survey of Preparers Survey of Users in

in Australia and Australia and
Singapore Singapore

Comparison

Following the ‘rich’ information gathering qualitative phase, the quantitative survey

method is used to gain insights into the extent of support (or opposition) of

Australian and Singaporean preparer and user views on fair value accounting for all

67
financial instruments held by financial institutions. Specific themes and issues

identified in the qualitative phase form the basis for the two similar surveys sent out

to preparers and users, respectively. This phase can be segregated into preparer and

user categories. Following this, comparisons are made between user and preparer

perceptions and between them and the ‘official’ stance (i.e. the JWGSS and

JWGBA).

Evidence is derived from preparer and user surveys (both mail and phone) in

Australia and Singapore, two significant financial markets in the Asian-Pacific

region.

3.6 RESEARCH METHODS – QUALITATIVE PHASE

Different research methods are adopted for the two distinct phases of this study (see

Figure 3.3). Key interviews and phone surveys/interviews are the tools utilised in the

qualitative phase to provide deeper insights into the crucial issues of the preference

for fair value accounting. Osteraker (2001) and Fontana and Frey (2000) argue that

interviews and surveys are most appropriate in opinion research where the research is

interested in perceptions. These interviews serve to set the parameters and scope of

the phone survey (and mail survey). Subsequently, two separate (but equivalent)

survey instruments are used to gather evidence in the quantitative phase for the

preparer and user groups respectively. The survey method proves a useful approach

to empirically study sociological characteristics and interrelationships (Roberts,

1999). Both mail and phone surveys are adopted to further enhance the collection of

evidence.

68
Figure 3.4 positions the numerous research methods available within a two-

dimensional axes defined in terms of two primary research objectives of “data

integrity” and “data currency” (Bonoma, 1985, p. 200). Data integrity refers to the

characteristics that affect error and bias in research results; an amalgamation of what

is variously termed internal validity (Campbell and Stanley, 1963) and reliability

(Guildford, 1965). Inversely, currency indicates the contextual relevance of data

gathered and pertains to the generalisability of the results, also termed “external

validity” by Cronbach and Meehl (1955). In the ideal world only methods that

provide data of high integrity and currency are used in all research (Bonoma, 1985).

Figure 3.4: A Knowledge Accrual Triangle

High

Laboratory
experiments

Models
DATA INTEGRITY

Stimulations

Field experiments
Interviews
Surveys

Case Studies
Science
Non-science

Low
Low CURRENCY High

Adapted from Bonoma (1985)

Both the interview and survey methods yield evidence with inherent currency and

integrity. These two methods are most appropriate for the purposes of theory

69
building in the relatively unexplored area of preference for fair value accounting for

financial instruments from both the preparer and user points of view. Following the

collection of data, univariate and multivariate statistical techniques (mainly multiple

regressions, one-way ANOVAs and T-tests) are then used for comparison and

explanation purposes. All the research methods utilised in this project are discussed

in the following subsections.

3.6.1 Key Interviews

The use of the interview technique to acquire information is so frequent in recent

times. Atkinson and Silverman (1997) brand the society today as the ‘interview

society’. Both qualitative and quantitative researchers tend to rely on interviews as

one of the basic methods of data collection to gather richer, in-depth experiential

accounts (Fontana and Frey, 2000). There is inherent faith that the results of the

interview process are trustworthy and accurate without undue bias (Silverman,

1993). Various forms of the interviewing technique have arisen, from structured to

unstructured types, formal to informal to spontaneous and face-to-face to electronic

means as a result of the persistent opinion polling since before the twentieth century,

(Fontana and Frey, 2000). It is however, crucial that the interviewer listens and

encourages respondents to talk and refrain from agreeing (or disagreeing) with them

or suggesting answers (Ticehurst and Veal, 2000). Even though critics of this

method are concerned with interviewer bias, Singer and Presser (1989), in their

research on interviewer effects show that interviewer characteristics of gender, age

and interviewing experience have relatively small impacts on responses.

Key interviews are used due to the complex nature of the issues being explored. The

interviews are relatively unstructured with the intention of probing for more

70
information from the key players in the fair value accounting debate. This

qualitative method is commonly used when the subjects of the research are relatively

few, or when high variation in the information likely to be obtained is expected or

when the topic to be explored is part of a preliminary stage of a larger (and possibly

quantitative) study (Ticehurst and Veal, 2000). The last two reasons make the use of

less structured interviews more appropriate in this thesis.

The interviews were carried out for an exploratory purpose to identify crucial issues

in the debate on fair value accounting for all financial instruments with standard-

setter representatives, active academics and practitioners in this field as well as key

representatives from both the user and preparer groups both locally and

internationally. The interviews conducted were informal with a semi-structured

question format and the interviewer’s (i.e. the researcher) role was moderately

directive (where the researcher will only use probe questions when key issues were

not covered). The semi-structured (also open-ended) format provides greater breadth

of data and the moderately directive roles ensures that essential issues are adequately

covered (Ticehurst and Veal, 2000).

3.6.2 Phone Surveys

Phone surveys provide an interactive environment where issues could be clarified

and additional insights and follow-ups obtained on the views of sophisticated users in

relation to fair value accounting for all financial instruments. Thus, it provides more

in-depth and arguably more reliable evidence than what could be obtained through

conventional means, such as mail surveys. For example, Lavrakas (1987) asserts that

telephone interviews provide the opportunity for quality control over the entire data

collection process. Although Dillman (1978) highlights certain deficiencies and

71
problems, phone surveys are used extensively in market and academic research

because of their speed and ease with which a widespread sample can be contacted as

well as cost effectiveness (Ticehurst and Veal, 2000; Weitz, 1990). Another

advantage of this method includes respondents feeling more anonymous and

therefore, may be more forthcoming in their opinions (Ticehurst and Veal, 2000). In

addition, the lack of past research evidence on actual information needs of users,

makes an interactive interview a more appropriate medium to probe and gather users’

views.

This method is adopted to gather deeper and more detailed insights into the views of

users about the usefulness of fair value accounting for all financial instruments.

Given the very busy professional lives of the respondent group, the use of phone

surveys enabled a higher response and two-way feedback and clarification of

complex terms and concepts. Two separate phone surveys were conducted on

Australian sophisticated users15 over a ten-month period to gather evidence as to the

views of sophisticated users about fair value accounting for financial instruments and

to determine if such views changed with the announcement in July 2002 that

Australian businesses are to fully adopt International Accounting Standards for

financial reporting purposes from 1 January 2005.

3.6.3 Statistical Analyses

In this phase, evidence from the interviews is narrated. The main issues and points

raised in the interviews are then summarised and key themes to be further explored

in the quantitative phase identified. For the phone survey portion, descriptive

statistics are employed (using univariate tools) to detail the demographics and to

15
Singaporean users were not phone surveyed due to financial constraints and lack of access.

72
summarise responses and T-tests conducted to test for differences between user

responses. Evidence collected is used to develop the surveys to be used in the

quantitative phase.

3.7 RESEARCH METHODS – QUANTITATIVE PHASE

3.7.1 Mail Surveys

Following the ‘rich’ information gathering qualitative phase, the quantitative survey

method is used to gather Australian and Singaporean preparer and user views on fair

value accounting for all financial instruments in financial institutions. Specific

themes and issues identified in the qualitative phase forms the basis for the two

similar surveys sent out to preparers and users, respectively. This phase can be

separated into preparer and user categories.

Surveys are a quick and inexpensive means of eliciting information about a

population’s views (Zikmund, 1997). The survey method is utilised to obtain

quantitative data and refers to investigations into a social matter involving systematic

measurements made over a series of cases resulting in a rectangle of data where the

variables are analysed for patterns (Marsh, 1982). It is a useful (and most commonly

used) approach to empirically study sociological characteristics and

interrelationships (Roberts, 1999; Ticehurst and Veal, 2000).

In survey research, patterns are proposed to exist among the variables of interest and

thus, the research model and its theoretical underpinnings are of critical importance

(Oppenheim, 1966; Roberts, 1999). The form of data collection and the method of

analysis for surveys differ from other research methods. Surveys prove an effective

means of collecting considerable volume of data on the same variables from large

73
numbers, allowing for greater scope but limiting the depth of the study (Bernstein,

Roy, Scrull and Wickens, 1988). In addition, good surveys allow the

interrelationships between variables to be rigourously tested and allow inferences to

be drawn about the population (i.e. provides more realism).

Surveys, if poorly conducted waste resources, possess bias (where results deviate

from population parameter’s true value), contain poorly phrased questions and cause

misleading results (Zikmund, 1999). This method is also susceptible to random

sampling error (difference between sample and census result) and non-response error

(statistical difference between a survey with only respondents and a survey that

includes both respondents and those who did not respond) (Dillman, 2000). Critics

of the method further argue that surveys lack clear definition, contain measurement

error causing unreliability, provide poor quality of responses and lowly response rate

(Ticehurst and Veal, 2000; Kerlinger, 1986). In addition, there is dispute as to

whether surveys can adequately establish causal connections between variables being

tested (Roberts, 1999).

Some of the main criticisms of the survey method can be addressed by making clear

the data that will be collected by the survey and explaining the importance of the

information before collection (Cavana, Delahaye and Sekaran, 2001). Safeguarding

techniques in the design of the survey and mail-out procedures are employed to

overcome these criticisms. In addition, the benefits of using the survey method in

this project far outweigh the failings of this technique, especially in enabling the

collection of evidence on the perceptions of fair value accounting from both users’

and preparers’ point of view. The qualitative phase conducted to develop the survey

better ensures the relevance and reliability/validity of survey questions.

74
In this project, the survey instrument is used to gather information about preparers’

and users’ (in both Australia and Singapore) preference for fair value accounting for

all financial instruments. Both financial institutions’ chief financial officers and key

sophisticated user groups in these countries are surveyed to facilitate the comparison

of preparer and user viewpoints on the proposed fair value accounting model. Two

similar versions of the questionnaire are developed; one each for preparer and user

(see Appendix A to D for actual surveys). In this capacity, the broad guidelines and

procedures for questionnaire development as suggested by Andrews (1984), Roberts

(1999) and Dillman (2000) are followed. The preparer survey is developed first and

then used to structure and develop the user version. These two surveys follow

suggested guidelines and underwent rigorous ethics approval by the Human Research

Ethics Committee at Murdoch University. The development of the two surveys

utilised in this project is detailed in the following subsections.

3.7.1.1 The Preparer Survey

Firstly, an extensive review of related instruments is conducted and a prototype of

the preparer questionnaire developed (see Appendix E). Following this, the

prototype questionnaire was adapted to the purposes of this study in the right context.

This is achieved with help from academic and professional experts in the field

through the interviews and phone surveys conducted in the qualitative phase.

Before the questionnaire is disseminated, the instrument is rigorously tested for

validity, relevance and optimal wording of questions through pilot testing. This is

done to identify any problems with the survey and the survey design itself, including

the procedures adopted (Hill, 1993). In this project, the refined prototype preparer

survey (as per Appendix F) is sent to 212 chief financial officers (essentially as a trial

75
run) to check for relevance and appropriateness of the survey. This pilot testing of

the questionnaire improved the reliability and validity of the data collected by

identifying potential problems. Questions that have ambiguous phrasing are either

excluded or reworded to ensure accuracy in data collected. The length of the survey

is also scrutinised to take into account the likely response rate for the actual mail-out

and the busy nature of the industry. See Appendix H for details.

Following the pilot testing, suggestions and comments from respondents result in a

shorter, more refined survey. This refined survey only focuses on the key themes as

identified from early interviews (and the pilot study) to facilitate the most efficient

collection of crucial evidence16. In the process of refining the survey questions, the

relevance, comprehensibility and feasibility of each question in relation to various

aspects of the problem being investigated is considered in accordance with the

suggestions of Swift (1979). The refined, trialed and tested questionnaire was then

sent to chief financial officers of financial institutions and in Australia and

Singapore. This final preparer survey forms the base for the user survey.

3.7.1.2 The User Survey

The two groups (i.e. preparer versus user) being examined in this research project

have varying levels of experience to warrant the use of slightly different surveys.

Questions in the preparer survey are deemed highly relevant (based on feedback

obtained through the pilot testing stage) and understandable to chief financial officers

of financial institutions but may not be so for a user. Therefore, questions specific to

preparers (for example questions relating to the asset base of the financial institution)

16
For example, the new survey excluded question 6 and 7 and amalgamated statements asked in
questions 5 and 10. The pilot-tested survey was 6 pages long and asked twice as many questions.
Questions that are unclear or outside the scope of this study were either reworded or excluded.

76
were either adjusted into the appropriate user context or simply removed. In

addition, users’ perceptions of the usefulness of fair value accounting for all financial

instruments needed to be adequately explored and questions to this effect were

added. Some examples are questions that focus on the qualitative characteristics of

useful information as defined in the IASB Framework (questions 3i to 3vi in

Appendix B). Finally, clarification on terms specific to financial institutions needed

to be added to ensure accuracy of survey evidence collected from the users.

However, the core questions addressing key themes of this project were identical

between the two surveys for comparison purposes (see Appendix B and D for the

actual user surveys – differences between the surveys are highlighted in blue).

3.7.2 Statistical Analyses

Responses to the surveys are first coded into numbers. For example responses of

strongly oppose, oppose, neutral, support, and strongly support are replaced with the

values 1, 2, 3, 4 and 5 respectively to quantify the strength of support for the

respective question. Then, a one-sample t-test is used to compare the means to the

neutral midpoint position (3 on the 5-point scale). Significant p-values indicate when

the means are significantly different from 3 (i.e. a neutral response). Averages of

survey responses are also calculated for a snapshot view of respondent perceptions.

These statistical techniques are used throughout the thesis. They are followed by

other main statistical techniques including the T-tests, multiple regression techniques

and ANOVAs to examine the perceptions of fair value accounting for financial

instruments and check for differences in opinions between users and preparers.

In this study, multiple regression techniques via the general linear model are used to

check for possible determinants for the preparer and user views on fair value

77
accounting for all financial instruments. Multiple linear regression is engaged when

the variation of the dependent variable is thought to be explained by the presence of

more than one independent variable (see for example Davis and Cosenza, 1985; Gay

and Diehl, 1992). The generalised form of the multiple linear regression model is:

Y = a0 + a1X1 + a2X2 + … + akXk + e

Where Y = the dependent variable;

X = the independent variable;
a = the coefficient to be estimated;
k = (1, 2, …, p);
p = the number of independent variables;
e = the error term.
Adapted from Cohen and Cohen (1983)

The objective of multiple regression analysis is to arrive at the best set of coefficients

for the independent variables that bring the dependent variables predicted from the

equation as close as possible to the actual values observed (Tabachnick and Fidell,

1996). Thus, it calculates an overall explanatory figure (r-squared) and individual p-

values for the independent variables to ascertain possible predictors. In all the

general linear models fitted for preparers, country and type of financial institution are

treated as factors (categorical variables) while size of financial institutions,

experience and order (of response) are treated as covariates (continuous variables).

Similarly, for general linear models of the user survey data, country and type of user

are factors while experience and order (of response) are covariates. Residual plots

for each model fitted are also checked to ensure the underlying assumptions were

met (See Section 3.8).

The two-sample T-test is a statistical technique for assessing the statistical

significance of the difference between two independent sample means (Hair,

Andersen, Tatham and Black, 1998). It is also known as a univariate hypothesis test

78
used when the population standard deviation is unknown and the sample size is small

(Zikmund, 1997). Independent sample t-tests are used to determine if observed

variations between two groups’ means are statistically significant (Shavelson, 1996).

The one-Sample T-Test procedure was used to test whether the mean of a single

variable differs from the neutral position for questions in the surveys.

Before the sample data was statistically analysed, it was checked for errors and

fulfillment of the underlying assumptions for inferential statistics. This is part of the

data cleaning process that was conducted as detailed in the following section.

3.8 DATA CLEANING

Prior to the statistical dissection of data for multivariate analyses, several issues

should be resolved. Tabachnick and Fidell (1996) and Hair, et al. (1998) argue that

the following issues need to be considered to achieve an appropriate analysis:

ƒ Accuracy of data,

ƒ Analysis of missing values,

ƒ Multicollinearity,

ƒ Fulfilment of statistical assumptions, and

ƒ Identification of outliers.

Each of these issues is dealt with separately in the following sub-sections.

3.8.1 Accuracy of Data

According to Tabachnick and Fidell (1996), the best way to ensure accuracy of data

is to proofread the original data against the computerised record. This was

conducted with the help of a fellow business student whose responsibility is to check

79
the computer file while the researcher calls out the data from the original surveys.

As a further check, frequencies and descriptive tables are also scrutinised for any

score that was not within the range of responses. Minimum and maximum scores are

carefully examined to check for scores that fell outside the scale of each item in the

survey. There were two instances where the score of “6” was found for items with

only a scale of 1 to 5. These were checked against the original surveys and rectified.

Following this, another business PhD student carried out a data entry process for a

sample of 10% each of the preparer and user surveys. The data file was then cross-

checked against the original data file and no discrepancies were found. This 100%

complete level of agreement indicates that the measurement and classification of the

data is reliable.

3.8.2 Missing Values Analysis

Tabachnick and Fidell (1996) proposed five methods to handle missing data, one of

the most pervasive problems in data analysis. Frequencies were calculated to

identify the occurrence of missing values. Fortunately, there were no instances of

surveys that had missing data. All eighty-three preparer respondents and sixty-five

user respondents from Australia and Singapore filled out their respective surveys

completely.

3.8.3 Multicollinearity

The third issue to resolve is multicollinearity. Hair, et al. (1998) stated that the

ability of the independent variables to predict the dependent variables is related, to a

certain extent, to the correlation between the independent variables.

Multicollinearity reduces any individual independent variable’s predictive power by

80
the extent to which it is associated with the other independent variables, making it

redundant. Inclusion of redundant variables in the same analysis inflates the size of

error terms and actually weakens an analysis (Tabachnick & Fidell, 1996).

Therefore, if independent variables are highly correlated, the unique variance

explained by each variable is decreased.

Bivariate correlations, one-way ANOVA’s, independent samples T-test and cross-

tabs are conducted to test for relationship between the independent variables for both

the preparer and user survey respectively. The results of these tests are discussed in

Sections 5.4 and 6.4. Overall, there are correlations between the independent

variables but none are severe enough to cause a multicollinearity problem.

Implications of such relationships are explored in the discussion on the results of the

regressions in both Chapters Five and Six.

3.8.4 Fulfilment of Statistical Assumptions

The assumption of normality underlies most statistical tests and some multivariate

procedures (Tabachnick and Fidell, 1996). Variables with normal distributions

improve the results obtained from the analysis. Normality of variables can be

assessed using either statistical or graphical methods. In this study, residual plots

were tested for each multiple linear regression (or general linear model) and there

were no instances of violation of underlying statistically assumptions.

3.8.5 Identification of Outliers

Hair, et al. (1998) defines outliers as observations with a unique combination of

characteristics that are distinctly different from the other observations. These

outliers can be problematic as they can distort statistics, thus causing misleading

81
results (Tabachnick and Fidell, 1996). The presence of an outlier could be a result of

data entry error; an extraordinary event that has an explanation; an unexplained

extraordinary event; or an outlier might fall within a variable’s ordinary range of

values but are unique in their combination of values across variables (Hair, et. al.,

1998).

Outliers can be detected using either graphical or statistical methods (Tabachnick and

Fidell, 1996). In using graphical methods, outliers are the cases that seem to be

unattached to the rest of the data. An example of a helpful graphical method is the

histogram. Conversely, a statistical procedure that can be used is the computation of

Mahalanobis distance for each case. Tabachnick and Fidell (1996, p. 67) defines the

Mahalanobis distance as “the distance of a case from the centroid of the remaining

cases where the centroid is the point created by the means of all the variables”.

Cases with extreme Mahalanobis scores (exceeding 22.458) separate from the rest of

the scores are considered outliers (Tabachnick and Fidell, 1996).

Cook’s distances were also calculated to further determine if the outliers are to be

deleted. Cook’s distances are measures of influence and cases with influence scores

of more than one are suspected of being outliers (Tabachnick and Fidell, 1996).

Examination of the scores indicated all cases have values of less than one. Both the

Mahalanobis and Cook’s distance were calculated for each multiple regression

analyses. The maximum scores were checked each and every time for outliers.

There were no instances of outliers that are problematic. Therefore, all data were

included in the analyses to improve the generalisability of the results (Hair et al.,

1998). The data set is now ready for further statistical testing.

82
3.9 SUMMARY

This chapter explained the research process adopted for this research project. The

objective-positivist approach adopted is deemed relevant and both qualitative and

quantitative methods are used in conjunction to derive evidence on user and preparer

perceptions of fair value accounting for all financial instruments. Figure 3.5

indicates the relevant evidence chapter for each phase of the research process.

Figure 3.5: Graphical Representation of Chapter Coverage

Qualitative Phase
Chapter Four
ƒ Key Interviews
ƒ Phone Surveys

Quantitative Phase

Survey of Preparers Survey of Users in

Chapter in Australia and Australia and Chapter
Five Singapore Singapore Six

Comparison

Chapter Seven

Evidence collected from the interviews and surveys will be analysed with univariate

and multivariate statistical tools to determine the level of support (or opposition) for

fair value accounting for all financial instruments and to identify factors that explains

user and preparer views. Data collected from both the qualitative and quantitative

phase will be detailed in Chapters Four to Seven. The next chapter highlights the

insights from the qualitative evidence gathered in this study.

83
 CHAPTER FOUR:
QUALITATIVE PHASE

4.1 INTRODUCTION

This chapter details the information gathered from the qualitative phase of interviews

and phone surveys and provides preliminary conclusions for this phase of the project.

4.2 INSIGHTS FROM KEY INTERVIEWS

Prominent Australian players in the fair value accounting debate were contacted17.

They were then asked it they were willing to assist in this research project by

discussing their thoughts on fair value accounting for all financial instruments. The

interviews were conducted over a period of time between July 2000 and July 2001

and included both formal and informal face-to-face interviews as well as one

electronic mail interview. The interviewer travelled across the country to Sydney

and Melbourne for most of the face-to-face interviews. In keeping with

recommendations that a knowledgeable and diverse sample of subjects be employed,

interviewees included preparers, sophisticated users, representatives from users

groups, standard setter representatives and academics who are active in the field of

fair value accounting for financial instruments. Overall, the group had knowledge

and expertise regarding the key financial issues. In all cases, the interviewee was

made aware of the nature and scope of the research project and guarantee or their

anonymity, ensuring that ethical standards were not breached. Sections 5.2.1 through

5.2.4 outlines the responses obtained from these semi-structured interviews.

17
These included key speakers for the JWGSS and JWGBA as well as experts in the field actively
involved in the development of accounting for financial instruments.

84
4.2.1 Preparer Views

Two senior officers from two Australian banks willingly discussed their opinions on

fair value accounting and the major concerns the banking industry has with the IASB

2000 draft standard proposing fair value accounting. Both preparers have different

views. Preparer 1 (P1 for short) strongly opposes the proposed fair value accounting

model, as P1 feels it does not represent the way banks are being managed, especially

in relation to the banking book. On the other hand, P2 (the second preparer) supports

the proposal by commenting that fair value accounting is used for all financial

instruments including hedging instruments. P2 also feels that fair value accounting is

the most relevant method as it reflects all relevant market information on the value of

a financial instrument.

P1’s oppose fair value accounting because it: (1) does not represent the underlying

fundamentals of the banking industry; (2) results in artificial volatility in reported

earnings and (3) lacks specific measurement guidance. According to P1, banks

manage the gap between assets and liabilities (also called Asset and Liability

Management, shortened to ALM) where cash flow information (not fair values) is

used for such purposes. Furthermore, risk management focuses on portfolios or

groups of assets and liabilities, not on an item to item basis, rendering fair value

accounting irrelevant in faithfully representing (and in P1’s view shows the lack of

understanding for) the workings of financial institutions. Conversely, P2 asserts that

fair values of financial instruments are known by managers of financial institutions,

contrary to JWGBA assertions.

P1 asserts that using fair values results in high volatility, “which in turn causes

artificial risk”. Following this, banks have to further hedge this artificial risk due to

85
market reactions to such high levels of volatility. P1 further indicates that the impact

of fair value measurement on management behaviour has short term and long term

consequences, especially from an incentives scheme perspective. P1 further points

out that implementation of the fair value accounting model will incur direct costs to

the entity as well as cost to the community. P1 argues that this is because the lay

person shareholder would be misled by fair values and thus any decision-making will

be flawed. On the other hand, P2 perceives that users are capable of understanding

the consequences of fair value accounting.

P1 questions the accuracy of fair values due to the myriad of factors that are

considered in the calculation of fair values and further challenges “How can the

assumptions or underlying factors be made comparable, verifiable and auditable?”

and “Who decides what factors should be assessed when calculating fair value for

non-marketable instruments?” According to P1, the definition of fair values (i.e. the

amount a willing knowledgeable buyer and a willing knowledgeable seller transacts

at arm’s length) results in “opportunities for banks to increase profits quite

substantially18” if fair values are mandated for all financial instruments. P1 then

questions the appropriateness of measurement rules based on definitions rather than

functions.

Both P1 and P2 perceive disclosure related to fair values to be important. However,

P1 supports fair value disclosure rather than taking the changes in fair values to the

Income Statement and argues the JWGSS proposal provides less information than the

current disclosure regime on fair values19. To illustrate this, P1 provides an example

18
P1 brings up the Origin mortgage portfolio situation where the actual price offered for the portfolio
was extremely high due to the use of fair value accounting.
19
Current disclosure regime includes other information on specific provisions, reprising interval and
average interest rate, to name a few versus the one number in fair value accounting.

86
of disclosure under the two different regimes as shown in Figure 4.1. In addition, P1

highlights the comprehensive fair value information disclosures in the annual report

(also known as the Gap Analysis) to prove the extensive amount of disclosure that is

currently being provided.

Figure 4.1: Differences in fair value accounting and current disclosures.

Fair value regime Current regime

Fair value $102 Face value $100
Specific provisions $20
Reprices every three months
Average interest rate 7.2%

Source: Example provided by P1.

Interviewee P1 also argues that the two categories of banking and trading are

fundamentally different and needs different measurement regimes and stresses that

fair values are NOT used for the banking book. In contrast, interviewee P2 states

that fair value accounting is being used for financial instruments in the banking book

as well and agreed with the JWGSS on the incredible amount of discretion available

in categorising financial instruments. Contrary to JWGBA views, P2 also stresses

that “they (bank managers) know the fair values of the financial instruments and it is

being used”. However, P2 admits that “there may be issues in assurance of the

reliability of the numbers”.

Finally, P1 suggests the research be approached using the conceptual framework to

assess the fair value model for financial instruments for financial institutions. This

study takes this stance in terms of the usefulness of fair value accounting as detailed

in Chapters 5 and 6.

87
P1 and P2 have mostly contrasting views on the proposed fair value accounting

model and P2 offered less information and comments as compared to P1. In general,

P1 disagrees with the arguments and position taken by the JWGSS in the proposed

accounting model. P2, however, agrees that fair value accounting is most relevant

for all financial instruments and should be adopted.

4.2.2 Standard Setter Thoughts

Two representatives from standard-setting bodies (hereon labeled S1 and S2) are

interviewed and unanimously agree that fair value accounting for all financial

instruments is the preferred approach. Both S1 and S2 state that fair value

accounting is more relevant than the mixed measurement model currently adopted

because the current model allows for too much latitude for financial institutions to

manage earnings. S1 even goes as far to say he personally knows of managers who

classify financial instruments in a way to manage earnings. Furthermore, both these

interviewees are of the view that the resulting volatility in earnings from the use of

fair value accounting will not have a big impact on user understanding.

Contrary to JWGBA assertions that banks cannot reliably determine the fair values

of financial instruments in the banking book, interviewee S2 believes that (bank)

managers of the banking book have a firm grasp on the fair values of the instruments

in the portfolio. S2 said “if they are paid big bucks to manage the banking book;

they do know the fair values of all financial instruments”. In addition, S2 perceive

there to be sound fair value models that will enable banks to achieve high reliability

in the estimation of the fair values of financial instruments.

88
S2 also suspects that fair value is used for internal management purposes especially

in the ALM (asset and liability management) department. Interviewee S1 supports

any research that “survey banks to find out the usefulness of fair value information

for both internal and external purposes”. Although both S1 and S2 favour fair value

accounting for all financial instruments, they are concerned that there is a lack of

knowledge on the accounting issues related to financial instruments amongst society

and even sophisticated users.

The two standard setter interviewees voiced support for the arguments put forward

by the JWGSS and did not support the assertions of the JWGBA. Overall, fair value

accounting was perceived to be relevant, reliable and resulted in useful information.

4.2.3 Sophisticated Users’ Perceptions

Three sophisticated users are interviewed (identified as U1, U2 and U3). Interviewee

U1 also happens to be the executive director and the president of two prominent

sophisticated financial user associations in the United States of America. U2 is

involved in the international deliberations on fair value accounting while U3 is a

prominent practitioner and researcher on fair value accounting.

U1 firmly states that the assertions of the banks (on the unreliability of fair values)

are completely false and stresses that “the CEO’s who are being paid millions of

dollars know the fair values of their financial instruments”. U1 argue that the current

mixed measurement model is preferred by banks for income-smoothing purposes. In

addition, U1 asserts that fair value accounting is the most relevant accounting

method for all financial instruments. U1 also feels that there are plenty of widely

accepted and robust tools to estimate the fair values of all financial instruments. U1

89
goes further to state that the opposition from banks is based on fiction and “lame

excuses”.

Similarly, U2 stresses that fair values of financial instruments in the banking books

are known and used by chief financial officers who are well remunerated.

Realistically, financial institutions would not pay chief executives so much money

and not expect them to know the fair value of financial instruments. U2 also believes

that banks want the mixed measurement model to smooth income. In addition, U2

disagrees that the banking and trading books are fundamentally different. This

interviewee also fully supports fair value accounting as the most relevant accounting

method for all financial instruments.

Interviewee U3 spoke from a more practical point of view but again showed support

for fair value accounting for all financial instruments. U3 believes that fair values

are known for all financial instruments and stresses that “the argument that volatility

will confuse ordinary users (and investors) does not stand”. According to U3, the

main users will have to be sophisticated users as most ordinary investors use brokers

and other advisors and, these sophisticated users would (and should) understand the

workings and implications of the fair value accounting model.

In summary, all three sophisticated users show support for JWGSS views on the

superiority of fair value accounting and little agreement with the JWGBA’s

assertions. Conversations with the user interviewees reveal that the usefulness of fair

value accounting financial statements and volatility in reported earnings are two

main issues in this debate.

90
4.2.4 Academic Views

Five prominent academics (A1 to A5) in the financial accounting field were

interviewed separately. A1 is especially active in the fair value accounting debate at

the international level. A1 believes fair value accounting for all financial instruments

is the most relevant accounting method and argue that the mixed measurement model

is really there as a “tool for managing earnings level”. Similarly, interviewee A1

does not believe the financial institutions’ argument that fair values are not known

and are highly unreliable. Both interviewee A2 and A3 have similar opinions.

Interviewee A2 states that fair value accounting is increasingly being mandated in

accounting standards and is the most relevant for all financial instruments.

Meanwhile A3 suggests that it will be the “prevalent accounting model in accounting

standards of the future”.

Although interviewees A4 and A5 were interviewed on separate occasions, their

views are highly aligned. Both agree that fair value accounting is most relevant for

all financial instruments and both do not believe the reasons stipulated by the

JWGBA are valid. Interviewee A4 has previously done research on fair value

accounting and, in that capacity, discovers that financial institutions do know (and

use) fair value accounting for instruments in the banking book. A4 further believes

the distinction for the supposedly different books in financial institution as

“deliberate attempts to smooth income”. A5 has the same sentiment and believes

banks are already using fair value accounting anyway but want to hold on to the

banking book distinction to manage earnings.

Furthermore, all of the interviewees do not believe that the resulting volatility in

reported earnings will result in confusion in investors and strongly believe that

91
existing fair value measurement tools are robust and highly reliable, given adequate

levels of disclosure.

The five academic interviews show strong support for the JWGSS stance on fair

value accounting and do not agree with most of the assertions put forward by the

JWGBA.

4.2.5 Summary of Interviews

Interviews show lack of agreement between user, preparer, standard setter and

academic views on this debate. The findings from all the interviews conducted are

summarised in Table 4.1.

Table 4.1: Preparers’, Standard Setters’, Sophisticated Users’ and Academics’

Views: Interview Summary

Items Preparers Standard Users Academics

Setters
Support for JWGSS position Mixed High High High
Support for JWGBA position Mixed Low Low Low
Support for JWGBA concerns on the Yes No No Mostly Not
volatility in earnings from the use of fair
value accounting
Support for JWGBA concerns on the lack Mixed No No No
of reliable fair values
Support for JWGSS views of the Mixed Yes Yes Yes
relevance of fair value accounting
Support for JWGBA view of fundamental Mixed No No No
difference between the trading and
banking book

Therefore, it seems that the JWGBA’s views are supported by preparers but not by

the other groups as they show support for the use of fair value accounting for

financial instruments. Notwithstanding this, there are still prominent key themes that

92
have arisen from the interviews conducted. These key themes are discussed in the

section below.

4.2.6 Key Themes Identified

Based on the evidence gathered during the qualitative interview phase, four major

themes evolved. The main theme is the preference for fair value accounting for all

financial instruments as compared to the current mixed measurement model that is

used for the banking and trading books. Subsequently, a second theme related to the

alleged differences that led to the distinction between the trading and banking book

can be identified. A third theme relates to whether fair value accounting for all

financial instruments results in information that is relevant, reliable, comparable and

understandable (characteristics of decision-useful information). The last theme

addresses the concern with the volatility in earnings that results from the use of fair

value accounting. Table 4.2 summarises the responses to the interviews categorized

into these four themes. These four themes were used as the foundation for the

development of the preparer and user surveys.

Following the initial person-to-person interview stage, phone surveys were carried

out to gather insights into users’ perceptions of fair value accounting for all financial

instruments in the financial institutions industry. Section 4.3.1 details the evidence

gathered from the first phone survey while Section 4.3.2 shows the results of a

second phone survey conducted ten months later to enrich the understanding of the

issues at hand.

93
Table 4.2: Summary of Preparer, User, Standard Setters’ and Academics’ Interviews

Theme User Interviews Preparer Interviews Standard Setters Academics

Distinction between trading and banking No distinction Mixed; both support and No difference No distinction
books opposition
Relevance of fair value accounting Agree One agreed, one disagreed Relevant Relevant
Reliability of fair values for traded Highly reliable Highly reliable Highly reliable Reliable
financial instruments
Reliability of fair values for non-traded Robust tools to ensure Not entirely reliable Reliable Reliable
financial instruments reliability
Comparability of fair value accounting Not asked Not asked Not asked Not asked
Volatility may be misunderstood by users Disagree Agree Not asked Disagree
Overall support for fair value accounting Support One supported, one opposed its Support Support
use in the banking book

94
4.3 PHONE SURVEY RESULTS

4.3.1 The First Phone Survey

A list of prominent Australian sophisticated users’ contact numbers was compiled

using resources on the internet. Sophisticated users from organisations ranging from

audit firms to securities’ firms to ratings agencies were included on the list. Each

user was phoned and then asked if they were willing to spare a few minutes of their

time to answer this phone survey. Once consent was given, each user was asked to

answer seven questions relating to the decision usefulness of fair value accounting

for all financial instruments in their capacity as a user for financial statements.

Their answers were recorded and any other opinions on the subject matter were also

included. All responses were then analysed and descriptive statistics derived.

4.3.1.1 Demographics

Of the ten organizations originally contacted, two companies declined to participate

in the phone survey. One company state that they have a company policy that

restricts them from participating in surveys while the other asserts that they have no

involvement in the accounting issues relating to financial instruments. Therefore, a

total of eight phone surveys are conducted from the initial list. Three more phone

surveys are conducted following recommendations from some initial respondents. In

total, there are 11 respondents to the first phone survey.

Seven users are from the assurance and advisory category. Four financial analysts

with experience in the financial institutions industry ranging from 5 to 20 years.

Table 5.1 details the eleven respondents’ level of business experience. A two-sample

t-test shows no difference in the experience of the two groups (p = 0.875).

95
Table 4.3: Characteristics of Phone Survey Respondents

Respondent User Category Years of experience

AUD1 Auditor/Assurance and Advisory 15
AUD2 Auditor/Assurance and Advisory 10
AUD3 Auditor/Assurance and Advisory 21
AUD4 Auditor/Assurance and Advisory 6
AUD5 Auditor/Assurance and Advisory 20
AUD6 Auditor/Assurance and Advisory 5
AUD7 Auditor/Assurance and Advisory 6
ANA1 Financial Analyst 10
ANA2 Financial Analyst 7
ANA3 Financial Analyst 12
ANA4 Financial Analyst 16
Average Level of Business Experience = 11.63 years
Legend: AUD = Auditor; ANA = Analyst.

4.3.1.2 Questions and Responses

The users were first asked to indicate the model that is most relevant to them in

relation to accounting for all financial instruments between three choices: historical

cost accounting, fair value accounting or a mixed measurement model. The users are

then segregated into auditors and analysts. Interestingly, all auditors surveyed prefer

fair value accounting for all financial instruments. This is consistent with the

JWGSS position. However, analysts still indicate preference for the current mixed

measurement model (the JWGBA stance). Table 4.4 shows the responses received

for the two groups of users.

Table 4.4: Relevance of Fair Value Accounting for all Financial Instruments
Section A: Support for fair value accounting
Question 1: Which of the following accounting model is most relevant to you?
Respondent Historical Cost for all Fair value accounting for Mixed measurement
financial instruments all financial instruments approach
Auditors (n=7) 0 7 (64%) 0
Analysts (n=4) 0 0 4 (36%)

96
The Fisher’s Exact Test reveals a clear difference (p-value = 0.000) in preferences

between auditors and financial analysts. All auditors indicate preference for the fair

value accounting model for ALL financial instruments. Conversely, all four

financial analysts prefer the mixed measurement model currently being utilised by

financial institutions.

The latter position may be due to the financial analysts’ preference to maintain the

status quo so that their knowledge base remains valuable to the investor class. In

other words, they may have developed techniques to adapt the current level of data to

give them a competitive advantage in the market place. The auditors’ preference for

fair value accounting is more problematic. These views are not generally consistent

with their clients and the use of full fair value accounting leads to complex

attestation issues. Perhaps they see the move towards fair value accounting as

unstoppable and are acknowledging the new reality. However, their support for fair

value accounting is based on their perception that it is the most relevant (i.e. better

reflect actual values) method for all financial instruments.

Further conversations indicate that financial analysts do not use the fair values

provided in financial reports because they lack transparency due to the dependence

on valuation models. They do, however, focus on marked-to-market information

provided in the disclosure sections. The general consensus among the analyst group

is fair value disclosure is more beneficial than fair value accounting for financial

instruments because the recognition of changes in fair values in the Income

Statement raises other problems. For example, one respondent suggests that taking

changes in fair values to the Income Statement adds another layer of ‘noise’ to

investors. One bank analyst also suggests that banks may not want to hold financial

97
assets because of the volatility in profitability that would result from the inclusion of

unrealised changes in fair values.

Conversely, the auditors signal strong support for the fair value accounting model.

However, one auditor indicates that the relevance of fair value accounting for all

financial instruments depends on the use of the information and the need to have a

consistent treatment for hedges. He also stresses that there are significant issues

involved in deriving fair values for non-traded instruments and the treatment in the

financial statements of a change in the credit rating of an entity in terms of the

valuation of its liabilities. One auditor admits that she was not completely familiar

with the proposed changes and voices concern on potential information overload that

may prove confusing to general users. She stresses the need for consistency in

accounting treatment and any changes must be clearly explained and appropriately

disclosed. Another auditor even suggests that fair value accounting should be used

for all items, not just financial instruments.

The next set of questions asks the user the extent he/she agrees or disagrees with four

different statements concerning the relevance, reliability, decision-usefulness and the

understandability of fair value accounting for all financial instruments. Another

statement asks for the extent the user agree or disagree with the separate

identification of fair values and historical cost where financial instruments are

measured using the mixed measurement model. Responses to these statements, the

average response score and the p-values from the two-sample t-test to compare the

two user groups’ responses are listed in Tables 4.5 and 4.6.

98
Table 4.5: Qualitative Characteristics of Fair Value Accounting
Section B: Qualitative characteristics of fair value accounting
Statement 1: Fair value accounting for all financial instruments provides relevant
information to you. (p = 0.068 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=7) 0 0 0 4 (37%) 3 (27%)
Analysts (n=4) 0 0 1 (9%) 3 (27%) 0
Total (mean = 4.2)* 0 0 1 (9%) 7 (64%) 3 (27%)
Statement 2: Fair value accounting for all financial instruments provides reliable
information to you. (p = 0.200 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=7) 0 0 1 (9%) 5 (46%) 1 (9%)
Analysts (n=4) 0 0 2 (18%) 2 (18%) 0
Total (mean = 3.8)* 0 0 3 (27%) 7 (64%) 1 (9%)
Statement 3: Fair value accounting for all financial instruments provides useful
information for your economic decision-making. (p = 0.074 for the difference
between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=7) 0 1 (9%) 0 4 (37%) 2 (18%)
Analysts (n=4) 0 2 (18%) 1 (9%) 1 (9%) 0
Total (mean = 3.5)* 0 3 (27%) 1 (9%) 5 (46%) 2 (18%)
Statement 4: You understand fair value accounting for all financial instruments.
(p = 0.037 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=7) 0 0 0 2 (18%) 5 (46%)
Analysts (n=4) 0 2 (18%) 0 1 (9%) 1 (9%)
Total (mean = 4.2)* 0 2 (18%) 0 3 (27%) 6 (55%)
* Average score based on a scale with 1 = Strongly Disagree to 5 = Strongly Agree

All but one user agrees that fair value accounting for all financial instruments

provides relevant information. According to the two-sample t-test, auditors (as

compared to analysts) tend to find fair value accounting for all financial instruments

to be more relevant (p-value = 0.068). Eight respondents agree that fair value

accounting provides reliable information and there are no significant difference in the

average response of the two user groups (the remaining three were neutral). Seven

99
users tend to agree that fair value accounting for all financial instruments is useful

for economic decision-making. The auditor group tend to find it more useful (p-

value = 0.074) but neither group unanimously agree with statement 3.

Most respondents strongly agree that they understand fair value accounting for all

financial instruments but two respondents admit that they need to upgrade their

knowledge to fully understand fair value accounting. Auditors indicate that they tend

to understand fair value accounting significantly more than the analysts (p-value =

0.037). The analysts are split.

As shown in Table 4.6, ten respondents agree that when some financial instruments

are measured at historical cost while others are measured at fair value, they should be

separately identified. The remaining respondent is neutral because he does not

support the mixed model at all. Overall, there is strong support for separate

identification of financial instruments measured using the two difference bases if the

mixed measurement model is used.

Table 4.6: Users’ Perception of the Need for the Separation of Fair Value
Figures from Historical Cost Numbers
Section C: Separation of fair value from historical cost in the accounts
Statement 5: When some financial instruments are measured at historical cost
while others are measured at fair value, they should be separately identified in the
accounts. (p = 0.305 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=7) 0 0 1 (9%) 3 (27%) 3 (27%)
Analysts (n=4) 0 0 0 1 (9%) 3 (28%)
Total (mean = 4.5)* 0 0 1 (9%) 4 (36%) 6 (55%)
* Average score based on a scale with 1 = Strongly Disagree to 5 = Strongly Agree

In summary, there is generally support for all four propositions on understandability,

relevance, reliability and usefulness (with means between 3.5 and 4.2). Overall,

100
respondents of the first survey support or strongly support the use of fair value

accounting for financial instruments and agree with all of the other statements on

relevance, reliability, usefulness and understandability. Auditors are generally more

supportive than financial analysts. However, the phone surveys highlight clear

differences between analysts and auditor views.

4.3.2 The Second Phone Survey

The second phone survey is conducted in 2002, 10 months after the first phone

survey. This is after the Financial Reporting Council (FRC) announced support for

the adoption of International Accounting Standards (all standards issued from 1

January 2002 will be known as the International Financial Reporting Standards

(IFRS)) by Australia by 1 January 2005 (FRC, 2002). After this date, Australian

companies need to comply with IFRSs and their audit reports will attest to this

compliance (FRC, 2002). Consequently, the Australian Accounting Standards Board

(AASB) is obligated to work towards the full implementation of (and convergence

to) IFRSs (Alfredson, 2003). Therefore, this second phone survey gathers user

perspectives after this announcement to gather insights into whether such national

(and international) movements potentially influence user views.

A new list of Australian sophisticated users (distinct from the first phone survey)

who are major players in the debate of fair value accounting is compiled and then

contacted. Similar questions are asked with an additional question related to their

opinion on the reasons why users are not actively using opportunities to express their

thoughts and opinions. The following sections show the demographics and results of

the second phone survey.

101
4.3.2.1 Demographics

Table 4.7 details the eight respondents’ (of the second phone survey) experience in

financial reporting, which ranged from 7 to 35 years. There are three users in the

auditor group and five from the analyst group. Two-sample t-test results show no

significant difference in the experience between the two groups (p = 0.938).

Table 4.7: Characteristics of Respondents of the Second Phone Survey

Respondent User Category Years of

experience
AUD8 Auditor/Assurance and Advisory 25
AUD9 Auditor/Assurance and Advisory 18
AUD10 Auditor/Assurance and Advisory 25
CA1 Corporate Advisory 30
FM1 Fund Manager 24
ANA5 Financial/Investment Analyst 7
ANA6 Financial/Investment Analyst 35
ANA7 Financial/Investment Analyst 20
Average Level of Business Experience = 23 years
Legend: AUD = Auditor; FM = Fund Manager; ANA = Analyst; and CA = Corporate
Advisor.

4.3.2.2 Questions and Responses

The first question again asks the user to indicate the model that is most relevant to

them in relation to accounting for all financial instruments between three choices:

historical cost accounting, fair value accounting or a mixed measurement model.

Table 4.8 shows the responses received.

Table 4.8: Phone Surveyed Users’ Views on the Relevance of Fair Value
Accounting for all Financial Instruments
Section A: Support for fair value accounting
Question 1: Which of the following accounting model is most relevant to you?
Respondent Historical Cost for all Fair value accounting for Mixed
financial instruments all financial instruments measurement
Auditors (n=3) 0 3 (38%) 0
Analysts (n=5) 0 4 (50%) 1 (12%)

102
This time, all but one user prefers fair value accounting for all financial instruments.

Almost all of the sophisticated users interviewed prefer fair value accounting for

ALL financial instruments. Only one user (ANA7) prefers the mixed measurement

model currently being utilised by financial institutions. The two-sample t-test

reveals a p-value of 0.482. Thus, there is initial evidence20 that the views of the two

user groups have shifted towards fair value accounting for all financial instruments

since the first phone survey. However, the second group is also more experienced

and perhaps that could be the explanation for the shift in preference.

Conversation with ANA7 indicates that analysts are predominantly interested in (and

want to understand) the “sustainable picture” of a financial institution and historical

cost is relevant. He feels that fair value accounting in financial statements will still

be out of date as the statements are produced 12 weeks after the financial year-end,

therefore rendering them irrelevant. However, seven of the eight users are of the

opinion that fair value information is necessary and relevant.

The next set of questions ask the user the extent he/she agrees or disagrees with four

different statements on the relevance, reliability, decision-usefulness and the

understandability of fair value accounting for all financial instruments. Another

statement then asks if the user agrees or disagrees that the mixed measurement model

is useful for their economic decision-making. Responses to these statements are

listed in Table 4.9 and 4.10 separated into the two user groups. However, further

statistical analyses indicate that there is no significant difference in the preferences of

the two groups with p-values ranging from 0.116 to 0.938.

20
The sample size is small and generalization is problematic.

103
Table 4.9: Phone Surveyed Users’ Views on the Qualitative Characteristics of
Fair Value Accounting for all Financial Instruments
Section B: Qualitative characteristics of fair value accounting
Statement 1: Fair value accounting for all financial instruments provides relevant
information to you. (p = 0.116 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=3) 0 0 0 0 2 (25%)
Analysts (n=5) 0 0 0 3 (37%) 3 (38%)
Total (mean = 4.6)* 0 0 0 3 (37%) 5 (63%)
Statement 2: Fair value accounting for all financial instruments provides reliable
information to you. (p = 0.260 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=3) 0 0 0 1 (13%) 2 (25%)
Analysts (n=5) 0 1 (13%) 0 3 (37%) 1 (12%)
Total (mean = 4.1)* 0 1 (13%) 0 4 (50%) 3 (37%)
Statement 3: Fair value accounting for all financial instruments provides useful
information for your economic decision-making. (p = 0.116 for the difference
between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=3) 0 0 0 0 3 (37%)
Analysts (n=5) 0 0 0 3 (37%) 2 (25%)
Total (mean = 4.6)* 0 0 0 3 (37%) 5 (63%)
Statement 4: You understand the implications of fair value accounting for all
financial instruments. (p = 0.339 for the difference between auditors and analysts)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=3) 0 0 0 (0%) 1 (12%) 2 (25%)
Analysts (n=5) 1 (12%) 0 1 (12%) 1 (12%) 2 (25%)
Total (mean = 4.2)* 1 (12%) 0 1 (12%) 2 (25%) 4 (50%)
* Average score based on a scale with 1 = Strongly Disagree to 5 = Strongly Agree

As shown in Table 4.9, there is a high level of support for the four statements (at

least 3 users have strong support for all of the statements). All users agree (five

strongly) that fair value accounting for all financial instruments provides relevant

information. Only one user thinks fair value accounting provides unreliable

information due to the underlying assumptions used in the calculations, while the

remaining seven agree that it provides reliable information. All eight users agree on

104
the usefulness of fair value accounting for all financial instruments for economic

decision-making. Although six of the users understand the implications of fair value

accounting for all financial instruments, FM1 indicate disagreement with the

statement (even with 24 years experience). Overall, there are no significant

differences of views from the two groups.

Table 4.10: Phone Surveyed Users’ Views on the Usefulness of the Mixed
Measurement Model
Section C: Decision-usefulness of the mixed measurement method
Statement 5: The current mixed measurement approach for financial instruments
provides useful information for your economic decision-making. (p = 0.582)
Respondent Strongly Disagree Neutral Agree Strongly
Disagree Agree
Auditors (n=3) 0 1 (12%) 0 2 (25%) 0
Analysts (n=5) 1 (12%) 1 (12%) 1 (12%) 2 (25%) 0
Total (mean = 3.0)* 1 (12%) 2 (25%) 1 (12%) 4 (50%) 0
* Average score based on a scale with 1 = Strongly Disagree to 5 = Strongly Agree

This question is included as a check that respondents are not just agreeing with

everything. Interestingly, even though there is general support for fair value

accounting, only three users are of the opinion that the current mixed measurement

model is not useful for decision-making. One respondent adopts a neutral position

but the other four think that the current mixed measurement model provides useful

information for their economic decision-making.

There is also a final question asking for the users’ thoughts on the low response rate

for a separate mail survey sent to sophisticated users even though that provides them

the opportunity to make their voices heard (see Appendix G). A few respondents cite

lack of knowledge and time as the reason for lack of participation. However, the

biggest concern is the perception that their (users’) voice would not be heard.

105
In summary, ten months later, users show strong support for fair value accounting

and all four propositions on understandability, relevance, reliability and usefulness

(with means between 4.1 and 4.6). Although these respondents have similar views to

those from the first phone survey, there is higher support for the usefulness of fair

value accounting information for decision making. However, a minority of

respondents (37%) still perceive the current mixed measurement model to be useful.

This time around, there are no statistical differences between the two user groups.

4.4 PRELIMINARY CONCLUSIONS

This preliminary evidence is based on telephone surveys of only 19 individuals and

therefore there are limitations in generalising from the results to the broader

population of sophisticated users. The phone survey approach provides deeper

insights into the actual views of the user group, a constituent that has been largely

ignored in past research, by gathering direct, albeit preliminary evidence.

Both the JWGBA and the JWGSS espouse views about their perceptions of users’

needs with limited support of truly compelling evidence. This study gathers the

views of sophisticated users’ in an interactive environment where issues are clarified

and additional insights and follow-ups obtained. Thus, it provides more in-depth

and, arguably, richer evidence than what could be obtained through other techniques.

For example, one respondent points out that fair values are only as reliable as the

assumptions that they are based on and the policies (and parameters) used as the

basis for fair value calculation are more important (and useful) for analysts to

understand the risk profile of the company. Other users voice some concern that fair

value accounting, without appropriate levels of disclosure, results in a distortion of a

company’s financial position and performance. Poon (2004) also argues for fair

106
value disclosures as the logical starting step before even contemplating a shift to fair

value accounting.

As indicated in the previous sections, despite the small sample size there seems to be

a shift in user preference towards fair value accounting for all financial instruments

over a crucial ten-month period. Users’ views may be converging in terms of their

information needs in relation to accounting for financial instruments by financial

institutions. There is a move towards more acceptance of fair value accounting

model (the JWGSS model). However, perhaps the extended level of experience of

the second group is the reason for a stronger preference for fair value accounting.

This experience factor is statistically examined for the mail surveys sent out to both

users and preparers (see Chapter 5 and Chapter 6 for details).

The first phone survey shows a significant difference between auditor and analyst

information needs whilst the second survey dos not. One possible explanation could

be the FRC decision on the adoption of IFRSs (and therefore the impending adoption

of IAS 39) caused this shift in users’ opinions. This certain move towards IAS 3921,

which prescribes fair value accounting for certain financial instruments, may be seen

as a fait accompli such that sophisticated users are beginning to resign themselves to

this new reality as highlighted by their perception that their voices would not be

heard (see Appendix G for details).

Users tend to prefer fair value accounting for all financial instruments and also

mostly agreed that it provides information that is relevant, reliable, understandable

and useful for decision-making purposes. This suggests overall support for the

21
The 2003 revised IAS 39 allow entities to use fair values for all financial instruments but stops
short of mandating its use for all financial instruments (see Chapters Two and Eight for further
details).

107
assertions made by the JWGSS, albeit on a weaker scale. A larger sample of users

would provide a better indication of the views of sophisticated users at large.

Discussions with the respondents also highlight the importance for further and more

detailed empirical research on the actual information needs of users and their

perceptions about fair value accounting for all financial instruments. The IASB

Framework implicitly assumes users to have at least a reasonable (if not

sophisticated) knowledge of business and economic activities and accounting.

Evidence from this thesis casts some doubt about this assumption. Some

respondents admit to a lack of understanding about the accounting issues with

financial instruments even though they are widely regarded as sophisticated users. It

should be noted that IAS 39 is indeed a very long, complex and arguably difficult

standard to comprehend.

4.5 SUMMARY

The in-depth insights gathered through this (exploratory) qualitative phase reveals

general consensus (except from preparers) for the use of fair value accounting for all

financial instruments for financial institutions with adequate levels of disclosure on

the principles and policies behind them. This result provides some support for the

JWGSS position as outlined earlier in this thesis. However, these preliminary

conclusions are based on a small sample and seem to indicate there is still variation

in the opinion of users and preparers alike as to the usefulness of fair value

accounting for all financial instruments. The following chapters detail the evidence

generated from a large sample through mail surveys of users and preparers in both

Singapore and Australia.

108
 CHAPTER FIVE:
QUANTITATIVE PHASE – PREPARER
VIEWS

5.1 INTRODUCTION

The qualitative phase reveals a general preference for fair value accounting for all

financial instruments in support of the JWGSS position. However, the interviews

also confirm some of the assertions made by the JWGBA such as the differentiation

between banking and trading books, the problems with earnings volatility and the

lack of a reliable measurement for certain financial instruments.

This chapter is the first of three analysing evidence gathered from the quantitative

phase. As detailed in Chapter Two, the quantitative phase involves two distinct

albeit connected groups in this debate in relation to support for fair value accounting

for all financial instruments. The survey evidence obtained from each group

(preparer and user) is provided in separate chapters (see Figure 3.5).

Here, insights obtained from the preparer surveys from both Australia and Singapore

are provided. Chapter Six deals with the user group from both countries and Chapter

Seven provides the comparison between preparer and user perceptions.

5.2 SURVEY PARAMETERS

The pilot-tested22, refined and shortened preparer survey is sent to chief financial

officers in Australia and Singapore (see Appendix A, F and H for details on the

22
The results of the pilot study are detailed in Appendix H.

109
changes made between the pilot tested survey and the final mail survey

disseminated). The only difference in the two final surveys sent to Australia and

Singapore preparers is the use of terminology for the Income Statement and Balance

Sheet. Statement of Financial Performance and Statement of Financial Position are

the terminology used in the Australian version of the survey as compared to the

Profit and Loss and Balance Sheet labels respectively in the Singaporean version.

A total of 401 preparer surveys were sent to chief financial officers in financial

institutions in Australia and Singapore. 201 preparer surveys were sent to chief

financial officers of 99 Australian financial institutions and corporate treasuries23 and

102 credit unions and building societies. However, 18 surveys are returned due to

errors in addresses or a company having moved premises. 41 completed surveys are

received resulting in an effective response rate of 22.4%. The Singaporean-adapted

survey is sent to 200 chief financial officers of financial institutions. An aggregate of

42 responses are received giving a 21% response rate.

The overall 21.7% response rate is relatively low. Though so, this low response rate

is consistent with the response rate of many studies in the social sciences (see Griffis,

Goldsby and Cooper, 2003 and Chiu and Brennan, 1990). However, due to the

nature of the industry, the profession and the inherent complexity of the topic, this

response rate is largely beyond the control of the researcher and not a result of a

flawed research study design (see also Appendix I and Section 5.3). The response

provides a broader coverage of the industry as compared to other studies, such as

Fargher (2000) that looks at only Australian financial institutions.

23
Corporate and government treasuries were also included as they are some of the important players
in the financial instrument market as per the Fargher (2001) study.

110
Nevertheless, generalisation of the evidence to the financial industry should be taken

cautiously. This thesis’s strength lies in the ability to provide insights into the views

of financial institutions from a broader spectrum not just the Big-Four banks in

Australia and their counterparts in Singapore. An especially important contribution

is the perceptions of preparers in the credit unions. This is important considering that

the Big-Four Australian Banks are represented on the JWGBA, and Singaporean

financial institutions are not represented at all. Overall, the survey results add

breadth on the perception of the financial institutions’ community on fair value

accounting for financial instruments issues in Australia and Singapore.

5.3 TESTING NON-RESPONSE BIAS

Chapman (1988) points out that making the assumption that non-response bias does

not exist is one of the seven deadly sins of survey research. The relatively low

response rate in this study signals possible non-response bias from the remaining

80% of preparers. Appendix I shows the general linear models conducted to test the

effect of response time on preparers’ experience and overall preference for fair value

accounting. The results show that response time is not a determinant of preparers’

preference for fair value accounting. However, for every 1-unit increase in response

(i.e. slower respondents), preparer experience decreases by 0.025 (p-value = 0.047).

This means that the last respondents are just one category of experience lower. Here,

experienced preparers tend to respond quicker and this suggests that experienced

preparers are more likely to respond to the survey. Thus, non-response bias may

exist in that the results are less relevant to inexperienced preparers (arguably, the

views of inexperienced preparers are less of an interest in this thesis due to the

111
complexity of the topic of fair value accounting). Therefore, non-response bias is not

considered a problem for the key issue of preference for fair value accounting.

The following subsections provide details of the type and size of assets of financial

institutions that responded and the level of experience of the respondents.

5.4 PREPARER DESCRIPTIVES

5.4.1 Type of Financial Institution by Country

Different categorisations of the type of financial institutions are used in the two

surveys due to the differing nature of financial institutions classification common to

Australia and Singapore, respectively. The categorisation is done based on the

classification listed for each country based on the Australian Prudential Regulatory

Authority (APRA) and Monetary Authority of Singapore’s (MAS) websites. The

nine different Australian categories and the four Singaporean ones are collapsed into

four common groups of local bank, foreign bank, credit union/building society and

the “other” category for the handful of merchant banks, treasuries and finance

companies. See Appendix A and C for full list of categories used.

Table 5.1 depicts the number of respondents from each category of financial

institution by country and in total.

Table 5.1: Breakdown of Type of Financial Institutions

Type of Financial Institution Australia Singapore Total Percent (%)

Local bank 5 11 16 19.3
Foreign bank 3 21 24 28.9
Credit Union/Building Societies 23 0* 23 27.7
Other financial institutions 10 10 20 24.1
Total 41 42 83 100.0
* There are no credit unions in Singapore.

112
The 83 respondents represent the width of the financial institutions industry in

Australia and Singapore. Of the 83 surveys that are received there are 16 from local

Australian and Singaporean banks24, 24 foreign banks, 23 credit unions/building

societies, and 20 other financial institutions.

No credit unions are sampled from Singapore because there are none in Singapore.

Although a small country, Singapore has a higher number of local and foreign banks

being a major financial focal point in the South East Asian region (International

Monetary Fund, 2004). The remaining 20 other financial institutions consist of

treasuries from Australia, as well as merchant banks and finance companies from

both countries.

In the sample, the fact that banks tend to be in Singapore and credit unions in

Australia is a form of collinearity (see Table 5.1) and care must be exercised when

interpreting regression models with one of these variables but not the other since

country and type of financial institution are partial surrogates for each other.

5.4.2 Size of Financial Institution by Country

The respondent financial institutions range from those with an asset base of less than

$1 million to large financial institutions with more than $100 billion of assets.

Although the surveys distinguished between the Australian and Singaporean dollar,

the exchange rates between these two currencies are highly stable at 1:125.

24
Although there are only five local banks in Singapore, the surveys were completed by branch
managers. The 11 surveys were from branches of the five local Singaporean banks. This thesis is
concerned with preparers’ opinions and although it may be argued that those from the same banks
may have the same opinion, it is not expected that it will have that big an effect on the overall
findings. Since respondents to the survey are anonymous, we do not know which respondents
work for the same bank.
25
The average exchange rate during the survey (1st May 2002 to 2nd September 2002) is 1:0.98069 as
per the OANDA currency conversion tool available on http://oanda.com/convert/fxhistory.

113
Therefore, no distinction (and conversion) of the currencies are made in this thesis.

This breakdown of the ‘size of financial institution’ is shown in Table 5.2.

Table 5.2: Size of Respondent Financial Institutions

Assets ($) Australia Singapore Total Percent (%)

< 1 million 0 7 7 8.4
1 to 10 million 2 11 13 15.7
10 to 100 million 14 3 17 20.5
100 million to 1 billion 6 4 10 12.0
1 to 10 billion 12 11 23 27.7
10 to 100 billion 6 4 10 12.0
> 100 billion 1 2 3 3.7
Total 41 42 83 100.0

The 83 financial institutions range from seven with less than $1 million asset base to

the three that have assets in excess of $100 billion. There were 20 small respondents

who have less than $10 million worth of assets and the remaining 27 are identified as

medium size financial institutions that have assets between $10 million and $1

billion. Overall, there is a good distribution of size of financial institutions in both

countries.

5.4.3 Level of Experience by Country

Preparer respondents are asked to indicate their level of experience from five ranges,

less than 2 years, 2 to 5 years, 6 to 10 years, 11 to 15 years and more than 15 years.

A wide range of preparers responded to the survey from various levels of experience

in both Australia and Singapore. This is shown in Table 5.3.

114
Table 5.3: The Level of Experience of Preparer Respondents

Australia Singapore Total Percent (%)

< 2 years 1 6 7 8.4
2 - 5 years 2 14 16 19.3
6 - 10 years 6 11 17 20.5
11 - 15 years 8 5 13 15.7
> 15 years 24 6 30 36.1
Total 41 42 83 100.0

Approximately three quarters of respondents have more than six years of experience

while thirty of the respondents have more than 15 years experience. The vast

majority of respondents have extensive experience in financial reporting but

Australian preparers in the sample tend to be more experienced (independent samples

T-test revealed p-value = 0.044). Here, experience and country are partial surrogates

for each other and any experience effect may be due to country effect and vice versa.

The consequences of this relationship between country and experience are

considered in the discussion of the results of the regressions.

5.4.4 Level of Experience by Size

Tabachnick and Fidell (1996) note that harmful levels of multicollinearity exist when

the bivariate correlation between independent variables are high (0.9 and above).

The correlation between the only two quantitative independent variables of size and

level of experience for the preparer survey shows a low absolute correlation of 0.375

( p-value = 0.000). Not surprisingly, this indicates that preparers in larger financial

institutions are more experienced than their counterparts. This correlation, although

statistically significantly different to zero, is considered low and not a collinearity

problem. Nevertheless, it should be remembered that any difference between the two

countries could be a result of Australian preparers tending to be more experienced.

115
5.4.5 Level of Experience by Type of Financial Institution

A one-way ANOVA is used to examine if average experience is significantly

different for each type of financial institution. The results of this analysis reveals a

p-value of 0.000, indicating a strong statistical relationship between level of

experience and type of financial institution. The descriptives as laid out in Table 5.4

show the average level of experience (based on the five categories detailed in Section

5.4.3) for the preparers from each type of financial institution.

Table 5.4: Breakdown of Level of Experience by Type of Financial Institution

Type of Financial Institution Average Experience Std. Dev. Total

Local bank 3.15 1.24 16
Foreign bank 3.42 1.41 24
Credit Union/Building Societies 4.48 0.89 23
Other financial institutions 3.15 1.35 20
Total 3.52 1.37 83

Preparers from credit unions and foreign banks tend to be more experienced relative

to those from local banks and other financial institutions. It seems that level of

experience is a partial surrogate for type of financial institution and thus, any

experience effect may be due to type of financial institution and vice versa.

Similarly, this relationship is considered in the discussion of the results of the

regressions.

5.4.6 Size by Type of Financial Institution

The relationship between these two independent variables is tested using a one-way

ANOVA. A p-value of 0.143 signals a lack of statistical relationship between these

two variables. Therefore, the size of the financial institution is not related to its type

and there is no collinearity between these two variables.

116
The following sections details their views on the key themes identified in this crucial

international debate as well as their views on other prominent issues asked in the

preparer survey.

5.5 CURRENT USE OF FAIR VALUE ACCOUNTING

The ‘fair value accounting’ term is used in this research to refer to the measurement

of all financial instruments at fair value and the recognition of changes in fair value

as revenues or expenses in the Income Statement in the period in which they arise

(ie. the JWGSS proposed measurement model). The global controversy raises

various issues. Before those issues are examined, it is worth exploring the extent of

the current use of fair value accounting for different categories of financial

instruments. In the preparer survey, chief financial officers from both Australia and

Singapore are asked to indicate the extent that fair value accounting was currently

used for different financial instruments to determine its current level of use (shown in

Table 5.5).

Table 5.5: The Extent Fair Value Accounting is Currently Used by Preparers
for Different Categories of Financial Instruments

Category n* Type of Financial Instruments Never Sometimes Always

Trading 54 Trading Derivatives 10 9 35

55 Trading Securities 9 9 37

Banking 54 Hedging Derivatives 21 20 13

70 Investment Securities 26 21 23

77 Loans and Receivables 34 18 25

60 Other Financial Assets 23 17 20

60 Other Financial Liabilities 25 15 20

* Respondents that do not have a particular type of financial instrument are excluded
where n is less than 83.

117
Further examination reveals smaller financial institutions to be the respondents

without some of the categories of financial instruments. It seems that some

respondents do not use fair value accounting for financial instruments typical in the

trading book (n equals 10 and 9 for trading derivatives and trading securities,

respectively). Across the breakdown of the categories, the majority of respondents

always use fair value accounting for trading financial instruments, consistent with the

JWGBA opinion. However, at least nine financial institutions do not use fair value

accounting for trading financial instruments contrary to the JWGBA arguments.

Table 5.5 shows that at least 13 respondents have employed fair value accounting for

certain financial instruments in the banking book, contradicting the JWGBA

position. The “Always” column shows at least 24% of the respondents use fair value

accounting for hedging derivatives, investment securities and for loans and

receivables. One of the major arguments of the JWGBA is related to the problematic

volatility introduced into earnings as a result of fair value accounting.

However, the survey evidence shows there are preparers already using fair value

accounting in the banking book. This may indicate that perhaps the volatility in

earnings concern expressed by the JWGBA is not shared by all financial institutions.

Therefore, the perceived volatility does not seem to be considered sufficiently

problematic to prevent some preparers from using fair value accounting. Responses

also show that the majority of financial institutions tend to use fair value accounting

for trading financial instruments rather than for those in the banking book. However,

this one-sided preference is not as extreme as asserted by the JWGBA.

Therefore, the trend of initial exploration of the level of use of fair value accounting

for financial instruments is generally as predicted but not as exaggerated as

118
advocated by the JWGBA. The fact that fair value accounting is used for financial

instruments in the banking book by some preparers lends weight to the JWGSS

proposals. Moreover, there are some financial institutions that do not use fair value

accounting for the trading book.

The following section provides evidence on preparer views in relation to the four

main controversial issues.

5.6 KEY THEMES AND POSSIBLE INFLUENCES

The four main themes identified in Chapter 4 form the focus of the quantitative

phase. The following subsections detail the results obtained from the 83 surveys in

relation to preparer opinions on the four key themes of fair value accounting, as

determined in Section 4.2.6. Potential determinants for each theme are examined via

multiple regressions by using the four preparer descriptors of country, type of

financial institution, size and level of experience as shown in Figure 5.1.

Figure 5.1: Possible Predictors of Preparer Views on Key Themes

Possible Predictors Key Themes

Country 1. Distinction between trading and

banking book
Type of financial 2. Qualitative characteristics of fair
institution value accounting information
3. Volatility that results from fair
Company Size value accounting
4. Preference for fair value accounting
Experience

119
The two-way interaction between country and the three other independent variables

are examined with regressions because the relationship between the themes and these

independent variables may be expected to differ between the two countries due to

inherent differences in the standard setting tendencies. Responses to the first three

themes: (1) distinction between trading and banking book, (2) problems with

earnings volatility, and (3) qualitative characteristics inherent in fair value

accounting, are also included as possible descriptors of preparer preference for fair

value accounting in Section 5.9.

5.7 DISTINCTION BETWEEN TRADING AND

BANKING

As detailed in Chapter Two, the main arguments against fair value accounting put

forward by the JWGBA was grounded on the purported fundamental difference

between the trading and the banking book. On the other hand, the JWGSS viewed

that the distinction (and related accounting options) allowed far too much latitude for

managers to manage earnings related to financial instruments. Therefore, one

question in the survey examines the extent to which respondents agree (or disagree)

with the assertion that the trading and banking books are sufficiently different

enough to warrant the use of different accounting measurement bases.

The responses of strongly disagree, disagree, neutral, agree and strongly agree were

replaced with the values 1, 2, 3, 4 and 5 respectively and a one-sample t-test used to

compare the means to the neutral response. This statistical technique is used

throughout this chapter and Chapters Six and Seven.

120
Table 5.6: Level of Preparers’ Support for the Statement on the Fundamental
Difference between the Trading and Banking Books

Disagree

Disagree
Strongly

Strongly

p-value
(n=83)
Mean
Neutral
Question

Agree

Agree
The banking book is so
fundamentally different from
the trading book to warrant the 7 20 16 29 11 3.2 .123
use of different accounting
measurement bases.
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

The p-value of 0.123 indicates the average preparer response (mean =3.2) is not

significantly different from the neutral (test value = 3) position. This finding

suggests neutrality of preparer views in relation to the purported fundamental

difference between the two books to warrant the use of different accounting

measurement bases.

This conclusion is confirmed with the following non-parametric test. If the 16

preparers who are neutral on this issue are excluded resulting in a sample of only 67,

almost 60% (40 out of 67) show support for the use of different measurement bases

for the two books. A one-sample proportion test reveals that this 60% is still not

significantly different from 50% (p-value of 0.143). Thus, there is no statistically

significant evidence that a majority of preparers agree with the fundamental

difference between trading and banking books. It is worth noting that while

preparers are neutral concerning the fundamental difference between the trading and

banking book on average, individual preparers express a variety of opinions with

only 16 out of 83 are in fact neutral.

121
5.7.1 Regression Results

The four preparer descriptors of country, type of financial institution, size and level

of experience are then fitted in general linear models to examine if any of them are

predicts preparers’ perception of the distinction between trading and banking books.

The two variables of size and level of experience are treated as covariates26 in all the

models, while the remaining two variables are fixed factors27. In the initial model,

the four possible descriptors are included with the inclusion of the two-way

interaction effects between country and each of the other three predictors to account

for possible country effects on the other three predictors. The backward elimination

is conducted by eliminating insignificant variables starting from the one with the

highest p-value to find the best regression estimates (Hair, et. al., 1998). Residual

plots for each model fitted are checked to ensure underlying assumptions are met.

Table 5.7 shows the initial general linear model fitted followed by the models

resulting from the backward elimination process.

Table 5.7: Regression Results for Preparers’ Perception of the Distinction

between the Trading and Banking Book

Initial Model
Variables in the equation Level of Significance
Country 0.171
Type of Financial Institution 0.009***
Size 0.408
Level of Experience 0.611
Country * Type of Financial Institution 0.004***
Country * Size 0.339
Country * Level of Experience 0.673
R-square = 0.242; Adjusted R-square = 0.137; p = 0.021

26
Covariate is an SPSS term for a continuous variable.
27
In SPSS, fixed factor is a label for nominal variables.

122
Table 5.7: Regression Results for Preparers’ Perception of the Distinction
between the Trading and Banking Book (continued)

Level 1 – Country * Level of Experience deleted

Country 0.180
Type of Financial Institution 0.009***
Size 0.440
Level of Experience 0.135
Country * Type of Financial Institution 0.003***
Country * Size 0.291
R-square = 0.240; Adjusted R-square = 0.147; p = 0.013
Level 2 – Country * Size deleted
Country 0.300
Type of Financial Institution 0.009***
Size 0.796
Level of Experience 0.494
Country * Type of Financial Institution 0.004***
R-square = 0.229; Adjusted R-square = 0.145; p = 0.010
Level 3 – Size deleted
Country 0.307
Type of Financial Institution 0.007***
Level of Experience 0.522
Country * Type of Financial Institution 0.002***
R-square = 0.228; Adjusted R-square = 0.156; p = 0.006
Level 4 – Level of Experience deleted
Country 0.156
Type of Financial Institution 0.007***
Country * Type of Financial Institution 0.002***
R-square = 0.224; Adjusted R-square = 0.162; p = 0.003

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The resultant models have limited predictive power with the final model having the

highest predictive power of 16.2% (adjusted R-squared). The significant interaction

implies both country and type of financial institution are significant predictors (see

Section 5.7.1.1 below). P-values for significant variables (in this case - type of

financial institution) that constitute part of a significant interaction variable (here,

country by type of financial institution) are not considered due to the difficulty in

123
correctly interpreting a model with an interaction term but without the factors

involved in the interaction. However, as stated in Section 5.4, level of experience

seems to be a partial surrogate for country as well as type of financial institution. In

this situation, level of experience has the most insignificant p-value at level 3 and

thus, country and type of financial institution are the more significant influences.

5.7.1.1 Country By Type of Financial Institution Effect

The influence of the interaction between country and type of financial institution is

depicted by the difference between countries depending on the type of financial

institution. For example, preparers’ support for the distinction between the two

books are quite different between Australia and Singapore if they are from a local

bank (a mean 4.60 as compared to 3.27). This is shown in Table 5.7 and Figure 5.2.

Table 5.8: Preparers’ Perception of the Distinction between the Trading and
Banking Book by Country by Type of Financial Institution
Country n Type of Financial Institution Mean Std. Error
Australia 5 Local Bank 4.60 .490
3 Foreign Bank 4.33 .633
23 Credit Union 3.30 .228
10 Other Financial Institution 2.10 .346
Singapore 11 Local Bank 3.27 .330
21 Foreign Bank 3.10 .239
10 Other Financial Institution 3.20 .346
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

Furthermore, there is substantial variation in preparers’ views if they originated from

Australia as compared to the small variation in the views of Singaporean preparers.

124
Figure 5.2: Australian and Singaporean Preparers’ on the Distinction between
the Trading and Banking Book

Average Response 5

4
Australia
3
Singapore
2

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

It seems that local and foreign banks in Australia are the biggest supporters (mean of

4.6 and 4.3) of using different accounting measurement bases for the banking and

trading book due to the fundamental difference between the two. This is consistent

with JWGBA assertions as local and foreign banks are represented on the JWGBA.

Australian preparers support the use of two different measurement methods more

than their Singaporean counterparts. There is also an indication that the support for

the distinction between the banking and trading books by local and foreign banks in

Australia is significantly different than their Singaporean equivalents.

Table 5.7 and Figure 5.2 also show that Singaporean preparers have similar average

responses tending around the neutral position regardless of the type of financial

institution. This could perhaps be due to the fact that the country follows

International Financial Reporting Standards or (International Accounting Standards)

more closely, and thus, the preparers have less resistance (or support for that matter)

towards the new proposed standard. The results also indicate that other financial

institutions in Australia (two merchant banks, four treasuries and four other financial

institutions) tend to oppose the differentiation of measurement bases between the two

125
books more than the other financial institutions in Singapore. This is rather curious

as the majority of comment letters from Australian treasuries received by the IASB

in response to the draft standard stressed the practical difficulties in valuing illiquid

instruments (i.e. those classified as financial instruments in the banking book).

5.8 QUALITATIVE CHARACTERISTICS

The support (and/or resistance) towards fair value accounting is further examined

with questions that assess the perceived usefulness of fair value accounting in

relation to the qualitative characteristics possessed by fair value information

according to the IASB Framework (2001). Table 5.9 shows preparer perceptions of

the relevance and comparability of information resulting from fair value accounting.

Table 5.9: Level of Preparers’ Perception of the Relevance and Comparability

of Fair Value Accounting
Disagree
Disagree
Strongly

Strongly

p-value
Means
Neutral

(n=83)
Agree

Agree

Fair value accounting improves

the relevance of the information 2 8 15 39 19 3.8*** .000
in the accounts for users.
Fair value accounting promotes
the comparability of Balance 3 12 17 37 14 3.6*** .000
Sheets between organisations.
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.
*** Highly Significant at the 0.01 level

Respondents tend to agree with the JWGSS logic that fair value accounting improves

the relevance of information for users and promotes comparability with average

responses of 3.8 and 3.6 respectively. These average scores are statistically highly

significantly different from a neutral response. Responses to these two questions are

also statistically highly correlated (r = 0.754, p = 0.000). Exclusion of neutral

126
responses results in 85% and 76% of preparers agreeing that fair value accounting

improves relevance and comparability, respectively, both statistically different from

50% (p = 0.000 in both one-sample proportion tests).

In order to assess the reliability of fair value accounting, the respondents are asked to

rate the reliability of the fair values of traded and non-traded financial instruments

(see Table 5.10). Responses show that the fair values of traded financial instruments

are generally considered reliable (mean = 4.0). Exclusion of the 11 preparers with

neutral opinions reveals 92% of preparers perceiving the fair values of traded

financial instruments to be highly reliable (p = 0.000).

Table 5.10: Level of Preparers’ Perception of the Reliability of Traded and

Non-traded Financial Instruments
Unreliable

Unreliable

p-value
Means
(n=83)
Reliable

Reliable
Neutral

Type of financial instrument

Highly

Highly

Traded financial instruments 1 5 11 39 27 4.0*** .000

Non-traded financial instruments 6 22 29 22 4 3.0 .665

Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

A mean of 3.0 indicates fair values of non-traded financial instruments are not

perceived to be reliable or unreliable, contrary to JWGBA assertions. Once the 29

preparers with no major opinion on the matter were excluded, only 48% of the

remaining preparers perceive these fair values to be reliable, which is not

significantly different to 50% (p = 0.892). This indicates a lack of consensus on the

achievable reliability of fair values for financial instruments in the banking book.

127
The following sections further examine these qualitative characteristics and possible

influence by country, size, type of financial institution and level of experience

together with the two-way interactions between country and the other three variables.

5.8.1 Regressions for Relevance of Fair Value Accounting

The four variables of country, size, type of financial institution and level of

experience together with the two-way interactions between country and the other

three variables are included in the initial general linear regression model to check for

predictors of preparer perception of the relevance of fair value accounting. The

models fitted are shown in Table 5.11.

Table 5.11: Regression Results for Preparers’ Perception of the Relevance of

Fair Value Accounting

Variables in the equation Level of Significance

Country 0.469
Type of Financial Institution 0.035**
Size 0.529
Level of Experience 0.587
Country * Type of Financial Institution 0.536
Country * Size 0.203
Country * Level of Experience 0.602
R-square = 0.194; Adjusted R-square = 0.082; p = 0.089
Level 1 – Country * Level of Experience deleted
Country 0.586
Type of Financial Institution 0.027**
Size 0.581
Level of Experience 0.598
Country * Type of Financial Institution 0.552
Country * Size 0.162
R-square = 0.191; Adjusted R-square = 0.092; p = 0.063

128
Table 5.11: Regression Results for Preparers’ Perception of the Relevance of
Fair Value Accounting (Continued)

Variables in the equation Level of Significance

Level 2 – Country * Type of Financial Institution deleted
Country 0.481
Type of Financial Institution 0.033**
Size 0.634
Level of Experience 0.685
Country * Size 0.123
R-square = 0.178; Adjusted R-square = 0.101; p = 0.034
Level 3 – Country * Size deleted
Country 0.028**
Type of Financial Institution 0.057*
Size 0.663
Level of Experience 0.512
R-square = 0.151; Adjusted R-square = 0.084; p = 0.047
Level 4 – Size deleted
Country 0.026**
Type of Financial Institution 0.059*
Level of Experience 0.387
R-square = 0.149; Adjusted R-square = 0.094; p = 0.027
Level 5 – Level of Experience deleted
Country 0.039**
Type of Financial Institution 0.048**
R-square = 0.141; Adjusted R-square = 0.097; p = 0.017

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Only two of the initial seven variables are significant predictors of preparer

perception of the relevance of fair value accounting. Country and type of financial

institution are the two significant predictors. However, in Section 5.4, country and

type of financial institution seem to be a partial surrogate for level of experience.

Level of experience is highly insignificant at level 4 and thus, country and type of

financial institution are more significant influences on preparers’ perception of the

relevance of fair value accounting.

129
5.8.1.1 Country Effect

The descriptives in Table 5.12 show Singaporean preparers’ perception of the

relevance of fair value accounting is higher than Australian preparers.

Table 5.12: Preparers’ Perception of the Relevance of Fair Value Accounting

by Country
Country n Mean Std. Error
Australia 41 3.45 .179
Singapore 42 4.04 .172
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

This finding is interesting as while the Singaporean preparers are neutral on the

difference between the banking and trading book, they are more strongly in favour of

the relevance of fair value accounting. Perhaps this signals that preparers in this

study do not regard the purported fundamental difference between the two books to

be a sufficient reason for the use of different measurement methods for financial

instruments.

5.8.1.2 Type of Financial Institution Effect

The other statistically significant predictor of preparer perception of the relevance of

fair value accounting is type of financial institution. Table 5.13 and Figure 5.3

depicts the marginal means of preparer responses to the statement.

Table 5.13: Preparers’ Perception of the Relevance of Fair Value Accounting

by Type of Financial Institution
Type of Financial Institution n Mean Std. Error
Local Bank 16 3.20 .241
Foreign Bank 24 3.78 .219
Credit Union 23 3.91 .241
Other Financial Institution 20 4.10 .210
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

130
Figure 5.3: Preparers’ Perception of the Relevance of Fair Value Accounting
by Type of Financial Institution

Average Response
4

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

It seems that preparers from the “other” financial institutions category believe fair

value accounting is more relevant than preparers from the other categories. This may

be due to the concentration of different types of financial instruments at the other

three categories of financial institutions. However, many credit unions do not have

trading financial instruments and as a result their support for the relevance of fair

value accounting for all financial instruments was somewhat surprising. Local banks

show less support for fair value accounting, as expected, from the assertions of

JWGBA, which represents major banking associations. Credit unions, treasuries,

merchant banks and finance companies have less (if any) representation on the

JWGBA, their (potentially unheard) opinions which should have some influence on

the future of fair value accounting for all financial instruments. Therefore, preparers

on average do feel that fair value accounting results in relevant information giving

some support to the JWGSS assertions. Australian local banks tend to show less

support for the relevance of fair value accounting.

131
5.8.2 Regressions for Reliability of Fair Values

The predictors used in the previous general linear models are again included in the

next regressions examining possible influences on preparers’ perception of the

reliability of fair values of traded financial instruments and non-traded financial

instruments. Residual plots (not shown) do not indicate any violation of underlying

assumptions.

5.8.2.1 Traded Financial Instruments

Table 5.14 and the following paragraphs detail the results of the general linear

models and backward elimination process.

Table 5.14: Regression Results for Preparers’ Perception of the Reliability of

Fair Values for Traded Financial Instruments

Variables in the equation Level of Significance

Country 0.029**
Type of Financial Institution 0.071*
Size 0.065*
Level of Experience 0.151
Country * Type of Financial Institution 0.195
Country * Size 0.160
Country * Level of Experience 0.162
R-square = 0.273; Adjusted R-square = 0.172; p = 0.007
Level 1 – Country * Type of Financial Institution deleted
Country 0.020**
Type of Financial Institution 0.055*
Size 0.112
Level of Experience 0.130
Country * Size 0.108
Country * Level of Experience 0.191
R-square = 0.240; Adjusted R-square = 0.157; p = 0.007

132
Table 5.14: Regression Results for Preparers’ Perception of the Reliability of
Fair Values for Traded Financial Instruments (Continued)

Variables in the equation Level of Significance

Level 2 – Country * Level of Experience deleted
Country 0.052*
Type of Financial Institution 0.080*
Size 0.168
Level of Experience 0.124
Country * Size 0.064*
R-square = 0.222; Adjusted R-square = 0.149; p = 0.007
Level 3 – Level of Experience deleted
Country 0.105
Type of Financial Institution 0.066*
Size 0.089*
Country * Size 0.106
R-square = 0.197; Adjusted R-square = 0.133; p = 0.009
Level 4 – Country * Size deleted
Country 0.771
Type of Financial Institution 0.007***
Size 0.361
R-square = 0.168; Adjusted R-square = 0.114; p = 0.013
Level 5 – Country deleted
Type of Financial Institution 0.004***
Size 0.365
R-square = 0.181; Adjusted R-square = 0.139; p = 0.003
Level 6 – Size deleted
Type of Financial Institution 0.003***
R-square = 0.159; Adjusted R-square = 0.127; p = 0.003

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

On the issue of reliability of traded financial instruments, the type of financial

institution remained the only significant predictor of preparer perception of the

reliability of traded financial instruments’ fair values. As noted in Section 5.4, type

of financial institution can be a partial surrogate for level of experience. However,

level of experience is eliminated at level 3 due to its statistical insignificance.

133
5.8.2.1.1 Type of Financial Institution Effect

The estimated marginal means are displayed in Figure 5.4 and Table 5.15. They

detail the influence of the type of financial institution on preparer perception of the

reliability of fair values for traded financial instruments.

Figure 5.4: Preparers’ Perception of the Reliability of Fair Values of Traded

Financial Instruments by Type of Financial Institution

5
Average Response

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

Table 5.15: Preparers’ Perception of the Reliability of Fair Values for Traded
Financial Instruments by Type of Financial Institution
Type of Financial Institution n Mean Std. Error
Local Bank 16 3.81 .211
Foreign Bank 24 4.08 .172
Credit Union 23 3.65 .176
Other Financial Institution 20 4.60 .189
Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

Preparers at “other” financial institutions tend to perceive fair values of traded

financial instruments to be highly reliable, followed by foreign banks.

The average response from local banks is perhaps not as strong as expected

according to JWGBA’s views. Overall, preparers perceive fair values for traded

financial instruments to be relatively reliable.

134
5.8.2.2 Non-Traded Financial Instruments

Table 5.16 and the following paragraphs detail the results of the general linear

models and backward elimination process in relation to possible predictors of

preparer perception of the reliability of non-traded financial instruments.

Table 5.16: Regression Results for Preparers’ Perception of the Reliability of

Fair Values for Non-Traded Financial Instruments

Variables in the equation Level of Significance

Country 0.455
Type of Financial Institution 0.232
Size 0.070*
Level of Experience 0.263
Country * Type of Financial Institution 0.602
Country * Size 0.875
Country * Level of Experience 0.799
R-square = 0.168; Adjusted R-square = 0.053; p = 0.174
Level 1 – Country * Size deleted
Country 0.354
Type of Financial Institution 0.229
Size 0.048**
Level of Experience 0.264
Country * Type of Financial Institution 0.593
Country * Level of Experience 0.771
R-square = 0.168; Adjusted R-square = 0.065; p = 0.121
Level 2 – Country * Level of Experience deleted
Country 0.120
Type of Financial Institution 0.191
Size 0.046**
Level of Experience 0.261
Country * Type of Financial Institution 0.587
R-square = 0.167; Adjusted R-square = 0.077; p = 0.080
Level 3 – Country * Type of Financial Institution deleted
Country 0.109
Type of Financial Institution 0.042**
Size 0.067*
Level of Experience 0.301
R-square = 0.155; Adjusted R-square = 0.088; p = 0.041

135
Table 5.16: Regression Results for Preparers’ Perception of the Reliability of
Fair Values for Non-Traded Financial Instruments (Continued)

Variables in the equation Level of Significance

Level 4 – Level of Experience deleted
Country 0.163
Type of Financial Institution 0.057*
Size 0.020**
R-square = 0.143; Adjusted R-square = 0.087; p = 0.034
Level 5 – Country deleted
Type of Financial Institution 0.079*
Size 0.062*
R-square = 0.121; Adjusted R-square = 0.076; p = 0.038

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The backward elimination reveals type of financial institution and size to be

moderately statistically significant predictors of preparer perception of the reliability

of fair values of non-traded financial instruments suggesting a possible trend. As per

Section 5.8.2.1, the type of financial institution may be a partial surrogate for level of

experience. Level of experience is again eliminated at level 3 for its statistical

insignificance. Thus, the type of financial institution is a more significant predictor

of preparers’ perception.

5.8.2.2.1 Type of Financial Institution Effect

The estimated marginal means depicts the influence of the type of financial

institution on preparer perception of the reliability of fair values for non-traded

financial instruments.

136
Figure 5.5: Preparers’ Perception of the Reliability of Fair Values of Non-
traded Financial Instruments by Type of Financial Institution

Average Response
4

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

Table 5.17: Preparers’ Perception of the Reliability of Fair Values for Non-
Traded Financial Instruments by Type of Financial Institution
Type of Financial Institution n Mean* Std. Error
Local Bank 16 2.38 .243
Foreign Bank 24 3.12 .203
Credit Union 23 3.03 .205
Other Financial Institution 20 3.12 .218
Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5
* Estimated marginal means when assets = 3.86.

Both Figure 5.5 and Table 5.17 show preparers at local banks as having the lowest

opinion on the reliability of the fair values for non-traded financial instruments while

the average response from financial institutions in the other three categories are

similar and tended to cluster around the neutral position. Local banks are the

prominent supporter of JWGBA views more than all other types of financial

institutions. It reflects that the JWGBA primarily represent the views of banks. This

could be due to banks having more demand deposits and loans, whose fair values are

arguably the most difficult to obtain.

5.8.2.2.2 Size Effect

The regression coefficient for size is -0.129. This indicates that for each unit

increase in the level of assets (i.e. the next category up or ten times more assets),

137
preparer response is estimated to decrease by 0.129; not a huge amount. Thus, size

(defined according to the scale used in the survey) needs to increase by a tremendous

degree before there is a substantial decrease in preparers’ perception of the reliability

of non-traded financial instruments’ fair values (even 1000 times bigger only results

in a 0.387 (3 x 0.129) decrease).

5.8.3 Regressions for Comparability of Fair Value Information

As shown in Table 5.9, preparers agree with the JWGSS in that fair value accounting

promotes comparability of financial information (mean of 3.6). General linear

models are then fitted to determine possible predictors of preparer perceptions of the

comparability of fair value accounting. Again, residual plots are checked and no

instances of violation of assumptions are found.

Table 5.18: Regression Results for Preparers’ Perception of the Comparability

of Fair Value Accounting

Variables in the equation Level of Significance

Country 0.713
Type of Financial Institution 0.067*
Size 0.514
Level of Experience 0.786
Country * Type of Financial Institution 0.719
Country * Size 0.731
Country * Level of Experience 0.861
R-square = 0.123; Adjusted R-square = 0.001; p = 0.443
Level 1 – Country * Level of Experience deleted
Country 0.745
Type of Financial Institution 0.056*
Size 0.525
Level of Experience 0.780
Country * Type of Financial Institution 0.717
Country * Size 0.701
R-square = 0.123; Adjusted R-square = 0.015; p = 0.350

138
Table 5.18: Regression Results for Preparers’ Perception of the Comparability
of Fair Value Accounting (Continued)

Variables in the equation Level of Significance

Level 2 – Country * Type of Financial Institution deleted
Country 0.688
Type of Financial Institution 0.030**
Size 0.614
Level of Experience 0.814
Country * Size 0.653
R-square = 0.115; Adjusted R-square = 0.032; p = 0.223
Level 3 – Country * Size deleted
Country 0.948
Type of Financial Institution 0.031**
Size 0.755
Level of Experience 0.868
R-square = 0.112; Adjusted R-square = 0.042; p = 0.158
Level 4 – Country deleted
Type of Financial Institution 0.029**
Size 0.710
Level of Experience 0.850
R-square = 0.112; Adjusted R-square = 0.055; p = 0.096
Level 5 – Level of Experience deleted
Type of Financial Institution 0.028**
Size 0.586
R-square = 0.112; Adjusted R-square = 0.066; p = 0.053
Level 6 – Size deleted
Type of Financial Institution 0.028**
R-square = 0.108; Adjusted R-square = 0.075; p = 0.028

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Only the type of financial institution proved to be a statistically significant predictor

of preparer opinions on the comparability of financial statements. Level of

experience, although found to be a partial surrogate for the type of financial

institution, was eliminated at level 4 due to its statistical insignificance. Here, the

type of financial institution is a more significant predictor of preparers’ perception.

139
5.8.3.1 Type of Financial Institution Effect

Preparer perceptions by type of financial institution is shown in Figure 5.6 and Table

5.19.

Figure 5.6: Preparers’ Perception of the Comparability of Fair Value

Accounting by Type of Financial Institution

5
Average Response

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

Table 5.19: Preparers’ Perception of the Comparability of Fair Value

Accounting by Type of Financial Institution
Type of Financial Institution n Mean Std. Error
Local Bank 16 2.94 .252
Foreign Bank 24 3.71 .206
Credit Union 23 3.52 .211
Other Financial Institution 20 3.95 .226

Figure 5.6 and Table 5.19 show preparers from “other” financial institutions are the

strongest supporter of the view that fair value accounting promotes the comparability

of financial statements between organisations. They are followed closely by foreign

banks and credit unions. Unsurprisingly, local banks have the lowest opinion on the

comparability of fair value accounting in line with JWGBA assertions. Perhaps local

banks are less supportive of this statement due to a higher concentration of non-

traded financial instruments where the comparability of such fair values is believed

to be harder to ensure.

140
5.9 CONCERNS REGARDING VOLATILITY

Another main theme identified in this debate relates to the supposed myriad of

problems that will arise due to the earnings volatility introduced by the use of fair

value accounting for all financial instruments. Respondents are asked their opinion

on this claim (see Table 5.20).

Table 5.20: Level of Preparers’ Perception of the Potential Problem with

Volatility in Reported Earnings Resulting from Fair Value
Accounting

Disagree
Disagree
Strongly

Strongly

p-value
Means
(n=83)
Neutral
JWGBA Views

Agree

Agree
Fair value accounting introduces
volatility in reported profits that 1 15 20 40 7 3.5*** .000
may be misunderstood by users
of the accounts.
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Preparers agree with the JWGBA that volatility in reported earnings may be

misunderstood with 75% (significantly different from 50% with p-value of 0.000) of

the preparers with a non-neutral opinion perceive that the earnings volatility

introduced may be misunderstood. This seems consistent with JWGBA’s argument

that it is confusing and will cause problems to both users and preparers.

5.9.1 Regressions on Concerns Regarding Volatility

As mentioned in the previous sub-section, preparers agree that the earnings volatility

introduced by fair value accounting may be misunderstood by users. Subsequently,

the descriptors of country, type of financial institution, size and level of experience

141
and the two-way interactions between country and each of the other three predictors

are then fitted in general linear models to determine possible predictors. Residual

plots for each model fitted and underlying assumptions are met. Table 5.21 shows

the initial general linear model fitted followed by the models resulting from the

backward elimination process.

Table 5.21: Regression Results for Preparers’ Concerns on Volatility in

Reported Earnings from the Use of Fair Value Accounting

Variables in the equation Level of Significance

Country 0.818
Type of Financial Institution 0.660
Size 0.833
Level of Experience 0.834
Country * Type of Financial Institution 0.740
Country * Size 0.839
Country * Level of Experience 0.841
R-square = 0.028; Adjusted R-square = -0.107; p = 0.995
Level 1 – Country * Level of Experience deleted
Country 0.878
Type of Financial Institution 0.664
Size 0.857
Level of Experience 0.838
Country * Type of Financial Institution 0.730
Country * Size 0.806
R-square = 0.027; Adjusted R-square = -0.092; p = 0.989
Level 2 – Country * Size deleted
Country 0.815
Type of Financial Institution 0.670
Size 0.951
Level of Experience 0.802
Country * Type of Financial Institution 0.733
R-square = 0.027; Adjusted R-square = -0.079; p = 0.979
Level 3 – Country * Type of Financial Institution deleted
Country 0.798
Type of Financial Institution 0.804
Size 0.981
Level of Experience 0.686
R-square = 0.018; Adjusted R-square = -0.059; p = 0.963

142
Table 5.21: Regression Results for Preparers’ Concerns on Volatility in
Reported Earnings from the Use of Fair Value Accounting
(Continued)

Variables in the equation Level of Significance

Level 4 – Size deleted
Country 0.787
Type of Financial Institution 0.769
Level of Experience 0.657
R-square = 0.018; Adjusted R-square = -0.045; p = 0.918
Level 5 – Country deleted
Type of Financial Institution 0.755
Level of Experience 0.717
R-square = 0.017; Adjusted R-square = -0.033; p = 0.846
Level 6 – Type of Financial Institution deleted
Level of Experience 0.657
R-square = 0.002; Adjusted R-square = -0.010; p = 0.657

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Table 5.21 reveals that none of the variables included in the models prove to be

statistically significant predictors. There are no determinants of preparers’

perception of the problems associated with volatility in reported earnings in terms of

user misunderstanding.

5.10 PREFERENCE FOR FAIR VALUE ACCOUNTING

The final key theme in Figure 5.1 to be examined is preparers’ preference for fair

value accounting. In the survey, chief financial officers are asked four questions to

indicate the extent (if any) that fair value accounting is preferred for financial

instruments. These questions relate to taking fair values to the Income Statement or

Balance Sheet for financial instruments in the trading or banking book. Their

responses are analysed as follows. Preparer responses of strongly oppose, oppose,

143
neutral, support, and strongly support are replaced with the values 1, 2, 3, 4 and 5

respectively to quantify the strength of support for each question. Then, a one-

sample t-test is used to compare the means to the midpoint position (3 on the 5-point

scale). Significant p-values indicate that the means are significantly different from 3

(i.e. a neutral response). In addition, an overall measure of preference for fair value

accounting is derived after examining the correlations between the four questions.

Then, regressions are conducted to find predictors of this overall preference for fair

value accounting.

5.10.1 Trading and Banking Books

The responses to the questions on preparer preference for the two aspects of fair

value accounting for financial instruments (distinguished by the trading and banking

books) are indicated in Table 5.22. There is support (means of 4.2 and 4.1, highly

significant at the .01 level) for fair value accounting in the trading book but neutrality

(means of 3.3 and 3.2) towards its use for financial instruments in the banking book.

The average response to the fourth statement (Q2b) is moderately significant with a

mean above the neutral position of 3. Exclusion of preparers with neutral opinions

for the two questions in relation to the banking book show that 59% and 57%of

remaining preparers support fair values in the Balance Sheet and the recognition of

fair value changes in the Income Statement respectively. However, these two

percentages are not significantly different from 50% according to the one-sample test

of proportions with p-values of 0.154 and 0.295 respectively.

144
Table 5.22: Preparers’ Level of Support for Fair Value Accounting in the
Trading and Banking Book

Support

Strongly
Strongly

Support
Neutral

p-value
Means
(n=83)
Oppose

Oppose
Label Question

Q1a Do you support the concept of

taking changes in fair values 1 8 4 36 34 4.1*** .000
of financial instruments in the
trading book to the Income
Statement?

Q1b Do you support the concept of

marking financial instruments 1 7 4 35 36 4.2*** .000
in the trading book to fair
value on the Balance Sheet?

Q2a Do you support the concept of

taking changes in fair values 7 25 9 28 14 3.2 .147
of financial instruments in the
banking book to the Income
Statement?

Q2b Do you support the concept of

marking financial instruments 7 22 12 27 15 3.3* .073
in the banking book to fair
value on the Balance Sheet?

Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.

* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The small number of preparers with a neutral opinion also highlights that nearly all

respondents have an opinion on this issue but their opinions are not aligned.

Moreover, they are on the two ends of the spectrum. Thus, of the preparers with a

non-neutral opinion, at least 57% of them support fair value accounting in the

banking book, contrary to the JWGBA position. These results also indicate that the

use of fair value accounting in the trading book and the opposition to its use in the

banking book is not unanimously supported.

145
5.10.2 Relationship between the Four Questions: Additional
Analysis

This section reports cross tabulations analysis for the four questions on fair value

accounting distinguished by the two books. Additionally, an overall score for

preparer preference for fair value accounting is derived from those four questions.

5.10.2.1 Cross-Tabulations

Cross tabulations were conducted for both pairs of related questions to examine

possible patterns in preparer responses. Tables 5.23 and 5.24 show that most

respondents are on the diagonal. This indicated that respondents who opposed one

aspect of fair value accounting (such as taking changes in fair values to the Income

Statement) tended to oppose the other requirement of fair value accounting.

Table 5.23: Cross Tabulation of Preparers’ Responses to the Two Questions on

Fair Value Accounting for the Trading Book
Do you support the concept of taking changes in
fair values of financial instruments in the trading
book to the Income Statement? (Q1a)
Strongly Oppose Neutral Support Strongly
Oppose Support
Do you support Strongly Oppose 1
the concept of Oppose 7
marking financial
instruments in the Neutral 1 2 1
trading book to Support 1 33 1
fair value on the
Strongly Support
Balance Sheet? 1 2 33
(Q1b)
Total (N = 83) 1 8 4 36 34

Most respondents tend to show support for fair value accounting in the trading book,

both the measurement at fair value and the recognition of changes in fair values as

revenues or expenses in the Income Statement. Similarly, preparers tend to either

oppose or support both the requirements of fair value accounting in the banking

book. For example, 13 respondents strongly support both the marking of financial

146
instruments to fair value on the Balance Sheet and taking resultant changes to the

Income Statement in the banking book (see Table 5.24).

Table 5.24: Cross Tabulation of Preparers’ Responses to the Two Questions on

Fair Value Accounting for the Banking Book
Do you support the concept of taking changes in
fair values of financial instruments in the banking
book to the Income Statement? (Q2a)
Strongly Oppose Neutral Support Strongly
Oppose Support
Do you support Strongly Oppose 5 2
the concept of
Oppose 1 20 1
marking financial
instruments in the Neutral 1 8 2 1
banking book to Support 1 2 1 23
fair value on the
Balance Sheet? Strongly Support
2 13
(Q2b)
Total (N = 83) 7 25 9 28 14

There are four instances of ‘outliers’ as highlighted in blue and purple in Table 5.24.

The first group of three (purple shading) are preparers who support measuring

financial instruments in the banking book at fair value on the Balance Sheet but

opposed the recognition of resultant changes to the Income Statement. These three

show support for the concerns of earnings volatility highlighted in the literature.

Examination of the characteristics of these three preparers does not reveal any

patterns; one is a highly experienced Australian preparer from a credit union while

the other two are Singaporean preparers of financial institutions with less than $1

million in assets and with less than five years experience. The preparer highlighted

in blue is more problematic28 in that it goes against expectations as per past research

and JWGBA assertions. Curiously, this Singaporean preparer at a local bank with

less than five years of experience supports the taking of changes to fair values in the

28
The original survey was also rechecked and confirmed this was not a data entry error.

147
banking book to the Income Statement but not fair value measurement in the Balance

Sheet.

Correlations between these four questions are then calculated (see Table 5.25). The

labels used for the correlations below were detailed in Table 5.22. There are

significant correlations between the responses to the first pair of questions in Table

5.25 (r = 0.94, p = 0.000) as well as the responses to the second pair of questions (r =

0.87, p = 0.000). Statistically, although there is significant correlation between

preparers’ response to questions in the first and second pair, these correlations are

not as strong (r = 0.270 to 0.326).

Table 5.25: Correlation between the Four Questions

Label Q1b Q2a Q2b

Pearson Correlation Q1a 0.936*** 0.283*** 0.309***
Q1b 0.270** 0.326***
Q2a 0.873***
Significance Q1a 0.000 0.010 0.004
Q1b 0.013 0.003
Q2a 0.000

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

As a further examination of this lower correlation between the questions in the first

and second pairs, more cross tabulations are prepared. This time, the minimum score

for Q1a and Q1b as well as for Q2a and Q2b is computed and compared in Table

5.26. Although there is a tendency along the diagonal, two other groups stand out.

Firstly, there are preparers who support fair value accounting in the trading book

who oppose its use in the banking book (highlighted in blue), lending weight to

JWGBA assertions. The second intriguing group (purple shade) consists of preparers

who support fair value accounting in the banking book but not for the trading book.

148
These two preparers, both Singaporean with less than five years experience from a

local bank of different sizes are contrary to expectations.

Table 5.26: Cross Tabulation of Preparers’ Minimum Responses to Q1a and

Q1b and Q2a and Q2b

Minimum score of preparer responses to Q1a and

Q1b on fair value accounting in the trading book
Strongly Oppose Neutral Support Strongly
Oppose Support
Minimum score of Strongly Oppose 1 2 6
preparer responses
Oppose 6 3 11 4
to Q2a and Q2b
on fair value Neutral 2 4 6
accounting in the Support 2 19 4
banking book
Strongly Support 13
Total (N = 83) 1 8 5 36 33

These statistically significant correlations between the four questions provide

credibility to additional analysis using only one measure for overall preference for

fair value accounting. The following sub-section details the derivation of one overall

measure for the preference for fair value accounting.

5.10.2.2 Four Becomes One

According to IAS 39 (also defined in the survey), fair value accounting refers to the

measurement of all financial instruments at fair value and the recognition of changes

in fair value as revenues or expenses in the Income Statement. Opposition towards

any of the four questions, by definition, is equal to opposition towards fair value

accounting. Therefore, the first option for an overall preference measure29 for fair

29
Separate analyses were also performed on the preference for fair value accounting in the trading
and banking book separately (see Appendix J). There is high level of similarity in the significance
of the variables in the regressions for the overall measure of preference for fair value accounting
and the Q1 pair as well as the Q2 pair.

149
value accounting score is to make it equal to the minimum score of the four

questions. For example, if a respondent strongly opposed the taking of changes in

fair value to the Income Statement but supported the other three statements, the

respondent is deemed to strongly oppose fair value accounting overall.

However, a possible second option for measuring overall preference for fair value

accounting is using the average responses to the four questions (calculated by adding

up the four respective scores, dividing the summation by 4 and rounding to the

nearest integer). These two different ways for measuring overall preference are

derived and correlations subsequently computed as further examination of the

viability of this new measure. The results show a statistically significantly high

correlation between these two options with r = 0.848 (p = 0.000), suggesting either

option is representative. Table 5.27 presents the cross tabulation between these two

measures as an additional check.

Table 5.27: Cross Tabulation of Preparers’ Minimum Responses and Average

Responses to the Four Questions
Average preparer responses to the four questions
Strongly Oppose Neutral Support Strongly
Oppose Support
Minimum of Strongly Oppose 1 7 1
preparer responses
Oppose 5 16 5
to the four
questions Neutral 2 10
Support 18 5
Strongly Support 13
Total (N = 83) 1 5 25 34 18

The majority of preparer responses are on the diagonal except for two distinct

groups. First of which are the 23 preparers’ whose average response to the four

questions results in a neutral position but the minimum score shows opposition to

preference for fair value accounting due to their opposition to its use in some form

150
for financial instruments in the banking book. The second group consists of

supporters who oppose fair value accounting according to the minimum score, again

due to opposition to fair value accounting in the banking book. In both instances, the

minimum score option is arguably more accurate in that opposition to fair value

accounting in the banking book is ultimately opposition to its use for all financial

instruments. Therefore, the minimum score better captures the essence of preference

for fair value accounting.

Although each option has their strengths (and weaknesses), the former option is

assumed and adopted for its theoretical strength. It is highly aligned with the concept

of fair value accounting as defined in the draft accounting standard by including the

Income Statement component of fair value accounting. It may seem like a bias

against fair value but it is more technically in line with the concept of fair value

accounting30 as per the proposed fair value accounting model. Therefore, in keeping

with the theme of this research, this new measure of overall preference for fair value

accounting is subject to further statistical examination in Section 5.10.3.

5.10.3 Overall Preference for Fair Value Accounting

An overall preference for fair value accounting score is calculated based on the

minimum score of the four questions asked. From this point onwards, preparer

preference for fair value accounting is measured using this new variable.

Subsequently, this new aggregated overall preference for fair value accounting (not

the initial four questions) is subject to further analyses.

30
The neutrality exhibited by respondents in Section 5.6 in relation to the differentiation between the
trading and banking books also adds credence to this new variable.

151
Table 5.28: Preparers’ Overall Support for Fair Value Accounting

Support

Strongly
Strongly

Support
Neutral

p-value
Means
(n=83)
Oppose

Oppose
Overall support for fair value
accounting as measured by the 9 26 12 23 13 3.1 .672
minimum score of all four
questions.
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.

As shown in Table 5.28, only 13 preparers show strong support for fair value

accounting and nine strongly oppose its use. However, the average response (mean

is 3.1) indicates that preparers are neutral on the use of fair value accounting. The

number of preparers supporting fair value accounting is just one more than opposing

preparers. If the 12 preparers with neutral opinion are excluded, 49% of preparers

oppose fair value accounting, which is not significantly different to 50% (p-value =

1.000). Thus, preparers’ opinions are highly divided, further highlighting the lack of

consensus among preparers in relation to fair value accounting for all financial

instruments.

5.10.3.1 Regression Results

The four variables of country, size, type of financial institution and level of

experience together with the two-way interactions between country and the other

three variables are included in the general linear regression model. However, logic

suggests that preparers’ responses to the key themes detailed previously will affect

their preference for fair value accounting. Therefore, in addition to the

aforementioned variables, preparer responses on the distinction between the banking

152
and trading book31, qualitative characteristics and problems with volatility of

reported earnings32 are also included as covariates in the initial general linear model.

The initial model is then refined using the backward elimination process as before.

Table 5.29 depicts the first general linear model fitted and successive models

resulting in the final general model. Again, residual plots for each model fitted are

checked and there is no violation of underlying statistical assumptions.

Table 5.29: Regression Results for Overall Preparers’ Preference for Fair
Value Accounting

Variables in the equation Level of Significance

Country 0.808
Type of Financial Institution 0.993
Size 0.809
Level of Experience 0.737
Country * Type of Financial Institution 0.032**
Country * Size 0.783
Country * Level of Experience 0.451
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.164
Support for Reliability of Traded Financial Instruments 0.462
Support for Reliability of Non-traded Financial Instruments 0.001***
Support for Comparability 0.063*
Volatility will not be Misunderstood 0.514
R-square = 0.752; Adjusted R-square = 0.692; p = 0.000

31
The data for this variable was recoded as the questions were expressed in the negative, against fair
value accounting, which is the opposite of the other three variables. Therefore, the descriptives
provided in Section 5.6 is the opposite for the variable included in this group of GLMs.
32
This variable was also recoded as the questions were expressed in the negative. The descriptives
provided in Section 5.7.4 is the opposite for the volatility variable included in this group of GLMs.

153
Table 5.29: Regression Results for Overall Preparers’ Preference for Fair
Value Accounting (Continued)

Variables in the equation Level of Significance

Level 1 – Country * Size deleted
Country 0.499
Type of Financial Institution 0.981
Size 0.903
Level of Experience 0.699
Country * Type of Financial Institution 0.030**
Country * Level of Experience 0.462
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.134
Support for Reliability of Traded Financial Instruments 0.407
Support for Reliability of Non-traded Financial Instruments 0.001***
Support for Comparability 0.063*
Volatility will not be Misunderstood 0.506
R-square = 0.752; Adjusted R-square = 0.696; p = 0.000
Level 2 – Country * Level of Experience deleted
Country 0.955
Type of Financial Institution 0.965
Size 0.706
Level of Experience 0.701
Country * Type of Financial Institution 0.029**
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.155
Support for Reliability of Traded Financial Instruments 0.498
Support for Reliability of Non-traded Financial Instruments 0.001***
Support for Comparability 0.059*
Volatility will not be Misunderstood 0.519
R-square = 0.750; Adjusted R-square = 0.698; p = 0.000
Level 3 – Size deleted
Country 0.803
Type of Financial Institution 0.901
Level of Experience 0.753
Country * Type of Financial Institution 0.028**
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.165
Support for Reliability of Traded Financial Instruments 0.463
Support for Reliability of Non-traded Financial Instruments 0.000***
Support for Comparability 0.048**
Volatility will not be Misunderstood 0.512
R-square = 0.749; Adjusted R-square = 0.702; p = 0.000

154
Table 5.29: Regression Results for Overall Preparers’ Preference for Fair
Value Accounting (Continued)
Level 4 – Level of Experience deleted
Country 0.905
Type of Financial Institution 0.881
Country * Type of Financial Institution 0.018**
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.133
Support for Reliability of Traded Financial Instruments 0.486
Support for Reliability of Non-traded Financial Instruments 0.000***
Support for Comparability 0.049**
Volatility will not be Misunderstood 0.503
R-square = 0.749; Adjusted R-square = 0.706; p = 0.000
Level 5 – Volatility will not be Misunderstood deleted
Country 0.953
Type of Financial Institution 0.861
Country * Type of Financial Institution 0.017**
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.093*
Support for Reliability of Traded Financial Instruments 0.489
Support for Reliability of Non-traded Financial Instruments 0.000***
Support for Comparability 0.059*
R-square = 0.747; Adjusted R-square = 0.708; p = 0.000
Level 6 – Support for Reliability of Traded Financial Instruments deleted
Country 0.870
Type of Financial Institution 0.730
Country * Type of Financial Institution 0.013**
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.116
Support for Reliability of Non-traded Financial Instruments 0.000***
Support for Comparability 0.051*
R-square = 0.745; Adjusted R-square = 0.710; p = 0.000
Level 7 – Support for Relevance deleted
Country 0.481
Type of Financial Institution 0.675
Country * Type of Financial Institution 0.012**
No Distinction between Banking and Trading Books 0.000***
Support for Reliability of Non-traded Financial Instruments 0.000***
Support for Comparability 0.000***
R-square = 0.737; Adjusted R-square = 0.704; p = 0.000

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

155
Two insignificant variables remain in the last model as they form the significant

interaction between country and type of financial institution. Once more, the

interaction between country and type of financial institution is found to be a

statistically significant predictor. Preparer perception of the banking and trading

distinction, the perceived reliability of fair values for non-traded financial

instruments and their perception of the comparability resulting from the use of fair

value accounting are also significant predictors. The country and type of financial

institution variables cannot be eliminated from the model even though they are

insignificant because they form the significant interaction between country and type

of financial institution variable.

However, the level of experience seems to be a partial surrogate for country as well

as type of financial institution. Because the level of experience has the most

insignificant p-value at level 3, country and type of financial institution (specifically

the interaction between the two) are more significant influences. Each significant

factor is discussed below.

5.10.3.2 Country by Type of Financial Institution Effect

Figure 5.7 and Table 5.30 show a considerable difference between the preference for

fair value accounting of foreign banks and financial institutions in Australia and

Singapore. As shown in Table 5.30, the relationship between foreign banks and their

preference for fair value accounting in Australia is statistically significantly different

from the relationship between foreign banks in Singapore and their fair value

accounting preference. The generally low mean scores (ranging from 2.23 to 3.36)

indicate low support for fair value accounting by financial institutions.

156
Figure 5.7: Australian and Singaporean Preparers’ on Overall Preference for
Fair Value Accounting

Average Response
4

3 Australia
2 Singapore

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

Table 5.30: Preparers’ Preference for Fair Value Accounting by Country by

Type of Financial Institution
Country n Type of Financial Institution Mean* Std. Error
Australia 5 Local Bank 2.94 .331
3 Foreign Bank 2.26 .415
23 Credit Union 3.10 .147
10 Other Financial Institution 3.23 .237
Singapore 11 Local Bank 2.93 .220
21 Foreign Bank 3.36 .154
10 Other Financial Institution 2.61 .224
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.
* Estimated marginal means when significant covariates are held constant at their mean
values: No distinction between the banking and trading book = 2.80, reliability of fair
values for non-traded financial instruments = 2.95 and support for comparability = 3.57.

Foreign banks in Australia have the lowest preference for fair value accounting for

all financial instruments whereas their counterparts in Singapore show the highest

level of support. Unfortunately, due to the anonymous nature of the surveys no

further information is available on these foreign banks to examine this trend further.

157
5.10.3.3 No Distinction between Banking and Trading Book Effect

It is contended that respondents who perceived that the trading and banking books

are not sufficiently different would prefer fair value accounting. Statistically, a

regression coefficient of 0.458 confirms this expectation. Here, for every 1-unit

increase in the perception that the trading and banking books are not different, the

estimated marginal means for the overall preference for fair value accounting in the

regression equation increases by 0.458. In other words, preparers who do not

perceive the two books to be different support fair value accounting. The interaction

between country and this variable is not a significant predictor when included in the

final model (p-value = 0.173). Thus, the reported relationship between perception of

the trading and banking book and preference for fair value accounting is not different

for the two countries.

5.10.3.4 Reliable Non-traded Financial Instrument Fair Values Effect

It is also professed that respondents who perceive that the fair values of non-traded

financial instruments are reliable will tend to support fair value accounting. This is

confirmed by a regression coefficient of 0.382 indicating that for every 1-unit

increase in the perception of the reliability of those fair values, the estimated

marginal means for the overall preference for fair value accounting increases by

0.382. Thus, for preparers who have faith in the reliability of the fair values of non-

traded financial instruments, support for fair value accounting is higher.

Additional analyses are performed by adding the interaction between country and

preparer’s perception of the reliability of fair values for non-traded financial

instruments to the final model as per Table 5.29. This resultant model shows this

interaction variable to be moderately significant (p-value = 0.077). Examination of

158
the parameter estimate shows a regression coefficient for the reliability of non-traded

financial instruments variable of 0.546 for Australia and 0.248 for Singapore.

Although this interaction variable is only moderately significant, it provides evidence

that the overall preference for fair value accounting increases more for Australian

preparers than Singaporean ones for each 1-unit increase in the perception of the

reliability of fair values for non-traded financial instruments. That is, perception

about level of reliability is a stronger determinant of preference for fair value

accounting in Australia than in Singapore.

5.10.3.5 Support for Comparability Effect

Finally, it is proposed that respondents who perceive that fair value accounting

results in comparable information will tend to prefer fair value accounting more than

their counterparts. The regression coefficient for this variable is 0.383; thus for

every 1 unit increase in the support for comparability, the estimated marginal mean

preference for fair value accounting increases by 0.383. This relationship is not

different for the two countries as the interaction between country and support for

comparability is not a significant predictor when included in the final model (p-value

= 0.579).

Table 5.31 depicts preparer support for comparability against overall support for fair

value accounting.

159
Table 5.31: Cross Tabulation of Preparers’ Responses on Comparability and
their Overall Preference for Fair Value Accounting
Preparer perception of the comparability of fair
value accounting
Strongly Disagree Neutral Agree Strongly
Disagree Agree
Overall preparer Strongly Oppose 2 4 2 1
support for fair
Oppose 1 6 8 11
value accounting
Neutral 1 6 4 1
Support 1 1 17 4
Strongly Support 5 8
Total (N = 83) 3 12 17 37 14

There is one preparer on the extreme with the opinion that fair value accounting

results in information that improves comparability but strongly opposes the use for

fair value accounting. Further scrutiny reveals that this preparer strongly opposes the

taking of changes in fair value of financial instruments in the banking book to the

Income Statement. Therefore, although this preparer thinks fair value accounting

improves comparability, the opposition to reporting fair value changes to the Income

Statement results in overall opposition to fair value accounting.

5.11 PREPARER VIEWS ON KEY THEMES: SUMMARY

The main findings of this section on preparer views on key themes identified in this

study are summarised in Table 5.32.

160
Table 5.32: Summary of Preparers’ Perceptions and Predictors on the Four Key Themes

Item Average Non—Neutral % Predictor Direction

Support Oppose (p-value)
Distinction between trading Neutral 60% 40% Country by Type of Financial Local and foreign banks in Australia show more
and banking books (3.2) Institution (0.002) support for using different measurement bases for each
book as compared to Singaporean counterparts
Relevance of fair value Agree 85% 15% Country (0.039) Singapore financial institutions agree more
accounting (3.8) Type of Financial Institution (0.048) Local banks have lowest opinion on relevance
Reliability of fair values for Reliable 92% 8% Type of Financial Institution (0.003) Financial institutions in the other category perceive
traded financial instruments (4.0) these fair values to be more reliable than counterparts
Reliability of fair values for Neutral 48% 52% Type of Financial Institution (0.079) Local banks perceive these to be unreliable
non-traded financial (3.0) Size (0.062) Bigger financial institutions have less faith in the
instruments reliability
Comparability of fair value Agree 76% 24% Type of Financial Institution (0.028) Local banks have the lowest opinion on the
accounting (3.6) comparability resulting from fair value accounting
Volatility may be Agree 75% 25% None
misunderstood by users (3.5)
Overall support for fair Neutral 51% 49% Country by Type of Financial Foreign banks in Australia and other financial
value accounting (3.1) Institution (0.012) institutions in Singapore oppose more than
Banking versus Trading book counterparts in other country
(0.000) No perceived distinction, higher support
Reliability of non-traded financial Higher perception of reliability of the fair values of
instruments (0.000) non-traded financial instruments and the comparability
Support for comparability (0.000) of fair value accounting results in higher support for
fair value accounting

161
Table 5.32 shows one common trait in the results, that is higher opposition towards

fair value accounting from the local banks (especially those in Australia). This is

caused by either overall opposition or lower perception of the qualitative

characteristics possessed by information resulting from its use. This confirms the

JWGBA’s position of representing the views of Australian banks. However, the

lower levels of opposition from the other types of financial institutions only adds to

the strength of this study by reinforcing that not all preparers agree with the official

JWGBA position.

5.12 PREPARER VIEWS ON OTHER ISSUES

Respondents are also asked questions on other issues raised in the controversy. Such

questions include the JWGBA postulated high cost of obtaining fair value

information and lack of ease of obtaining reliable fair values. The level of preparer

agreement with respective statements on these other issues are shown in Table 5.33.

Table 5.33: Level of Preparers’ Agreement on Other Issues Raised in the Fair
Value Accounting Debate
Disagree
Disagree
Strongly

Strongly

p-value
Means
(n=83)
Neutral

Statements
Agree

Agree

Fair value accounting brings financial

reporting in line with current financial 2 8 15 44 14 3.7*** .000
risk management policies.
Reliable and independent market
valuations are impossible to obtain for 1 21 12 40 9 3.4*** .000
some products.
The cost of obtaining fair value
information will be unacceptably high. 1 30 29 21 2 2.9 .381

Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

*** Highly Significant at the 0.01 level

162
Preparers tend to agree (mean of 3.7, significantly different from the neutral) that fair

value accounting will bring financial reporting in line with current financial risk

management policies. This adds further support towards the relevance of fair value

accounting. However, preparers still perceive that reliable market valuations are

impossible to obtain for some products (mean of 3.4, p = .000). Thus, the earlier

evidence of preparers’ reasonably high perception of the reliability of fair values

could be hampered by the obstacles to obtaining reliable fair values. This suggests

that when reliable fair values are easily available, there is one less obstacle towards

perceived reliable fair values (especially for non-traded financial instruments).

Perhaps full acceptance and preference for fair value accounting for all financial

instruments can then be achieved.

Contradicting the JWGBA claims, respondents are neutral in relation to the

purported high cost of obtaining fair value information with a mean of 2.9 which is

not (statistically) significantly different from a neutral response. In addition, if the

29 respondents with a neutral opinion are excluded, only 43% of preparers agree

with the JWGBA assertion. The other 57% disagree (not statistically different from

50%) highlighting a lack of consensus in preparer views on this matter.

The survey also asks for the percentage of financial assets and liabilities that

respondents estimate to be able to determine reliable fair values (see Table 5.34).

More than half of the respondents indicate that the fair values of at least 51% of

financial assets and liabilities could be reliably determined by preparers.

Furthermore, at least 20% of the respondents state that fair values for over 90% (of

both financial assets and liabilities) could be reliably determined.

163
Table 5.34: The Percentage of Financial Assets and Liabilities which Reliable
Fair Values can be Determined by Preparers

Financial assets Financial liabilities

Frequency (%) Frequency (%)
0 to 10% 7 8.4 8 9.6
11 to 25% 8 9.6 6 7.2
26 to 50% 22 26.5 19 22.9
51 to 75% 11 13.3 14 16.9
76 to 90% 18 21.7 18 21.7
91 to 100% 17 20.5 18 21.7
Total 83 100.0 83 100.0

Therefore, preparers tend to agree on the relevance of fair value accounting but the

difficulty in obtaining reliable fair values seems to be the outstanding obstacle. An

interesting finding is preparer neutrality on the purported unacceptable high cost of

implementing fair value accounting for all financial instruments. As such, this

finding disputes the cost-benefit argument put forward by JWGBA.

5.13 SUMMARY OF PREPARER VIEWS

Preparer responses indicate that fair value accounting is used in some instances in the

trading book and also in the banking book. Respondents are relatively neutral on the

purported difference between the trading and banking book to warrant the use of

different measurement bases. Interestingly, local banks are the biggest supporter of

using different measurement bases for different books as compared to other financial

institutions. It seems that views remained moderate since Fargher (2001). This is

despite the time lag and apparent movement towards fair value accounting for

financial instruments and by the proposal to allow this option in IAS 39 as issued in

December 2003.

164
Preparers agree that fair value accounting improves the relevance of information and

promotes comparability of financial statements. Respondents perceive the fair values

of traded financial instruments to be reliable but are neutral on the reliability of fair

values for non-traded financial instruments. Preparers also believe that the earnings

volatility introduced by fair value accounting may be misunderstood by users but

none of the preparer descriptives prove to be predictors (see Table 5.32).

Although there is preparer support for fair value accounting in the trading book but

neutrality for its use in the banking book, results show preparer neutrality towards

overall support for fair value accounting for all financial instruments. Again, banks

in Australia show higher opposition than their Singaporean counterparts. However,

Singaporean financial institutions in the “other” category oppose it more than their

Australian equivalents. In addition, preparers’ prefer fair value accounting when the

(1) banking and trading books are not considered different, (2) fair values of non-

traded financial instruments are reliable and (3) comparability can be achieved.

In addition, preparers agree that fair value accounting brings financial reporting in

line with management practices (support for JWGSS assertions) and reliable fair

values are impossible to obtain for some products (supporting the official views of

the JWGBA). Another finding is preparer neutrality on the assertion of unacceptably

high cost of obtaining fair value information. This indicates that respondents do not

agree with the cost-benefit argument put forward by the JWGBA.

This evidence shows some significant contradictions from the official JWGBA and

JWGSS positions. The thesis evidence shows that there are financial institutions that

currently employ fair value accounting for some financial instruments in the banking

book. This is contrary to JWGBA (1999) assertion that fair value accounting is not

165
used for the banking book. They also argue that the trading and banking books are

fundamentally different but this is not unanimously supported. Where findings

reveal that preparers tend to agree with the assertions of JWGBA, it is to a lesser

extent than what the official position would lead one to believe. There was

considerable variation in views and both the JWGBA and the JWGSS do represent

the views of some (but never all) preparers.

Overall, the survey results support assertions of both parties to the international

debate, both for and against fair value accounting for all financial instruments. Both

the JWGBA and JWGSS contend that their views are supported by the users of

financial statements. The next chapter furthers this research by providing details on

the views of sophisticated users in Australia and Singapore.

166
 CHAPTER SIX:
QUANTITATIVE PHASE – USER VIEWS

6.1 INTRODUCTION

This second of three chapters analysing evidence gathered from the quantitative

phase is concerned with sophisticated user views. Insights obtained from the user

surveys from both Australia and Singapore is detailed; first descriptives are provided

and then multiple regressions are conducted to check for possible predictors of user

views. This chapter is followed by a comparison of the perceptions of the two

groups in Chapter Seven (see Figure 3.5 for further clarification).

6.2 USER SURVEY PARAMETERS

The user survey is very similar to the preparer survey (reported in Chapter 5) but

adapted to a user focus. Questions that are linked to current preparer practices and

demographics are excluded (as they are not relevant to users who do not prepare

financial statements). Other questions relating to criteria for certain qualitative

characteristics are added to facilitate a better examination of the usefulness of fair

value accounting from the perspective of users (see Appendix B and D). Other than

that, the Australian and Singaporean user survey differs only in terminology.

A total of 379 user surveys are sent to the head of research or chief investment

officers of sophisticated user organisations in Australia and Singapore. 229 user

surveys are disseminated33 to the head of research and chief investment officers of

33
A total of 33 (a coincidence that this number is the same as the surveys completed) surveys are
returned due to various reasons (people had moved job; the company had moved location etc.).

167
sophisticated user organisations in Australia. Only 33 responses are received giving

a low effective response rate of 16.8% (i.e. 33 out of 196). Similarly, only 32

responses are received from the 150 surveys sent to chief investment officers of

sophisticated user organisations in Singapore. Fifteen surveys are returned due to

errors in the address resulting in a response rate of 23.7% (i.e. 32 out of 135).

Overall, the aggregated response rate of 19.6% is relatively low but consistent with

the response rate of many studies in the social sciences (see Griffis, Goldsby and

Cooper, 2003 and Chiu and Brennan, 1990). Although complete generalisation of

the results is restricted, the strength of this study lies in the ability to provide insights

into the actual views of sophisticated users in both Australia and Singapore.

Therefore, the evidence and results add new breadth on the perceived usefulness of

fair value accounting for all financial instruments from the users’ point of view.

Given that accounting rules and standards are promulgated in order to provide useful

information to users, this evidence is considered important.

6.3 TESTING NON-RESPONSE BIAS

This low response rate signals possible non-response bias from the remaining 80.4%

of users. In order to test for non-response bias, general linear models are conducted

to test the effect of response time on users’ experience and overall preference for fair

value accounting. The results show that response time is not a determinant of users’

preference for fair value accounting (see Appendix I). However, for every 1-unit

increase in response (i.e. slower respondents), user experience decreases by 0.049 (p-

value = 0.003). This means that the last user respondents are just one and a half

categories of experience lower than the first respondent. Experienced users tend to

respond quicker and this suggests that this group of users is more likely to respond to

168
the survey. Although non-response bias may exist in that the results are less relevant

to inexperienced users, the views of inexperienced users are of less interest in this

thesis due to the complexity of fair value accounting. Therefore, no non-response

bias is detected for the key issue of preference for fair value accounting.

The following subsections detail user descriptives such as the type and level of

experience of the respondents.

6.4 USER DESCRIPTIVES

6.4.1 Type of User by Country

Seven categories of users are included in the survey namely auditors (labeled

assurance and advisory), bank analyst, financial analyst, investment analyst, fund

manager, industry association and an other grouping. Respondent users represent the

various groups in the survey. These categories are then collapsed into the three

dominant user groups of auditors (termed assurance and advisory), analysts (made up

of financial, bank and investment analysts) and ‘other’ users (see Appendix B and D

for full list). Table 6.1 depicts the number of respondents from different categories

of users.

Table 6.1: Breakdown of Type of Respondent Users

Type of User Australia Singapore Total Percent (%)

Assurance and Advisory 8 7 15 23.0
Analysts 11 14 25 38.5
Other 14 11 25 38.5
Total 33 32 65 100.0

This move also enhances statistical analyses by increasing cell sizes. The 25 users

from the other category include fund and trust managers and industry associations.

169
There are 15 auditors and 25 analysts. Overall, Table 6.1 showed that users were

adequately spread out between the categories enhancing the generalisability of the

findings.

6.4.2 Level of Experience by Country

Users are asked to indicate their level of experience from five ranges (less than 2

years, 2 to 5 years, 6 to 10 years, 11 to 15 years and more than 15 years). A

breakdown of respondents across the five groups are detailed in Table 6.2.

Table 6.2: The Level of Experience of User Respondents

Australia Singapore Total Percent (%)

< 2 years 1 4 5 7.7
2 - 5 years 9 18 27 41.5
6 - 10 years 4 7 11 16.9
11 - 15 years 7 2 9 13.8
> 15 years 12 1 13 20.0
Total 33 32 65 100.0

More than 40% of respondents have between two to five years experience in

financial reporting. About half of the respondents have more than six years

experience with thirteen users having more than 15 years experience. Australian

users in this sample tend to be more experienced as shown by a p-value of 0.001 in

the independent samples T-test conducted. In this instance, level of experience and

country are partial surrogates for each other and any country effect may be due to an

experience effect and vice versa. This surrogacy between experience and country is

considered in the discussion of the regression results.

170
6.4.3 Type of User by Level of Experience

A one-way ANOVA is used to examine if average experience is significantly

different for each type of user. This analysis reveals a p-value of 0.995 indicating no

statistical relationship between level of experience and type of user. Therefore, there

is no collinearity problem between these two independent variables.

The following sections detail users’ views on the key themes identified as well as

their views on other prominent issues asked in the mail survey.

6.5 USAGE OF FAIR VALUE INFORMATION

In the survey, users are asked the extent they would use fair value information on

financial instruments in the two different categories in their analysis of financial

institutions if it were available. Table 6.3 depicts the responses.

Table 6.3: The Extent that Users Use Fair Value Information where Available
Category of Financial Instruments Never Sometimes Always
Banking Book 17 28 20
Trading Book 2 20 43

Interestingly and contrary to the JWGBA assertions, survey responses indicate that at

least 28 and 20 users do use fair value information for financial instruments in the

banking book when available. Table 6.4 breaks down Table 6.3 by type of user.

171
Table 6.4: Usage of Fair Value Information for the Banking and Trading
Book by Type of User

Type of User Never Sometimes Always

Banking Assurance and Advisory 3 8 4
Analyst 6 11 8
Other 8 9 8
Total 17 28 20
Trading Assurance and Advisory 0 8 7
Analyst 1 3 21
Other 1 9 15
Total 2 20 43

Among the 17 users who would never use fair value in the banking book, there are

three auditors, six analysts and eight ‘other’ users who agree with the JWGBA

assertions. Interestingly, an analyst and one ‘other’ user would not use fair value

information in the trading book. Although the majority of users seem to lend support

for fair value accounting as indicated by their willingness to use the resultant

information, there are 17 users who would never use fair value information provided

for the banking book.

6.6 KEY THEMES AND POSSIBLE INFLUENCES

In the user survey, fair value accounting is specifically defined to refer to the

measurement of all financial instruments at fair value and the recognition of changes

in fair value in the Income Statement in the period in which they arise. The four

main themes identified in Chapter 4 form the focus of this quantitative phase. These

are (1) distinction between the trading and banking book, (2) whether fair value

accounting for all financial instruments results in decision-useful information, (3) the

problems with earnings volatility that results from the use of fair value accounting

and (4) overall preference for fair value accounting.

172
The following subsections outline the results obtained from the 65 user surveys in

relation to the four themes and examines potential determinants via the general linear

model using three34 variables of country, type of user and level of experience. Figure

6.1 depicts the variables and the key themes examined.

Figure 6.1: Diagram of Possible Predictors of User Views on Key Themes

Possible Predictors Key Themes

1. Distinction between trading and

Country banking book
2. Qualitative characteristics of fair
value accounting information
Type of User
3. Volatility that results from fair
value accounting
Experience 4. Preference for fair value accounting

For all of the following regressions, country and type of user are inserted as fixed

factors while level of experience is treated as a covariate (using the SPSS

terminology). Most of the general linear models fitted also included two-way

interactions between country and the other two independent variables. Results of

those models are detailed as follows for each key theme.

34
This is one less variable than the four possible predictors of preparer views. Size of company was
excluded due to its irrelevance. The user surveys were aimed at individuals, not companies.

173
6.7 DISTINCTION BETWEEN TRADING AND
BANKING

Similar to the preparer survey, users are asked the extent to which they agree (or

disagree) with the assertion that trading and banking books35 are sufficiently different

to warrant the use of different accounting measurment bases. User responses of

strongly oppose, oppose, neutral, support, and strongly support are replaced with the

values 1, 2, 3, 4 and 5 respectively. Then, a one-sample t-test is used to compare the

means to the midpoint position and significant p-values indicate that the means are

significantly different from 3, the neutral response. This method is used throughout

this chapter. As shown in Table 6.5, the JWGBA’s assertion is not substantiated

(mean = 3.2; p = .159) as respondents are neutral in relation to the purported

difference between the two books.

Table 6.5: Level of Users’ Support for the Statement on the Fundamental
Difference between the Trading and Banking Books
Strongly

Strongly

p-value
Support

Support

Means
(n=65)
Neutral
Oppose

Oppose

Question

The banking book is so

fundamentally different from the
trading book to warrant the use of 3 15 20 21 6 3.2 .159
different accounting measurement
bases.
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.

If the 20 user respondents with neutral opinions are excluded, 27 users out of the

remaining 45 (60%) actually support the use of two different accounting

measurement bases. However, the one-sample proportion test36 indicate that 60% is

35
Both the trading and banking categories were separately defined in the user survey (see the actual
user survey in Appendix B and D).
36
The one-sample proportion test is used throughout this chapter to test for difference in the
proportions from the mid-way percentage (50%).

174
not statistically different from 50% (p = 0.233) and thus, users with opinions are

divided on the issue. The JWGBA argument that the two books are distinctly

different has both support (27/65) and opposition (18/65) from users.

6.7.1 Regression Results

The three user descriptors of country, type of user and level of experience are then

fitted in general linear models to examine if any of them are possible predictors. In

addition, the two-way interaction effects between country and each of the other two

predictors are also included in the models to account for possible difference between

countries. Scrutiny of the residual plots for each model fitted confirms that the

underlying assumptions are met. Table 6.6 shows the initial general linear model

fitted followed by the models resulting from the backward elimination process.

Table 6.6: Regression Results for Level of Users’ Agreement with the
Purported Fundamental Difference between the Trading and
Banking Books
Initial Model
Variables in the equation Level of Significance
Country 0.798
Type of User 0.132
Level of Experience 0.135
Country * Type of User 0.268
Country * Level of Experience 0.657
R-square = 0.139; Adjusted R-square = 0.033; p = 0.263
Level 1 – Country * Level of Experience deleted
Country 0.722
Type of User 0.140
Level of Experience 0.147
Country * Type of User 0.254
R-square = 0.136; Adjusted R-square = 0.046; p = 0.189

175
Table 6.6: Regression Results for Level of Users’ Agreement with the
Purported Fundamental Difference between the Trading and
Banking Books (continued)

Variables in the equation Level of Significance

Level 2 – Country * Type of User deleted
Country 0.808
Type of User 0.161
Level of Experience 0.151
R-square = 0.094; Adjusted R-square = 0.033; p = 0.199
Level 3 – Country deleted
Type of User 0.161
Level of Experience 0.126
R-square = 0.093; Adjusted R-square = 0.048; p = 0.112
Level 4 – Type of User deleted
Level of Experience 0.126
R-square = 0.037; Adjusted R-square = 0.021; p = 0.126

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

None of the variables included are found to be significant predictors of users’ support

for the fundamental different between banking and trading books argument.

6.8 QUALITATIVE CHARACTERISTICS

The support (and/or resistance) towards fair value accounting is further examined

with questions that assess the perceived usefulness of fair value accounting in

relation to the qualitative characteristics in the IASB’s Framework (2001). Users are

asked to indicate their level of agreement with three statements that pertain to

qualitative characteristics that fair value accounting need to possess to be useful for

decision making purposes. Table 6.7 shows users’ perception of the relevance and

comparability of information resulting from fair value accounting.

176
Table 6.7: Level of Users’ Perception of the Relevance and Comparability of
Fair Value Accounting

Disagree

Disagree

p-value
Strongly

Strongly

Means
(n=65)
Neutral

Agree

Agree
Fair value accounting provides
information that help users evaluate 0 8 23 26 8 3.5*** .000
past, present and/or future events.
Fair value accounting enables users
to compare financial statements of 0 8 36 19 2 3.2* .010
different entities
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The average response of 3.5 is statistically significantly different from the neutral

response. This indicated that users perceived fair value accounting for financial

instruments results in relevant information. However, 23 users have neutral

opinions. Once excluded, 81% (p = 0.000 in one-sample proportion test) of users

perceive that fair value accounting provides information that help users evaluate past,

present and/or future events, confirming its relevance as per the IASB framework’s

definition.

Users seem to agree that the financial statements are comparable across entities with

an average response of 3.2 which is moderately statistically different from the neutral

position according to the one-sample t-test. This is reinforced by the one-sample

proportion test conducted with neutral users excluded which revealed 72%

(statistically significantly different from 50% with p = 0.026) of users believe the

financial statements to be comparable across entities. Therefore, of users with a non-

neutral opinion, most tend to perceive financial statements prepared using fair value

accounting to be comparable across entities.

177
Users are then asked to rate the reliability of the fair values of traded and non-traded

financial instruments. User responses are shown in Table 6.8.

Table 6.8: Level of Users’ Perception of the Reliability of Traded and Non-
traded Financial Instruments

Unreliable

Unreliable

Reliable

Reliable
Neutral

p-value
Highly

Highly

Means
(n=65)
Type of financial instrument

Traded financial instruments 0 2 7 24 32 4.3*** .000

Non-traded financial instruments 2 27 22 14 0 2.7** .014

Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

User responses indicate that the fair values of traded financial instruments are

perceived to be reliable (mean = 4.3) in line with expectations of high reliability. In

fact, of the 58 users with an opinion on this, 97% of them perceive these fair values

to be reliable (statistically significantly different from 50% as indicated by a p-value

of 0.000).

In terms of non-traded financial instruments, a mean of 2.7 (significantly different

from the neutral position) signals that their fair values are perceived to be unreliable

as per JWGBA assertions. Further examination of this reveals that of the 43 users

with an opinion on the matter, 67% (p = 0.033) perceive the fair values of non-traded

financial instruments to be unreliable. However, there is still a lack of consensus on

the achievable reliability of fair values for financial instruments in the banking book

as 14 (of 65) users believe they can be reliably measured.

178
The following sections further examine these qualitative characteristics and possible

influence by country, type of user and level of experience together with the two-way

interactions between country and the other two variables.

6.8.1 Regressions for Relevance of Fair Value Accounting

According to the IASB (1989) “Framework for the preparation and presentation of

financial statements”, information must be relevant to the decision making needs of

users by helping them evaluate past, present or future events. Table 6.9 shows the

general linear models fitted to find possible determinants of users’ perception of the

relevance of fair value accounting. Residual plots for each model fitted are checked

to ensure the underlying assumptions are met.

Table 6.9: Regression Results for Users’ Perception of the Relevance of Fair
Value Accounting for all Financial Instruments

Initial Model
Variables in the equation Level of Significance
Country 0.963
Type of User 0.025**
Level of Experience 0.466
Country * Type of User 0.431
Country * Level of Experience 0.734
R-square = 0.174; Adjusted R-square = 0.072; p = 0.125
Level 1 – Country * Level of Experience deleted
Country 0.523
Type of User 0.025**
Level of Experience 0.343
Country * Type of User 0.440
R-square = 0.172; Adjusted R-square = 0.086; p = 0.079
Level 2 – Country * Type of User deleted
Country 0.701
Type of User 0.022**
Level of Experience 0.261
R-square = 0.148; Adjusted R-square = 0.091; p = 0.044

179
Table 6.9: Regression Results for Users’ Perception of the Relevance of Fair
Value Accounting for all Financial Instruments (continued)

Variables in the equation Level of Significance

Level 3 – Country deleted
Type of User 0.022**
Level of Experience 0.126
R-square = 0.146; Adjusted R-square = 0.104; p = 0.021
Level 4 – Level of Experience deleted
Type of User 0.025**
R-square = 0.112; Adjusted R-square = 0.084; p = 0.025

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Type of user was the only statistical significant predictor of users’ perceptions of the

relevance of fair value accounting information.

6.8.1.1 Type of User Effect

The influence of type of user is depicted by the marginal means in Table 6.10. This is

also graphically depicted by a bar chart to distinguish between the average response

in relation to user percption on the relevance of fair value accounting (Figure 6.2).

Table 6.10: Users’ Perception of the Relevance of Fair Value Accounting by

Type of User
Type of User n Mean Std. Error
Assurance and Advisory 15 3.20 .215
Analyst 25 3.88 .166
Other 25 3.36 .166
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5

The analyst group of users agrees that fair value accounting results in relevant

information more than the other groups of users even though the preliminary phone

survey (see Section 4.3.1) suggested otherwise.

180
Figure 6.2: Users’ Perception of the Relevance of Fair Value Accounting by
Type of User

Average Response
4

0
Assurance and Advisory Analyst Other
Type of User

It seems that ‘other’ users are not as supportive of the view that fair value accounting

provides relevant information. Perhaps the time analysts spend on analysing

financial statements make them really appreciate and understand the benefits of fair

value accounting.

6.8.2 Regressions for Reliability of Fair Values

As before, country, type of user and level of experience are fitted in general linear

models to find possible determinants of users’ perception of the relevance of fair

value accounting. The interaction between country and type of user as well as

country and level of experience are also included. Residual plots for each model

fitted are checked to ensure the underlying assumptions are met.

6.8.2.1 Traded Financial Instruments

The regression (and backward elimination) results checking for possible

determinants of user views on the reliability of traded financial instruments are

shown in Table 6.11.

181
Table 6.11: Regression Results for Users’ Perception of the Reliability of Fair
Values for Traded Financial Instruments

Initial Model
Variables in the equation Level of Significance
Country 0.568
Type of User 0.027**
Level of Experience 0.104
Country * Type of User 0.390
Country * Level of Experience 0.256
R-square = 0.171; Adjusted R-square = 0.070; p = 0.131
Level 1 – Country * Type of User deleted
Country 0.569
Type of User 0.024**
Level of Experience 0.150
Country * Level of Experience 0.313
R-square = 0.144; Adjusted R-square = 0.071; p = 0.095
Level 2 – Country * Level of Experience deleted
Country 0.376
Type of User 0.025**
Level of Experience 0.254
R-square = 0.129; Adjusted R-square = 0.071; p = 0.078
Level 3 – Country deleted
Type of User 0.030**
Level of Experience 0.420
R-square = 0.117; Adjusted R-square = 0.074; p = 0.054
Level 4 – Level of Experience deleted
Type of User 0.029**
R-square = 0.107; Adjusted R-square = 0.079; p = 0.029

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The regression models show the type of user to be the only significant predictor of

user perception of the reliability of traded financial instruments’ fair values.

182
6.8.2.1.1 Type of User Effect

The estimated marginal means as shown in Table 6.12 details the influence of type of

user on users’ perception of the reliability of fair values for traded financial

instruments.

Table 6.12: Users’ Perception of the Reliability of Fair Values for Traded
Financial Instruments by Type of User
Type of User n Mean Std. Error
Assurance and Advisory 15 4.47 .196
Analyst 25 4.56 .152
Other 25 4.00 .152
Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

Figure 6.3: Users’ Perception of the Reliability of Fair Values of Traded

Financial Instruments by Type of User

4
Average Response

0
Assurance and Advisory Analyst Other
Type of User

Users in the ‘other’ category (fund managers, trust managers and industry

associations’ representatives) tend to perceive fair values of traded financial

instruments to be less reliable than the other two groups of users. Auditors and

analysts have a higher opinion of the reliability of fair values of traded financial

instruments with average responses of 4.47 and 4.56.

183
6.8.2.2 Non-traded Financial Instruments

As before, general linear models are fitted to check for possible determinants of user

views on the reliability of non-traded financial instruments. Results are shown in

Table 6.13.

Table 6.13: Regression Results for Users’ Perception of the Reliability of Fair
Values for Non-traded Financial Instruments

Initial Model
Variables in the equation Level of Significance
Country 0.284
Type of User 0.259
Level of Experience 0.238
Country * Type of User 0.051*
Country * Level of Experience 0.521
R-square = 0.201; Adjusted R-square = 0.102; p = 0.065
Level 1 – Country * Level of Experience deleted
Country 0.237
Type of User 0.272
Level of Experience 0.309
Country * Type of User 0.043**
R-square = 0.195; Adjusted R-square = 0.111; p = 0.043
Level 2 –Level of Experience deleted
Country 0.048**
Type of User 0.258
Country * Type of User 0.043**
R-square = 0.180; Adjusted R-square = 0.111; p = 0.035

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The backward elimination process reveals the interaction between country and type

of user to be a significant predictor of user perception of the reliability of non-traded

financial instruments’ fair values. However, as found in Section 5.3, country seems

to be a partial surrogate for level of experience. Level of experience has the most

184
insignificant p-value at level 1 and is eliminated. Thus, in this instance, country is

the more significant predictor of users’ perception.

6.8.2.2.1 Country by Type of User Effect

Figure 6.4 distinguishes between the average response in relation to the reliability of

fair values for non-traded financial instruments between the two countries. The

influence of the interaction between country and type of user is depicted also by the

marginal means in Table 6.14.

Figure 6.4: Users’ Perception of the Reliability of Fair Values of Non-traded

Financial Instruments by Country by Type of User

5
Average Response

3 Australia
2 Singapore

0
Assurance and Analyst Other
Advisory
Type of User

Table 6.14: Users’ Perception of the Reliability of Fair Values for Non-traded
Financial Instruments by Country by Type of User
Country n Type of User Mean Std. Error
Australia 8 Assurance and Advisory 3.25 .278
11 Analyst 2.64 .237
14 Other 2.93 .210
Singapore 7 Assurance and Advisory 2.57 .297
14 Analyst 2.93 .210
11 Other 2.09 .237
Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

185
Singaporean users in the ‘other’ category perceive these fair values to be unreliable

with an average response of 2.09. The remaining users tend around the neutral point

in relation to the reliability of such instruments. Singaporean users in the ‘other’

category are made up of fund and trust managers. Perhaps their extensive dealings in

non-traded financial instruments could explain their lower opinion of reliability.

6.8.3 Regressions for Comparability of Fair Value Information

Users’ average support for the comparability of financial statements between entities

is further examined with general linear models. Country, type of user and level of

experience are fitted in general linear models to find determinants of users’

perception of the comparability of financial statements prepared using fair value

accounting (see Table 6.15). The interaction between country and type of user as

well as country and level of experience are also included. Residual plots for each

model fitted are checked to ensure the underlying assumptions are met.

Table 6.15: Regression Results for Users’ Perception of the Comparability of

Fair Value Accounting

Initial Model
Variables in the equation Level of Significance
Country 0.416
Type of User 0.933
Level of Experience 0.253
Country * Type of User 0.344
Country * Level of Experience 0.209
R-square = 0.177; Adjusted R-square = 0.076; p = 0.114
Level 1 – Country * Type of User deleted
Country 0.483
Type of User 0.883
Level of Experience 0.293
Country * Level of Experience 0.208
R-square = 0.146; Adjusted R-square = 0.074; p = 0.089

186
Table 6.15: Regression Results for Users’ Perception of the Comparability of
Fair Value Accounting (continued)

Variables in the equation Level of Significance

Level 2 – Country * Level of Experience deleted
Country 0.258
Type of User 0.930
Level of Experience 0.097*
R-square = 0.123; Adjusted R-square = 0.064; p = 0.093
Level 3 – Type of User deleted
Country 0.236
Level of Experience 0.095*
R-square = 0.120; Adjusted R-square = 0.092; p = 0.019
Level 4 – Country deleted
Level of Experience 0.010**
R-square = 0.100; Adjusted R-square = 0.086; p = 0.010

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The backward elimination process, where the least significant variable is removed at

each step is again conducted. This process reveals the level of experience to be the

only significant predictor of user perception of the comparability of different entities’

financial statements prepared on fair value accounting basis. However, level of

experience may be a partial surrogate for country which is less significant in level 3

and subsequently eliminated. Here, level of experience is the more significant

predictor of users’ perception.

A regression coefficient of 0.171 indicates that for every 1 unit increase in the level

of experience, user perception of the comparability of fair value accounting financial

statements across entities increases by 0.171. As such, users with more experience

tend to perceive financial statements prepared using fair value accounting more

comparable than their less experienced counterparts. Perhaps more experienced

187
users have a greater capacity for understanding the ramifications of the complex fair

value accounting and, thus, are better equipped to compare different financial

statements.

6.9 CONCERNS REGARDING VOLATILITY

As previously mentioned, another main theme identified in this fair value accounting

debate is the purported volatility introduced by the use of fair value accounting for

all financial instruments. Users are asked their opinion on this claim (see Table

6.16). The average score of user responses and the significant p-value shows that

users tend to agree with the JWGBA position. They feel that the volatility resulting

from the use of fair value accounting for all financial instruments will be

misunderstood by users.

Table 6.16: Level of Users’ Perception of the Potential Problem with Volatility
in Reported Earnings
Disagree
Disagree
Strongly

Strongly
Neutral

p-value
Means
(n=65)
Agree

Agree

Statement

Fair value accounting for all

financial instruments introduces 4 12 14 32 3 3.3** .033
volatility that will be
misunderstood by users.
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

There is still considerable variance among users’ perception of the potential

problems the volatility in reported earnings from the use of fair value accounting

would pose. In general, of the users with an opinion on the matter, 69% perceive that

the volatility in earnings will be misunderstood by users, which is significantly

different to 50% (p = 0.012). This finding indicates agreement with JWGBA

188
concerns on the understandability of the earnings volatility resulting from the use of

fair value accounting.

6.9.1 Regressions on Concerns Regarding Volatility

The descriptors of country, type of user and level of experience and the two-way

interactions between country and each of the other two predictors are fitted in general

linear models to determine possible predictors. Residual plots are checked and

underlying assumptions are met. Table 6.17 shows the initial general linear model

fitted followed by the models resulting from the backward elimination process.

Table 6.17: Regression Results for Users’ Perception of Volatility in Reported

Earnings from the Use of Fair Value Accounting

Initial Model
Variables in the equation Level of Significance
Country 0.990
Type of User 0.867
Level of Experience 0.573
Country * Type of User 0.044**
Country * Level of Experience 0.674
R-square = 0.131; Adjusted R-square = 0.024; p = 0.302
Level 1 – Country * Level of Experience deleted
Country 0.371
Type of User 0.849
Level of Experience 0.663
Country * Type of User 0.039**
R-square = 0.128; Adjusted R-square = 0.038; p = 0.221
Level 2 – Level of Experience deleted
Country 0.431
Type of User 0.835
Country * Type of User 0.035**
R-square = 0.126; Adjusted R-square = 0.051; p = 0.150

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

189
As shown in the backward elimination process, the interaction between country and

type of user is the only significant predictor of user perception that volatility in

reported earnings would be misunderstood by users. Although insignificant, the

country and type of user variables cannot be eliminated from the model because they

form the significant variable of interaction between country and type of user. As

found in Section 5.3, country seems to be a partial surrogate for level of experience.

In this situation, level of experience has the most insignificant p-value and, thus,

country turns out to be the more significant influence.

6.9.1.1 Country by Type of User Effect

Figure 6.5 graphically distinguishes between the average user responses and Table

6.18 shows the influence of the interaction between country and type of user via the

mean response.

Figure 6.5: Users’ Perception that the Volatility in Earnings Introduced by

Fair Value Accounting will be Misunderstood by Users by Country
by Type of User

5
Average Response

3 Australia

2 Singapore

0
Assurance and Analyst Other
Advisory
Type of User

The bar chart (Figure 6.5) also shows that auditors from both countries have similar

perceptions. Australian users in the ‘other’ category tend around the neutral position,

whereas their Singaporean counterparts are more concerned with the potential

190
problems of volatility in earnings. Conversely, Australian analysts expressed

stronger concerns about volatility misrepresentations than Singaporean analysts.

Table 6.18: Users’ Perception that the Volatility in Earnings Introduced by

Fair Value Accounting will be Misunderstood by Users by Country
by Type of User
Country n Type of User Mean Std. Error
Australia 8 Assurance and Advisory 3.37 .352
11 Analyst 3.73 .300
14 Other 3.14 .266
Singapore 7 Assurance and Advisory 3.29 .377
14 Analyst 2.71 .266
11 Other 3.64 .300
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5

As reported in Table 6.18, 14 Singaporean analysts stand out as the group of users

that are least concerned (2.71 out of 5) with the problems that may arise from the

volatility in reported earnings as a result of using fair value accounting. This is a big

contrast to Australian analysts who are more concerned (3.73 out of 5) with potential

user misunderstanding due to the volatility in reported earnings.

6.10 SUPPORT FOR FAIR VALUE ACCOUNTING

In the user survey, fair value accounting is specifically defined as the measurement

of all financial instruments at fair value and the recognition of changes in fair value

as revenues or expenses in the Income Statement in the period in which they arise.

Sophisticated users are asked the extent they support (or oppose) fair value

accounting for all financial instruments. The responses are indicated in Table 6.19.

191
Table 6.19: Users’ Support for Fair Value Accounting for All Financial
Instruments

Support

Strongly
Strongly

Support
Neutral

p-value
Means
(n=65)
Oppose

Oppose
Question

Do you support the use of fair

value accounting for all financial 1 16 12 20 16 3.5*** .001
instruments?
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The average response of users is a statistically significant 3.5 score indicating

support for comprehensive fair value accounting. In addition, when the 12 users with

neutral opinions are excluded, 68% (p = 0.013 in one-sample proportion test) of users

supports the use of fair value accounting for all financial instruments. At first

glance, respondent users are supportive of fair value accounting for all financial

instruments. However, user preferences are still varied as evidenced by the 17 users

who oppose this approach and the 12 users who are neutral on the matter.

6.10.1 Regression Results

The three user descriptors of country, type of user and level of experience are fitted

in general linear models to find possible determinants of users’ general support for

fair value accounting. As before, country and type of user are inserted as factors

while level of experience is treated as a covariate in all the models and the interaction

between country and type of user as well as country and level of experience are also

included. It is also expected that users’ perception of the banking and trading

distinction, the qualitative characteristics of fair value accounting as well as the

impact of earnings volatility on users’ decisions will affect their preference for fair

value accounting. Therefore, in addition to the five aforementioned variables, user

192
responses for the questions related to the trading/banking distinction, qualitative

characteristics and problems with volatility of reported earnings37 are also included

as covariates in the initial general linear model fitted. Residual plots for each model

fitted are checked to ensure the underlying assumptions are met.

Table 6.20: Regression Results for Users’ Preference for Fair Value Accounting
for all Financial Instruments

Initial Model
Variables in the equation Level of Significance
Country 0.198
Type of User 0.843
Level of Experience 0.073*
Country * Type of User 0.921
Country * Level of Experience 0.302
No distinction between trading and banking book 0.009***
Support for Relevance 0.139
Support for Comparability 0.899
Support for Reliability of Traded Financial Instruments 0.666
Support for Reliability of Non-traded Financial Instruments 0.068*
Volatility will not be misunderstood 0.002***
R-square = 0.612; Adjusted R-square = 0.514; p = 0.000
Level 1 – Country * Type of User deleted
Country 0.184
Type of User 0.808
Level of Experience 0.058*
Country * Level of Experience 0.295
No distinction between trading and banking book 0.008***
Support for Relevance 0.113
Support for Comparability 0.915
Support for Reliability of Traded Financial Instruments 0.686
Support for Reliability of Non-traded Financial Instruments 0.066*
Volatility will not be misunderstood 0.002***
R-square = 0.611; Adjusted R-square = 0.531; p = 0.000

37
The data for variables where the questions were expressed in the negative, against fair value
accounting, are recoded to be aligned with other variables. Therefore, the descriptives provided in
previous sections are complete opposite.

193
Table 6.20: Regression Results for Users’ Preference for Fair Value Accounting
for all Financial Instruments (continued)

Level 2 – Country * Level of Experience deleted

Country 0.360
Type of User 0.703
Level of Experience 0.011**
No distinction between trading and banking book 0.010**
Support for Relevance 0.111
Support for Comparability 0.699
Support for Reliability of Traded Financial Instruments 0.885
Support for Reliability of Non-traded Financial Instruments 0.088*
Volatility will not be misunderstood 0.002***
R-square = 0.603; Adjusted R-square = 0.529; p = 0.000
Level 3 – Support for Reliability of Traded Financial Instruments deleted
Country 0.334
Type of User 0.704
Level of Experience 0.009***
No distinction between trading and banking book 0.009**
Support for Relevance 0.109
Support for Comparability 0.668
Support for Reliability of Non-traded Financial Instruments 0.086*
Volatility will not be misunderstood 0.001***
R-square = 0.603; Adjusted R-square = 0.538; p = 0.000
Level 4 – Type of User deleted
Country 0.335
Level of Experience 0.008***
No distinction between trading and banking book 0.003**
Support for Relevance 0.139
Support for Comparability 0.646
Support for Reliability of Non-traded Financial Instruments 0.056*
Volatility will not be misunderstood 0.001***
R-square = 0.598; Adjusted R-square = 0.548; p = 0.000
Level 5 – Support for Comparability deleted
Country 0.365
Level of Experience 0.006***
No distinction between trading and banking book 0.002**
Support for Relevance 0.092*
Support for Reliability of Non-traded Financial Instruments 0.051*
Volatility will not be misunderstood 0.001***
R-square = 0.596; Adjusted R-square = 0.554; p = 0.000

194
Table 6.20: Regression Results for Users’ Preference for Fair Value Accounting
for all Financial Instruments (continued)

Level 6 – Country deleted

Level of Experience 0.008***
No distinction between trading and banking book 0.002**
Support for Relevance 0.090*
Support for Reliability of Non-traded Financial Instruments 0.067*
Volatility will not be misunderstood 0.000***
R-square = 0.590; Adjusted R-square = 0.556; p = 0.000
Level 7 – Support for Relevance deleted
Level of Experience 0.006***
No distinction between trading and banking book 0.001**
Support for Reliability of Non-traded Financial Instruments 0.017**
Volatility will not be misunderstood 0.000***
R-square = 0.570; Adjusted R-square = 0.541; p = 0.000

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Although the elimination process could have been stopped at level 6, an extra

elimination is performed as support for relevance with a p-value of 0.09, though still

moderately significant, is rather insignificant. Therefore, according to the final

model: (1) level of experience, (2) no distinction between the trading and banking

book, (3) reliability of non-traded financial instruments and (4) volatility will not be

misunderstood influence users’ preference for fair value accounting are significant

predictors of users preference for fair value accounting. It is crucial to note that level

of experience seems to be a partial surrogate for country, the least significant

variable eliminated at level 6. Thus, in terms of users’ support for fair value

accounting, level of experience is more influential than country.

195
6.10.1.1 Level of Experience Effect

According to the regression coefficient of 0.226, for every 1 unit increase in the level

of experience, user support for fair value accounting increases by 0.226. Therefore,

users with more experience tend to support fair value accounting for all financial

instruments more than their less experienced counterparts. The complexity of fair

value accounting and financial instruments in general could be the reason for this.

Perhaps more experienced users are more familiar with the complexity and thus more

comfortable with the change to fair value accounting for all financial instruments.

6.10.1.2 No Distinction between Banking and Trading Book Effect

Users who perceive that the trading book is not fundamentally different from the

banking book tend to prefer fair value accounting as compared to their counterparts.

The regression coefficient for this variable is 0.351, indicating that for every 1-unit

increase in the perception of no difference between the two books, user average

support for fair value accounting increases by 0.351. It seems that so long as these

two books are seen to be fundamentally different, there will be continued opposition

to fair value accounting for all financial instruments.

6.10.1.3 Reliability of Non-traded Financial Instruments Effect

Respondent users that perceive fair values of non-traded financial instruments to be

reliable tend to support fair value accounting for all financial instruments. A

regression coefficient of 0.341 shows that for every 1-unit increase in the perception

of the reliability of fair values for non-traded financial instruments, the average

support for fair value accounting increases by 0.341. This lends weight to JWGBA

assertions of problems with obtaining reliable fair values for financial instruments

196
that have no active markets. Therefore, it seems that users are more likely to support

the use of fair value accounting for all financial instruments if mechanisms for

calculating reliable fair values for non-traded financial instruments can be

established.

6.10.1.4 Volatility Will Not be Misunderstood Effect

Users with the view that the volatility in reported earnings resulting from the use of

fair value accounting for all financial instruments will not be misunderstood by users

tend to support its use. A regression coefficient of 0.408 shows that for every 1-unit

increase in the perception that the resultant volatility will not be misunderstood,

average user support for fair value accounting increases by 0.408. The descriptives

in Section 6.8 reveal users to be concerned that the resulting volatility will be

misunderstood; giving some support to the JWGBA assertions. It seems that users

are concerned with the volatility introduced by fair value accounting and will only

support its use if users will not misunderstand the effects of its use.

6.11 USER VIEWS ON KEY THEMES: SUMMARY

The findings in relation to users’ perceptions of the key themes previously identified

are summarised in Table 6.21.

197
Table 6.21: Summary of Users’ Perceptions of the Four Key Themes

Item Average Non—Neutral % Predictor Direction

Support Oppose (p-value)

Trading and banking Neutral 60% 40% None

distinction (3.2)
Relevance of fair value Agree 81% 19% Type of User (0.025) Analysts perceive it to be more relevant than users in the
accounting (3.5) other category followed by auditors
Reliability of fair values Reliable 97% 3% Type of User (0.029) Analysts, closely followed by auditors while users in the
for traded financial (4.3) other category have the lowest perception of the reliability
instruments of fair values for traded financial instruments
Reliability of fair values Unreliable 33% 67% Country by Type of User (0.043) Singaporean auditors and users in the other category have
for non-traded financial (2.7) statistically lower opinion on the reliability of fair values
instruments for non-traded financial instruments
Comparability of fair Agree 72% 28% Level of Experience (0.010) User perception of comparability of fair value accounting
value accounting (3.2) increases with the level of experience
Volatility may be Agree 69% 31% Country by Type of User (0.035) Singaporean analysts are statistically more concerned about
misunderstood by users (3.3) the volatility than their Australian counterparts
Overall support for fair Support 68% 32% Level of Experience (0.006) Support for fair value accounting increases with experience
value accounting (3.5)
No Distinction between trading and Support for fair value accounting increases with the
banking (0.001) perception of no difference between two books
Support for Reliability of non-traded Support for fair value accounting increases with support for
instruments (0.017) reliability of non-traded instruments
Volatility will not be misunderstood Support for fair value accounting increases with perception
(0.000) that volatility will not be misunderstood

198
The average responses (and the 60% of user supporters) show general support for fair

value accounting for all financial instruments. The only contradiction is users’

perception that fair values for non-traded financial instruments are unreliable. These

results further highlight the lack of consensus in user preference with noticeable

support for arguments from both sides of the debate (JWGSS and JWGBA).

In addition, the findings reveal that users will support fair value accounting so long

as there is no perceived difference between the banking and trading books, fair

values of non-traded financial instruments are reliable and volatility will not be

misunderstood. It is also found that user experience increases the level of support for

the proposed fair value accounting model indicating that the accounting for financial

instruments remains a complex issue where experience increases users’

comprehension of all the issues.

6.12 USER VIEWS ON OTHER ISSUES

The IASB framework states that the objective of financial statements is to provide

useful information for users. In order to be useful, information provided in financial

statements must possess certain qualitative characteristics. The main qualitative

characteristics of relevance, reliability and comparability have been examined in

previous sections. However, in keeping with the exploration of the usefulness of fair

value accounting, respondents are asked further questions in relation to other aspects

of qualitative characteristics necessary for information to be useful.

199
6.12.1 General Usefulness of Fair Value Accounting

The survey asks users to indicate their level of agreement with four statements on the

general usefulness of fair value accounting for all financial instruments. Their

responses are shown in Table 6.22.

Table 6.22: Users’ Level of Agreement with General Usefulness of Fair Value
Accounting for All Financial Instruments
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5

Disagree
Disagree
Strongly

Strongly

p-value
Means
(n=65)
Neutral

Agree

Agree
Fair value accounting for all
financial instruments provides 1 8 16 26 14 3.7*** .000
useful information for economic
decision making.
Fair value accounting ensures that
financial statements show a true
and fair view of the financial 0 13 12 28 12 3.6*** .000
position, performance and changes
in financial position of an entity.
The benefits of using fair value
accounting for all financial 2 11 24 23 5 3.3** .021
instruments outweigh the costs.

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The average user response indicates relative agreement (means of 3.7, 3.6 and 3.3)

that fair value accounting provides useful information, ensures financial statements

present a true and fair view and that the benefits of using fair value accounting

outweigh the costs. However, 18% to 37% of users have a neutral opinion in those

three instances.

First, when users with neutral opinions are excluded, 82% (statistically different from

50% with a p-value = 0.000) of the remaining users perceive that fair value

accounting provides useful information for economic decision making as compared

200
with only the 18% who do not. This signals support for fair value accounting being

useful to users for decision making. Second, 75% (different from 50%; p-value =

0.000) of user respondents perceive that fair value accounting results in financial

statements that show a true and fair view lending weight to JWGSS assertions that

fair value accounting is appropriate. Third, 68% (statistically different from 50%

with a p-value of 0.029) of users with non-neutral opinions perceive that the benefits

of using fair value accounting outweigh the costs.

Although there is variation in the responses, overall, it seems that sophisticated users

are of the opinion that fair value accounting would result in financial statements that

are useful to their decision making needs and that will achieve the objective of

financial reporting.

6.12.2 More on Relevance

Users are also asked their opinion on the relevance of fair value accounting for six

categories of financial instruments. The results of which are discussed in Section

6.12.2.1. The IASB Framework, in addition to the general definition of relevance,

stipulates that the relevance of financial information is affected by its nature and

materiality. Subsequently, the user survey also asks two questions on the materiality

of financial statements resulting from the use of fair value accounting. These results

are shown in Section 6.12.2.2.

6.12.2.1 Fair Value Accounting for Different Categories

User responses indicate that fair value accounting is generally relevant for both

trading derivatives and securities whose fair values are generally viewed as highly

reliable. Except for two users, fair value accounting is uniformaly regarded as

201
relevant for the financial instruments in the trading book. However, the fact that

some users perceive it to be only ‘sometimes relevant’ shows lack of complete

consensus.

Table 6.23: The Extent Users Regard Fair Value Accounting as Relevant for
Different Categories of Financial Instruments

Category Type of Financial Instruments Never Sometimes Always

Trading Trading Derivatives 2 15 48

Trading Securities 0 16 49

Banking Hedging Derivatives 14 24 27

Investment Securities 10 25 30

Loans and receivables 14 35 16

Deposits and Payables 14 35 16

As per JWGBA assertions, 15% to 21% (10/65 to 14/65) of users regard fair value

accounting as not relevant for financial instruments in the banking book. However,

there are at least 16 (of 65) users who indicate that fair value accounting is always

relevant for the banking book in line with JWGSS views. The majority of users

perceive it to be sometimes relevant. Again, there is a high variability in user

preferences when it comes to fair value accounting for the banking book.

6.12.2.2 Materiality

Users are also asked their level of agreement with two statements on the nature and

materiality of fair value accounting information. As per the IASB Framework, the

relevance of information is affected by its nature and materiality. Average responses

of 3.6 to 3.8 (see Table 6.24), all statistically significantly different from the neutral

position, indicate that users perceive fair value accounting to result in material, and

therefore, relevant information.

202
Table 6.24: Level of Users’ Perception of the Materiality of Information
Resulting from Fair Value Accounting

Disagree
Disagree
Strongly

Strongly

p-value
Means
(n=65)
Neutral

Agree

Agree
The nature of the information
resulting from fair value
accounting affects the assessment 0 5 15 36 9 3.8*** .000
of the risks and opportunities
facing the entity
The omission of fair values can
cause the financial statements to 1 3 19 38 4 3.6*** .000
be false or misleading
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Users with a neutral opinion are again excluded to conduct further analysis. This

results in 87% and 88% (both different from 50%; p-values = 0.000) of users

perceiving fair value accounting information to be of an important nature and its

omission can cause financial statements to be false or misleading. There is a

noticeable variance in the opinion of users in both instance and substantial neutral

opinions. However, on average, users perceive fair value accounting to result in

information that is relevant as per the IASB framework’s definition.

6.12.3 More on Reliability

The IASB framework states that information must be reliable to be useful and

defines reliable information to be “free from material error and bias and can be

depended upon by users to represent faithfully that which it either purports to

represent or could reasonably be expected to represent” (IASB, 1989, para 31, p. 49).

In the user survey, respondents are asked their level of agreement with the two

203
criteria of reliability in relation to fair value accounting for all financial instruments.

The responses are shown in Table 6.25.

Table 6.25: Level of Users’ Perception of the Reliability of Fair Value

Accounting for All Financial Instruments

Disagree

Disagree
Strongly

Strongly

p-value
Means
(n=65)
Neutral
Statements

Agree

Agree
Information resulting from fair
value accounting is free from 4 28 26 7 0 2.6*** .000
material error and bias
Fair value accounting represents the
substance and economic reality of 0 9 19 34 3 3.5*** .000
the underlying transaction
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Exclusion of neutral respondents shows 82% (significantly different from 50%) of

users believe that fair value accounting results in information that contains material

error and bias. This finding is in line with JWGBA concerns. However, 83% (p-

value = 0.000) of respondents agreed that fair value accounting would represent the

substance and economic reality of the underlying transactions. Given that these two

statements are integral to the definition of reliable information (according to the

IASB framework, they are different elements that reliable information must contain),

the contrasting user opinions raise interesting questions about the definition of

reliability. Users are of the opinion that information resulting from fair value

accounting contains material errors but feel that fair value accounting represents the

substance of the underlying transaction. A possible explanation is that fair value

accounting represents the substance and economic reality of the transactions but is

not sufficiently free from material errors and bias in the opinion of respondent

sophisticated users. Alternatively, users may perceive information resulting from

204
fair value accounting to have more material errors and bias relative to existing

accounting practice.

6.12.4 Understandability and Comparability Over Time

Users are also asked their perception of the understandability of fair value

accounting. Their responses are listed in Table 6.26. The average score of 3.6

(statistically significant at the 0.01 level) indicates that user respondents perceive fair

value accounting in financial statements to be understandable. Exclusion of the 13

neutral users results in 81% of users perceiving that fair value accounting is

understandable by users with reasonable knowledge of accounting; contradicting the

JWGBA assertions.

Table 6.26: Level of Users’ Perception of the Understandability and

Comparability Over Time of Fair Value Accounting in Financial
Statements
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5
Disagree

Disagree
Strongly

Strongly
Neutral

p-value
Means
(n=65)
Agree

Agree

JWGBA Views

Fair value accounting in financial

statements is understandable by 0 10 13 32 10 3.6*** .000
users with reasonable knowledge of
business activities and accounting.
Fair value accounting enables users
to compare financial statements of 0 10 27 22 6 3.2 .184
an entity through time

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

However, users are neutral in terms of the comparability of financial statements over

time with average responses of 3.2. Once the neutral users are excluded, 74%

(statistically significantly different from 50% with p = 0.006) of users actually agree

205
that fair value accounting enables users to compare financial statements over time.

Therefore, of users with an opinion on this, most tend to perceive financial

statements resulting from fair value accounting to be comparable across time.

6.12.5 Usefulness of Mixed Measurement

User responses indicate support for fair value accounting thus far in the analyses.

However, when asked about the mixed measurement model, an average response of

3.6, significantly different from the neutral position signals that users do perceive the

current mixed (but separately identified) measurement of financial instruments to be

useful (see Table 6.27).

Table 6.27: Users’ Level of Agreement with Usefulness of Current Mixed

Measurement of All Financial Instruments
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5
Disagree
Disagree
Strongly

Strongly

p-value
Means
(n=65)
Neutral

Agree

Agree

The use of a combination of historical

cost and fair value measurement bases
that are separated in the financial 1 5 17 37 5 3.6*** .000
statements provides useful
information.

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Exclusion of the 17 neutral users results in 88% (42 of 65) of users with the opinion

that the current mixed measurement model used for financial instruments does

provide useful information. Results thus far generally indicate support for fair value

accounting (and subsequently support for JWGSS) and yet, users do perceive the

current mixed measurement model to be useful as well (lending support to JWGBA).

Therefore, although users would prefer fair value accounting for all financial

206
instruments, they feel that the current mixed measurement model provides useful

information.

6.12.6 User Views on Other Issues: Summary

The findings in relation to users’ perceptions of the other issues discussed in this

section are summarised in Table 6.28.

Table 6.28: Summary of Users’ Perceptions of the Other Issues

Item Average Non—Neutral %

Support Oppose
Fair value accounting useful for economic decision Agree 82% 18%
making (3.7)
Fair value accounting ensures true and fair view Agree 75% 25%
(3.6)
Benefits of fair value accounting outweighs its costs Agree 68% 32%
(3.3)
Fair value accounting results in information that is Agree 87% 13%
important in nature (3.8)
Omission of fair values can cause financial statements Agree 88% 12%
to be false or misleading (3.6)
Fair value information free from material error and bias Disagree 18% 82%
(2.6)
Fair value accounting represent substance and economic Agree 83% 17%
reality of underlying transactions (3.5)
Fair value accounting in financial statements is Agree 81% 19%
understandable (3.6)
Fair value accounting enables comparability over time Neutral 74% 26%
(3.2)
The combination of historical cost and fair value Agree 88% 12%
measurement bases provide useful information (3.6)

Overall, there is general support for fair value accounting and the qualitative

characteristics it possesses, yet the current mixed model is also deemed useful. The

results in this chapter show some support for assertions from both the JWGSS and

the JWGBA, however, on balance was closer to the position of the JWGSS.

207
6.13 MORE ON QUALITATIVE CHARACTERISTICS

The IASB Framework provides guidance in terms of defining the qualitative

characteristics or attributes that make financial statement information useful.

According to the framework, these attributes include understandability, relevance,

reliability, comparability and true and fair presentation. Specifically, additional

guidance is provided on determining relevance and reliability. In this thesis’ user

survey, questions addressing the additional criteria set out in the guidance for

relevance and reliability are included as validity checks. This inclusion allows the

testing of alignment between these so-called additional criteria and their relevant

attribute from a sophisticated user perspective, as they are the end-users of financial

information. Results are discussed in the following sub-sections.

6.13.1 Relevance

The IASB Framework states that the relevance of information is influenced by the

nature and materiality of the underlying information. Accordingly, if the nature of

the information is such that it affects the assessment of users’ decision making, then

it is relevant. In addition, information is considered material if its omission can

influence users economic decisions and in turn adds to the relevance of said

information. Questions on these additional characteristics of materiality are included

to assess whether the characteristics are indicative of relevance, from a users

perspective. The correlation between users’ responses to the question on relevance

as defined by the ability to help users evaluate past, present and/or future events and

the two additional criteria are shown in Table 6.29.

208
Table 6.29: Correlation between the Two Materiality Criteria and Relevance
from a Sophisticated User Perspective
Pearson R (p-value)
Nature Omission
Fair value accounting provides information that 0.327*** (0.008) 0.354*** (0.004)
help users evaluate past, present and/or future
events. (relevance)
The nature of the information resulting from 0.748*** (0.000)
fair value accounting affects the assessment of
the risks and opportunities facing the entity
The omission of fair values can cause the
financial statements to be false or misleading

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

All three questions are highly statistically significantly correlated with p-values of

less than 0.008. This is comforting as users perceive the additional criterion set out

by the IASB to affect relevance. Therefore, users agree with the IASB Framework

that materiality is assessed by reference to the nature and relevance of information.

6.13.2 Reliability

The IASB Framework states that reliable information represents the economic

substance of the underlying transaction or event. Similarly, in this thesis, users are

asked their perception of whether fair value accounting represents the substance and

economic reality of the underlying transactions. They are also asked whether fair

value accounting results in information that is free from material error and bias. User

responses to those two questions are compared and are found to be statistically

significantly correlated (Pearson’s r = 0.532; p = 0.000)38. Users agree that fair value

accounting is reliable if it represents the substance and economic reality of a

transaction in line with the IASB Framework.

38
Average user responses to these two statements were different (see Table 6.25) possibly due to a
lack of confidence in fair values being free from material error and bias.

209
6.13.3 Qualitative Characteristics Equals Useful Information?

The IASB Framework states that the four principal qualitative characteristics of

understandability, relevance, reliability and comparability are the attributes that make

financial information useful. Subsequently, user responses to the questions on the

components of the attributes listed in the IASB Framework are examined for

correlation with their response to the question on general usefulness of fair value

accounting information. The correlations are shown in Table 6.30.

The first row of the table is of most interest in the sense that it tests for relationships

between the various attributes of useful information. It also tests whether users

perceive those attributes to be part and parcel of useful information. All the listed

attributes in the IASB Framework are found to be highly statistically correlated with

the overall measure of useful information. This gives credence to the role of the

IASB Framework in that, according to sophisticated users in Australia and

Singapore, it provides guidance on attributes that result in useful information.

Overall, there are frequent significant correlations between the attributes as expected

according to the IASB Framework. There are only five instances of a lack of

correlation between a few of the attributes of useful information (as highlighted in

the table). It seems that materiality of information is not related to its

understandability. According to user responses, reliability of information is not

related to its materiality and comparability across entities and materiality of

information does not affect its comparability. However, this is expected as the

Framework treats these attributes as mutually exclusive.

210
Table 6.30: Correlations between the Various Attributes of Usefulness and Overall Usefulness of Accounting Information: Based on User
Responses

Pearson R (p-value)
True and Understand- Relevance Nature Reliability Substance Materiality Compare- Compare-
Fair able over Form Entities Time
Useful Information .780*** .481*** .449*** .470*** .478*** .650*** .321*** .614*** .463***
(.000) (.000) (.000) (.000) (.000) (.000) (.000) (.000) (.000)
True and Fair .430*** .420*** .402*** .448*** .650*** .279** .406*** .440***
(.000) (.001) (.001) (.000) (.000) (.024) (.001) (.000)
Understandable .351*** .455*** .388*** .297** .103 .462*** .416***
(.004) (.000) (.001) (.016) (.416) (.000) (.001)
Relevance .327*** .261** .404*** .354*** .513*** .414***
(.008) (.036) (.001) (.004) (.000) (.001)
Nature .278** .779*** .748*** .435*** .357***
(.025) (.000) (.000) (.000) (.004)
Reliability .532*** .063 .206 .251**
(.000) (.620) (.100) (.044)
Substance over form .711*** .357*** .417***
(.000) (.003) (.001)
Materiality . .095 .046
(.452) (.715)
Compare-Entities .660***
(.000)

* Moderately Significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

211
6.14 USER VIEWS: SUMMARY

The evidence in this chapter supports the proposed fair value accounting model from

the users’ point of view giving support to JWGSS assertions. However, there are

instances where users agree with the JWGBA, for example, that fair value

accounting may introduce confusing volatility and lacks reliability. At least 68% of

users also believe the benefits of fair value accounting outweighed the costs. This

perception is crucial in the move to fair value accounting.

Overall, users perceive financial statements prepared using fair value accounting as

relevant for both trading and banking financial instruments (although to a lesser

extent for the latter) and agree that it contains characteristics of useful information.

Users feel that the fair values of traded financial instruments are reliable but those for

non-traded financial instruments are not. Results also indicate that users understand

fair value accounting in financial statements but are neutral in terms of the

comparability of those financial statements between entities and over time.

Respondent users; views are consistent with the JWGBA on the possible

misunderstanding of the volatility resulting from the use of fair value accounting for

all financial instruments (see Table 6.21). In addition, user responses on the

qualitative characteristics of fair value information and their overall perception of

fair value accounting provide assurance that the IASB Framework’s definitions and

qualitative criteria have merit.

Regressions conducted reveal that level of experience, perception of the distinction

between the banking and trading book, support for reliability of non-traded financial

instruments and perception of the ensuing volatility in reported earnings influence

users’ preference for fair value accounting. More experienced users tend to prefer

212
fair value accounting more than their less experienced peers. In addition, users who

perceive no difference between the banking and trading books, believe fair values of

non-traded financial instruments are reliable and that perceived volatility will not be

misunderstood tend to support fair value accounting. There are instances where

Australian users tend to be more supportive of the fair value accounting model. This

support may be explained by a very mature accounting standard process and a well

developed due process in Australia as compared to Singapore (Soewarso, Tower,

Hancock and Taplin, 2003). Australia’s active involvement in the standard setting

process both locally and internationally may have led to a better dissemination

amongst users of the workings and the ramifications of using fair value accounting

for all financial instruments.

The evidence in this chapter shows that most user respondents would use fair value

accounting information if it is provided. Results show general support for fair value

accounting for all financial instruments although there are varied responses for each

theme discussed. This finding provides further support of the need for more research

into this complex area of accounting. Although users generally perceive fair value

accounting for all financial instruments results in information useful for decision

making purposes as per the IASB Framework, there is still considerable

disagreement. Responses to the user survey reveal that the current mixed

measurement accounting model provides useful information even though users do

not believe that different measurement bases are warranted for the two groups of

financial instruments. Perhaps it is only perceived as useful because it is the current

model used and fair value accounting, if used, may be considered to be more useful.

The next chapter provides a comparison of user and preparer perceptions. This is

important to determine the level of consensus of the views of these two groups.
213
 CHAPTER SEVEN:
USER VERSUS PREPARER VIEWS

7.1 INTRODUCTION

Chapters Five and Six detailed the evidence gathered from the two surveys

distributed in the quantitative phase to preparers and users respectively. As the last

chapter analysing data contained in the disseminated surveys, this chapter provides

further statistical analysis by comparing the responses of users and preparers to

identical (or similar) questions on the key themes in both surveys (see Figure 3.5).

7.2 PREPARER VERSUS USER EXPERIENCE

It was previously mentioned that the 83 preparer and the 65 user respondents were

surveyed to represent the preparer and users in Australia and Singapore. Most of the

demographics collected for each group are not compared as they are not identical.

However, the level of experience for users and preparers can be compared. Table 7.1

shows the distribution of user and preparer experience as well as the average level of

experience.

Table 7.1: Comparison of Preparer and User Respondents’ Experience

Preparer User
(mean = 3.52) (mean = 2.97)
< 2 years 7 (8%) 5 (8%)
2 - 5 years 16 (19%) 27 (42%)
6 - 10 years 17 (21%) 11 (17%)
11 - 15 years 13 (16%) 9 (13%)
> 15 years 30 (36%) 13 (20%)
Total 83 65
Legend: <2 years = 1; 2-5 years = 2; 6-10 years = 3; 11-15 years = 4; >15 years = 5.

214
The average score of user and preparer respondents indicate that preparers tend to

have a higher level of experience as compared to users. The independent samples t-

test (with equal variances assumed) reveals a p-value of 0.015. This difference in

experience could explain differences in opinions held by preparers and users.

Such a significant statistical difference between the two groups may be an issue if

level of experience is a prominent predictor of any of the key themes examined. In

terms of preparer views, level of experience is not a significant predictor of the

themes examined. Level of experience is barely a moderately significant predictor of

user views on the comparability of financial statements resulting from fair value

accounting but is a significant predictor of users’ overall preference for fair value

accounting for all financial instruments.

7.3 COMPARISON OF PREPARER AND USER VIEWS

User and preparer responses to each key theme examined in Chapters 5 and 6 are

compared in a similar fashion. The corresponding percentages are also displayed to

indicate the distribution of user and preparer perceptions of each of the key themes.

These are distinction between the trading and banking books, the qualitative

characteristics possessed by fair value accounting, the resultant earnings volatility

and overall preference for fair value accounting. Independent samples t-test are then

conducted to statistically test for differences between user and preparer perceptions.

7.3.1 Distinction Between Trading and Banking

One of the main reasons put forward by the JWGBA against fair value accounting is

the fundamental difference between the trading and banking books that warrants the

215
use of different measurement methods. Both users and preparers are asked the same

question and the responses are summarised in Table 7.2.

Table 7.2: Comparison of Preparer and User Perceptions of the Trading

Versus Banking Distinction
Preparer User
(Mean = 3.20) (Mean = 3.20)

The banking book is so Strongly Disagree 7 (8%) 3 (5%)

fundamentally different Disagree 20 (24%) 15 (23%)
from the trading book to
Neutral 16 (20%) 20 (31%)
warrant the use of
different accounting Agree 29 (35%) 21 (32%)
measurement bases. Strongly Agree 11 (13%) 6 (9%)
Total 83 65
T = 0.108; Significance = 0.914
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

A p-value of 0.914 signals no significant difference between the opinions of the two

groups. The average user and preparer aggregate response are both 3.2, a neutral

opinion on the difference between the trading and the banking books. Although the

average preparer and user responses are neutral, the percentages indicate that

preparers are less likely to have a neutral response to this with 20% without a view as

compared to 31% of users. This distinction is one of the main reasons put forward

by the JWGBA against fair value accounting. On the whole, it seems that the

JWGBA view is not widely held by both preparers and users in the sample surveyed.

However, as noted in Section 7.2, there is a significant difference in the preparers’

and users’ level of experience. Therefore, a general linear model is fitted to test the

effect of experience. The three variables in this model are (1) experience (a

covariate), (2) preparer or user and the (3) interaction between the first two variables.

Results (not tabulated) show that the differences between users and preparers do not

depend on the level of experience (p-value = 0.147).

216
7.3.2 Qualitative Characteristics

Both the user and preparer surveys ask respondents questions on relevance and

comparability. The questions are highly aligned but worded slightly differently. The

questions on the reliability of non-traded and traded financial instruments are

identical. Therefore, preparer and user responses are compared in the following

subsections and independent samples t-test conducted to examine if user perceptions

are significantly different to preparer views concerning relevance, reliability and

comparability of fair value accounting for financial instruments.

7.3.2.1 Relevance of Fair Value Accounting

Responses indicate that on average, user and preparer perceptions of the relevance of

fair value accounting is somewhat similar with means of 3.80 and 3.50; providing

some agreement that fair value accounting results in relevant information.

Table 7.3: Comparison of Preparer and User Perceptions of the Relevance of

Fair Value Accounting
Preparer User
(Mean = 3.80) (Mean = 3.50)

Fair value accounting Strongly Disagree 2 (2%) 0 (0%)

improves the relevance of Disagree 8 (10%) 8 (12%)
the information in the
Neutral 15 (18%) 23 (36%)
accounts for users.
Agree 39 (47%) 26 (40%)
Strongly Agree 19 (23%) 8 (12%)
Total 83 65
T = 1.675; Significance = 0.096
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

Statistically, as shown in Table 7.3, a p-value of 0.096 signals a borderline moderate

difference between the opinions of the two groups. The percentages again indicate

preparers are likely to have a neutral opinion on the relevance of fair value

217
accounting information and more likely to find it relevant. Therefore, both users and

preparers somewhat agree that fair value accounting results in relevant information

for decision making purposes. However, contrary to expectations, preparers more

strongly agree with the relevance of fair value accounting. The results of the general

linear model (not tabulated) show that the differences between users and preparers do

not depend on the level of experience (p-value = 0.242).

7.3.2.2 Reliability of Fair Values of Financial Instruments

Another qualitative characteristic of useful information is reliability. Both users and

preparers are asked (with identical questions) to assess the reliability of the fair

values of traded and non-traded financial instruments. User and preparer responses

to the two statements on the reliability of fair values are detailed below.

7.3.2.2.1 Traded Financial Instruments

On the reliability of fair values for traded financial instruments, Table 7.4 shows both

groups, on average, perceive the fair values for traded financial instruments to be

quite reliable (mean of 4.0 and 4.3) in line with JWGBA and JWGSS assertions.

Table 7.4: Comparison of Preparer and User Perceptions of the Reliability of

Fair Values for Traded Financial Instruments
Preparer User
(Mean = 4.00) (Mean = 4.30)

Reliability of fair values Highly Unreliable 1 (1%) 0 (0%)

for traded financial Unreliable 5 (6%) 2 (3%)
instruments
Neutral 11 (13%) 7 (11%)
Reliable 39 (47%) 24 (37%)
Highly Reliable 27 (33%) 32 (49%)
Total 83 65
T = -2.023; Significance = 0.045
Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

218
Independent samples t-tests reveal that users have a statistically significant higher

perception of the reliability of the fair values of traded financial instruments as

compared to preparers (who are the ones preparing the fair value numbers). Perhaps

users’ are overconfident in the reliability of such numbers as they are not aware of

the inherent difficulties and complexities since they do not prepare these fair value

financial statements themselves. The results of the general linear model (not

tabulated) again show that the differences between users and preparers do not depend

on the level of experience (p-value = 0.453).

7.3.2.2.2 Non-Traded Financial Instruments

In relation to the reliability of fair values for non-traded financial instruments, Table

7.5 shows both groups, on average, to have somewhat similar views tending around

the neutral position. The range of preparer and user responses also show the same

varied response on the perception of the reliability of fair values for non-traded

financial instruments.

Table 7.5: Comparison of Preparer and User Perceptions of the Reliability of

Fair Values for Non-Traded Financial Instruments
Preparer User
(Mean = 3.00) (Mean = 2.70)

Reliability of fair values Highly Unreliable 6 (6%) 2 (3%)

for non-traded financial Unreliable 22 (27%) 27 (42%)
instruments
Neutral 29 (35%) 22 (34%)
Reliable 22 (27%) 14 (21%)
Highly Reliable 4 (5%) 0 (0%)
Total 83 65
T = 1.374; Significance = 0.172
Legend: Highly Unreliable= 1; Unreliable= 2; Neutral= 3; Reliable= 4; Highly Reliable= 5

219
Neither group is particularly confident in the reliability of non-traded financial

instruments. It does seem that users tend to perceive fair values for non-traded

financial instruments to be unreliable. This lack of consensus on the reliability of

fair values for non-traded financial instruments is in line with the JWGBA views.

The independent samples t-test reveals no statistically significant difference (p-value

= 0.463) between the two groups’ perception.

Additional general linear models fitted (results not tabulated) with preparer or user,

experience and the interaction between these two variables show the interaction

variable to be moderately significant (p-value = 0.084). According to the parameter

estimates, the experience slope is -0.051 and 0.153 for preparers and users

respectively. This means that for every 1-unit increase in experience, users’

perceptions of the reliability of fair values for non-traded financial instruments

increase by 0.153. Therefore, level of experience does not significantly influence the

perception of the reliability of fair values for such financial instruments.

7.3.2.3 Comparability of Fair Value Accounting

According to the IASB Framework (1989), fair value accounting should result in

financial statements that are comparable to be useful. Both users and preparers are

asked to assess the comparability of the financial statements resulting from fair value

accounting across entities39. Their responses are compared in Table 7.6.

39
Although preparers were asked specifically on the comparability of the Balance Sheets across
entities, most of the changes introduced by this standard affect the Balance Sheet more than the
Income Statement. Therefore, the questions asked to both parties are sufficiently similar for
comparative purposes.

220
Table 7.6: Comparison of Preparer and User Perceptions of the
Comparability of Fair Value Accounting Financial Statements
Across Entities
Preparer User
(Mean = 3.60) (Mean = 3.20)

Comparability of fair value Strongly Disagree 3 (4%) 0 (0%)

accounting financial Disagree 12 (14%) 8 (12%)
statements across entities
Neutral 17 (20%) 36 (55%)
Agree 37 (45%) 19 (30%)
Strongly Agree 14 (17%) 2 (3%)
Total 83 65
T = 2.217; Significance = 0.028
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

The independent samples t-test shows preparers to have a statistically higher

perception of the comparability of financial statements prepared using fair value

accounting than users. Furthermore, a majority of users are neutral on the subject.

Both groups do agree on the comparability of these financial statements. The general

linear model (results not tabulated) confirms that the differences between users and

preparers do not depend on the level of experience (p-value = 0.117).

7.3.3 Concerns Regarding Volatility

Another main argument of the JWGBA concerned possible problems caused by the

resulting earnings volatility from fair value accounting for all financial instruments.

As shown in Table7.7, there is no statistical difference between user and preparer

views.

221
Table 7.7: Comparison of Preparer and User Perceptions of the Earnings
Volatility Resulting from Fair Value Accounting
Preparer User
(Mean = 3.50) (Mean = 3.30)
Fair value accounting for Strongly Disagree 1 (1%) 4 (6%)
all financial instruments 15 12
Disagree (19%) (19%)
introduces volatility that
will be misunderstood by Neutral 20 (24%) 14 (22%)
users. Agree 40 (48%) 32 (49%)
Strongly Agree 7 (8%) 3 (4%)
Total 83 65
T = 1.050; Significance = 0.295;
Legend: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5.

Users and preparers do agree (although to a lesser extent) with the JWGBA concern.

Also independent samples t-test confirms no statistical difference between the

opinions of the two groups. The bulk of users and preparers with an opinion on the

matter are concerned with the resulting earnings volatility. In addition, the general

linear model (results not tabulated) confirms that the difference between user and

preparer views do not depend on the level of experience (p-value = 0.811).

7.3.4 Preference for fair value accounting

As discussed in Chapter 5, an overall aggregate measure for preparer preference for

fair value accounting was derived from responses to four separate questions. The

composite overall measure derived is arguably theoretically and conceptually highly

similar with the single question on preference for fair value accounting40 asked in the

user survey. That overall measure is compared with users’ responses to the single

question. This is shown in Table 7.8.

40
The term is defined in the survey to specifically include the components separately asked in the
preparer survey.

222
Table 7.8: Comparison of Preparer and User Preference for Fair Value
Accounting
Preparer User
(Mean = 3.10) (Mean = 3.50)
Do you support the use of Strongly Oppose 9 (11%) 1 (2%)
fair value accounting for 26 16
Oppose (31%) (25%)
all financial instruments?
Neutral 12 (14%) 12 (18%)
Support 23 (28%) 20 (31%)
Strongly Support 13 (16%) 16 (24%)
Total 83 65
T = -2.262; Significance = 0.025
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.

The independent samples t-test result in a p-value of 0.025 highlighting that user and

preparer preference for fair value accounting is statistically significantly different.

Users tend to prefer fair value accounting more than preparers. The distribution of

responses highlights a lack of consensus among users and preparers when it comes to

preference for fair value accounting for all financial instruments. A general linear

model is also fitted to test the effect of experience as a covariate with preparer or

user, experience and the interaction between these two variables. The results (not

tabulated) show the interaction variable to be significant (p-value = 0.046). This

indicates that the differences between users and preparers depend on the level of

experience. According to the parameter estimates, the experience slope is 0.029 and

0.334 for preparers and users respectively. Therefore, for each 1-unit increase in

experience, users’ preference for fair value accounting increases by 0.334 on average

while preparers’ preference increases by only 0.029. This indicates that users with

more experience have an increased preference for fair value accounting.

223
7.3.5 Comparison of Preparer and User Views: Summary

A summary of user versus preparer views is shown in Table 7.9.

Table 7.9: Summary of the Comparison of Preparer and User Views

Preparer User P-Value

Distinction between trading and banking Neutral Neutral 0.914
books (3.2) (3.2)
Relevance of fair value accounting Agree Agree 0.096*
(3.8) (3.5)
Reliability of fair values for traded Agree Agree 0.045**
financial instruments (4.0) (4.3)
Reliability of fair values for non-traded Neutral Disagree 0.172
financial instruments (3.0) (2.7)
Comparability of fair value accounting Agree Agree 0.028**
(3.6) (3.2)
Volatility may be misunderstood by users Agree Agree 0.295
(3.5) (3.3)
Overall support for fair value accounting Neutral Support 0.025**
(3.1) (3.5)
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The average user and preparer response to each of the main themes tend to be in the

same direction. Users and preparers have statistically different views about the

reliability of fair values for traded financial instruments, comparability of fair value

accounting and overall support for fair value accounting. Though so, the average

response for both groups tends to be on the same side.

It is interesting to find that users show statistically more support for the use of fair

value accounting relative to preparers’ neutral response (on average). If the objective

of financial statements as per the IASB Framework is to be achieved, perhaps user

views should be the strongest focus of accounting standard-setting.

224
7.4 SUMMARY

Comparisons of the preparer and user groups indicate that their perceptions are

generally not statistically significantly different. Although highly similar, user and

preparer views show support for some of the arguments forwarded by both the

JWGBA and JWGSS. The general consensus is that fair value accounting is useful

but can lack reliability and the resulting earnings volatility is considered problematic.

Overall, there is slight support for the proposed fair value accounting model but

concerns on the reliability of numbers and the resulting earnings volatility persists.

Chapter 8 provides a summary of the thesis, details the implications of all the

findings of this research and suggests possible future research ideas.

225
 CHAPTER EIGHT:
IMPLICATIONS & CONCLUSIONS

Everything that has a beginning has an end… – The Oracle,

The Matrix Revolutions, 2003

8.1 OVERVIEW

The first three chapters of this thesis covered the introduction to the research project,

relevant literature and theoretical framework, and the research process adopted in

this study. Chapter 4 detailed the qualitative data collection phase which sets the

scene and finalised the surveys used to collect quantitative data from the target

sample of preparer and users. Chapters 5 and 6 dealt with the descriptive and

inferential statistics related to the data collected from both preparer and user surveys.

The analyses in these two separate quantitative phases were compiled and contrasted

in Chapter 7. This chapter presents the implications and conclusions of this research

project to answer the key research questions stated in Chapter 1. These two research

questions are:

RQ 1: To what extent do preparers and users support the use of fair value
accounting for all financial instruments?

RQ 2: What factors help to explain preparer and user support or opposition for
the use of fair value accounting for all financial instruments?

8.2 SUMMARY OF FINDINGS

The summary of the main findings of both the qualitative and quantitative phases of

this thesis as detailed in Chapters 4 to 7 are discussed in the following sections.

226
8.2.1 Preference for Fair Value Accounting

The summary of the statistical findings in relation to the two research questions on

preparer and user support for fair value accounting and possible determinants of their

perception is shown in Table 8.1.

Table 8.1: Summary of Findings for Both Research Questions

Research (RQ1) (RQ2) Statistical Findings

Question Factors/Determinants Direction
Mean
Extent of preparer Neutral Country by type of financial institution NA
support for fair (3.1) No difference between the banking and
value accounting +
trading book
(see Section 5.11)
Reliability of fair values for non-traded +
financial instruments
Comparability of fair value accounting +
Extent of user Support Experience +
support for fair (3.5) No difference between the banking and
value accounting +
trading book
(see Section 6.11)
Reliability of fair values for non-traded
+
financial instruments.
Volatility may be misunderstood by users −
Legend: Scores based on a 5 point preparer and user survey; Scale: Strongly Disagree = 1;
Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5; Non-neutral opinions are
statistically different from the neutral position according to the one-sample t-test.

Table 8.1 reveals that views remained moderate (although the level of general

support had increased) since Fargher (2001). Despite considerable variation (see

Section 6.10) within users’ perceptions, overall, users show a higher level of support

for fair value accounting (3.5 on the 5 point scale). This lends weight to the

JWGSS’s views that fair value accounting is the preferred method for financial

instruments by users. There is also variation within preparers’ views (see Section

5.10) and the average neutral response somewhat contradicts the JWGBA’s argument

that fair value accounting is not preferred for all financial instruments. Results of the

statistical tests (general linear models) indicate that when users and preparers

227
perceive that fair value accounting information possesses the qualitative

characteristics listed on the IASB framework, they tend to favour fair value

accounting for all financial instruments. These findings also show that views have

changed in the ten years since the comprehensive AICPA (1994) study on user

information needs. However, this shift towards fair value accounting may be due to

other reasons such as different participants or perhaps simply a country difference.

On average, there is considerable support for fair value accounting for all financial

instruments but there are still concerns regarding earnings volatility, reliability of fair

values of non-traded instruments and the comparability of fair value accounting.

8.2.2 Alignment of Views

This thesis examines the diverse claims (about user information needs) between the

JWGSS and JWGBA when the JWGSS first proposed the use of fair value

accounting for all financial instruments. Given the different membership and

interests of these two groups, it is expected that preparers’ views will be more

aligned with JWGBA assertions and users’ perceptions closer to the JWGSS

position. In order to graphically represent the positions of the various constituents,

the average score as per the preparer and user survey results are indicated on a

spectrum of scores one to five. Then, the explicit (or implicit) stances taken by the

JWGSS and JWGBA respectively are estimated (as best possible) based on

assertions made in their respective official documents on this debate. In doing so,

where the global body is silent on a particular issue (for example, the JWGSS is not

explicit on whether fair value accounting results in comparable information), their

positioning is based on interpretation of their discussions. Figure 8.1 depicts this

interpretation and survey results.

228
Figure 8.1: Spectrum of JWGBA versus JWGSS Views: Compared to Thesis
Evidence from Users and Preparers

Key Themes
Distinction between trading and banking books
Fundamental Difference Neutral No Difference
1 5
JWGBA JWGSS
User & Preparer

Relevance of fair value accounting

Less relevant Neutral Relevant
1 5
JWGBA JWGSS
User Preparer
Reliability of fair values of traded financial instruments
Less reliable Neutral Reliable
1 5
Preparer User
JWGSS & JWGBA
Reliability of fair values of non-traded financial instruments
Less reliable Neutral Reliable
1 5
JWGBA JWGSS
User Preparer
Comparability of fair value accounting
Less comparable Neutral Comparable
1 5
JWGBA JWGSS
User Preparer
Earnings volatility may be problematic
More problematic Neutral Less
1 5
JWGBA JWGSS
UserPreparer
Overall support for fair value accounting
Less Neutral More
1 5
JWGBA JWGSS
Preparer User

Legend: Average scores based on preparer and user survey; JWGSS and JWGBA positions
are based on interpretations from official documents even when they are explicitly
silent on a particular issue.

229
Figure 8.1 shows a lack of alignment between user and JWGSS views as well as

preparer and JWGBA assertions in four out of six points of contention in this debate.

An interesting trend is the high level of agreement between average user and average

preparer views on many of the key issues. Users and preparers also tend not to agree

with the two global bodies but sit somewhere in between the two official views.

However, in respect of the relevance of fair value accounting and overall support for

fair value accountings, both users and preparers are closer to the position of the

JWGSS. Yet the conclusions these groups derive regarding the fair value accounting

model, are partly based on their key assumptions on the perceptions of preparers as

well as users. Thus, the arguments of the JWGBA and the JWGSS need to be

reassessed in the light of the empirical evidence presented in this thesis and Fargher

(2001).

8.2.2.1 Alignment of User Views

As mentioned in Chapter Six, there is considerable variation between users’

perceptions of the main issues of this debate. Table 8.2 summarises the various user

views as gleaned from the Australian interviews and phone surveys as well as mail

surveys to both countries as compared to those asserted by the JWGSS and JWGBA

on their behalf. Generally, user responses vary from neutrality to some support for

the JWGSS views on fair value accounting. However, users disagree with the

JWGSS in that they perceive that the fair values for non-traded financial instruments

are unreliable whereas the JWGSS is optimistic that reliability can be achieved.

Similarly, the assertions by the JWGBA that users do not perceive fair value

accounting to be relevant is disputed by actual user views. This JWGBA assumption

of user contentment with the current mixed measurement model is not consistent

with the views expressed by the sampled users.

230
Table 8.2: Comparison of User Interviews, User Surveys and JWGSS and JWGBA Assertions

Theme User User Phone User JWGSS Assertions JWGBA Assertions Consensus
Interviews Survey Survey
Distinction between No Not asked Neutral Value of financial instruments The two books fundamentally No difference
trading and banking distinction (3.2) does not depend on realisation different to warrant the use of
books process – No difference different methods
Relevance of fair value Agree Agree Agree Fair value accounting provides Users did not perceive fair value Relevant
accounting (3.5) more relevant information for information to contribute to decision
users making process
Reliability of fair values Highly Agree Agree Highly reliable Highly reliable Fair values of
for traded financial reliable (4.3) traded instruments
instruments reliable
Reliability of fair values Reliable Agree Disagree Reliability can be achieved Lacks reliability due to its Contrasting views
for non-traded financial (2.7) subjectivity
instruments
Comparability of fair Not asked Not asked Agree Comparable information with Mixed measurement is known and Comparable
value accounting (3.2) common accounting tested, which aids comparability information
measurement policy
Volatility may be Disagree Not asked Agree Silent Volatility in reported earnings will be Contrasting views
misunderstood by users (3.3) misunderstood
Overall support for fair Support Support Support Fully support fair value Half support; not relevant in the Support for full
value accounting (3.5) accounting and firmly believes banking book and lack of complaint fair value
in its relevance to all users means users prefer mixed accounting
measurement
Legend: Scores are based on a 5 point preparer and user survey; Scale: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5; Non
neutral opinions are statistically different from the neutral position according to the one-sample t-test.

231
Overall, a slight variation exists between the views of all the users but generally, they

tend to agree with the JWGSS on most aspects of fair value accounting. However,

they do agree with the JWGBA on the potential problems with volatility in earnings

as well as the lack of reliable fair values for non-traded financial instruments.

Hence, the JWGSS and the JWGBA are sometimes right in their views on user

information needs.

8.2.2.2 Alignment of Preparer Views

The JWGBA assertions sit on the extremes and the preparer survey respondents’

perceptions range from neutral to mild agreement with their various positions on the

four key themes examined. Table 8.3 summarises the various preparer views as

gleaned from the interviews (conducted with Australian preparers only) and mail

surveys as compared to those asserted by the JWGSS and JWGBA on their behalf.

There is also a noticeable variation among preparers’ perceptions and they seem to

support some of the assertions made by each global group. Preparer responses vary

from neutrality to some support for the JWGBA arguments against fair value

accounting. However, they disagree with the JWGBA who asserted that fair value

accounting is not relevant and not comparable. There is also evidence that the

reliability of fair values for non-traded financial instruments can be achieved

according to the sampled preparers.

232
Table 8.3: Comparison of Preparer Interviews, Preparer Surveys and JWGSS and JWGBA Assertions

Theme Preparer Interviews Preparer JWGSS Assertions JWGBA Assertions Consensus

Survey
Distinction between Mixed; both support Neutral Value of financial instruments does The two books fundamentally Neutral opinion
trading and banking and opposition (3.2) not depend on realisation process – different to warrant the use of overall
books No difference different methods
Relevance of fair value One agree, one Agree Fair value accounting provides more Users did not perceive fair value Relevant
accounting disagree (3.8) relevant information for users information to contribute to
decision making process
Reliability of fair values Highly reliable Agree Highly reliable Highly reliable Fair values of
for traded financial (4.0) traded instruments
instruments reliable
Reliability of fair values Not entirely reliable Neutral Reliability can be achieved Lacks reliability due to its Contrasting views
for non-traded financial (3.0) subjectivity
instruments
Comparability of fair Not asked Agree Comparable information with Mixed measurement is known and Comparable
value accounting (3.6) common accounting measurement tested, which aids comparability information
policy
Volatility may be Agree Agree Silent Volatility in reported earnings will Agree
misunderstood by users (3.5) be misunderstood
Overall support for fair Mixed, one support Neutral Fully support fair value accounting Half support; not relevant in the Support for full fair
value accounting and one opposed (3.1) and firmly believes in its relevance banking book and lack of value accounting
to all users complaint means users prefer
mixed measurement
Legend: Scores are based on a 5 point preparer and user survey; Scale: Strongly Disagree = 1; Disagree = 2; Neutral = 3; Agree = 4; Strongly Agree = 5; Non
neutral opinions are statistically different from the neutral position according to the one-sample t-test.

233
This lack of alignment between the preparers (see Table 8.3) and the JWGBA in this

debate suggests that standard setters around the globe need to be wary of “industry

voices”. The ‘official industry view’ is not wholly consistent with the preparers’

evidence gathered in this thesis. Standard setters thus may make incorrect

assumptions regarding industry views. It is acknowledged that the JWGBA did not

explicitly state that they represent all financial institutions but implied that their

views are that of the financial institutions in the respective countries of all the

banking associations involved in the JWGBA.

8.3 IMPLICATIONS OF FINDINGS

The key findings generated by this research include some support for fair value

accounting, and the lack of major differences between user and preparer respondents

but variation within each group. These have implications for the accounting for

financial instruments especially in light of the convergence towards a single set of

accounting standards. The implications include the effectiveness of the IASB

framework’s list of qualitative characteristics and the clear need for extensive

education and training are discussed in the following sections.

8.3.1 Accounting for Financial Instruments: Political Process?

Tosen (2003, p. 13) said “When IAS 39 was released, it became clear that the

standard setters had taken on a mammoth task, attempting to reconcile the views of

the proponents of respectively fair value and non-fair value accounting”. True to

these words, the accounting for financial instruments remains a highly contentious

and topical area (Pozzi, 2003; Bradbury, 2000). The fact that a total of four

international exposure drafts relating to various aspects of IAS 39 were released in

234
2004 is evidence of the continuing problems the IASB has with financial

instruments. These releases are: (1) ED Proposed Amendments to IAS 39 Financial

Instruments: Recognition and Measurement: The Fair Value Option; (2) ED

Financial Instruments: Financial Guarantee Contracts and Credit Insurance; (3) ED

Financial Instruments: Cash Flow Hedge Accounting of Forecast Intragroup

Transactions; and (4) ED Financial Instruments: Transition and Initial Recognition of

Financial Assets and Liabilities (IASB, 2004a). According to the European Central

Bank (2004, p. 79), the move to fair value accounting “can be truly qualified as a

paradigm shift since backward-looking accounting measures based on the concepts

of prudence and reliability give way to measures based on prevailing economic

values.” It argues that fair value accounting may have positive consequences in

terms of relevance but lacks reliability, comparability and transparency.

Figure 8.2 depicts the developments undergone by IAS 39 until now as well as

highlighting the various influences that played a part in each development.

235
Figure 8.2: Recent Developments of IAS 39 and the Influences that Played a
Part

IAS 39 - December 1998

ƒ Effective 1 January 2001
ƒ Mixed measurement
ƒ No option to fair value all financial instruments

⇐ 2000 - E66 Limited Revisions to IAS 39

⇐ JWGSS on Financial Instruments extending
from the IASC 1997 Discussion Paper

IAS 39 JWGSS Draft Standard - December 2000

ƒ Fair value accounting for all financial instruments

⇐ Opposition to fair value accounting for ALL

financial instruments especially with no hedge
accounting

IAS 39 - December 2003

ƒ Mixed measurement
ƒ Option to fair value all financial instruments

⇐ Huge reaction from major European financial

institution regulators who were concerned that
the option will be abused resulting in the
demise of various banks.

ED on Fair Value Option - April 2004

ƒ Restriction on the fair value option by imposing a
verifiability condition

The development of an accounting standard for financial instruments is highly

dynamic and much has happened since IAS 39 was first introduced. IAS 39 was

finalised in December 1998 as a result of Exposure Draft E62 Financial Instruments

(issued in June 1988) and issued as an interim standard because the then IASC could

not wait for the completion of the JWGSS’s proposals. This interim standard was a

compromise on the fair value approach recommended by the 1997 Discussion Paper

236
and the 2000 draft standard (Bradbury, 2003). Initially, IAS 39 was consistent with

the views of the JWGBA as it mandated a mixed measurement approach, but the

changes introduced in 2000 (as well as the revisions in 2003) allowed entities to

adopt fair value for virtually all financial instruments and thus provided some

support to the JWGSS position. However, the April 2004 ED on the fair value

option proposed to limit the widespread use of fair values and hence moved it back

towards the JWGBA position. The extensive lobbying from banks around the world

resulted in the proposed revised IAS 39 containing only a limited extension of fair

value accounting principles (IASB, 2004b). Thus, present rules state that banks will

be permitted to choose amortised cost for the banking book (Smith Barney, 2004).

All these dominating influences point towards standard setting being “fundamentally

as much a political process as an economic one” (Scott, 2003, p. 447) and affected by

organisational and institutional influences (Gaa, 1986; McSweeney, 1984).

As shown in Figure 8.2, an ED limiting the use of the fair value option based on

verifiability was issued in April 2004 for comments in response to the 2003 revised

IAS 39 standard. This ED was a result of the influence asserted by various European

constituents, especially the European banking regulators (arguably a very important

user of financial statements from financial institutions). The main concerns voiced

by this group centred on the lack of reliability of certain fair values which they felt

would result in artificial volatility that may ultimately lead to the demise of many

financial institutions. These views were also mirrored in this thesis as findings

indicated that when reliability (arguably similar to the concept of verifiability

introduced by IAS 39) can be achieved, there was higher support for fair value

accounting. The concerns of the European regulators needed to be dealt with

237
because the IASB wished to have the European Union fully adopt all IFRSs, hence

the very recent ED on the fair value option.

In addition, the 2003 revised IAS 39 categorised financial instruments into four

groups: (1) financial instruments at fair value through profit or loss; (2) available-for-

sale financial assets; (3) held-to-maturity investments; and (4) loans and receivables.

There are significant similarities between the first two categories and the financial

instruments classified into the banking book as per JWGBA (1999a) and between the

last two categories and the instruments typically found in the trading book.

According to IAS 39, all financial instruments are to be measured at fair value upon

recognition. Subsequent measurement depends on the category of financial asset as

outlined below:

ƒ financial assets at fair value through profit or loss - measured at fair value with

fair value changes in profit or loss,

ƒ available-for-sale financial assets - measured at fair value in the Balance Sheet

and changes are recognised directly in equity, through the statement of changes in

equity,

ƒ loans and receivables - measured at amortised cost, and

ƒ held-to-maturity investments - measured at amortised cost.

A similar mixed measurement model is also prescribed for financial liabilities with

two categories: financial liabilities at fair value through profit or loss and other

financial liabilities measured at amortised cost using the effective interest method.

Subsequently, paragraph 46 of IAS 39 (revised 2003) stipulates that financial assets

and liabilities (including derivatives) should be measured at fair value except loans

and receivables, held-to-maturity investments and non-derivative financial liabilities

238
which should be measured at amortised cost using the effective interest method while

investments in equity instruments with no reliable fair value measurement (and

derivatives indexed to such equity instruments) should be measured at cost.

Therefore, IAS 39 (revised 2003) continues to adopt a mixed measurement model

with an option for entities to choose the fair value option for all financial instruments

except for investments in unquoted equity investments.

However, as the IASB examines proposals to wind back the fair value option, the

lack of high alignment between preparer views and the JWGBA assertions (noted in

this thesis) should not be forgotten. Perhaps standard setters should not completely

assume that official representative groups’ positions are representative of all its

constituents. Conversely, these global representatives should put more effort into

obtaining the actual views of their constituents. The process of responding to EDs or

proposed standards is very costly and so time consuming that some constituents

cannot afford to participate (or feel that their concerns will not be addressed). This is

ironic since the due process is in place to supposedly ensure that users can have

direct representation in the formulation of standards as they are the direct

beneficiaries (Harding and Mckinnon, 1997). The ancient and ongoing calls for

greater user involvement in the standard-setting process (Rahman, 1991, Walker,

1990, Masel, 1983; Balmford, 1977) also raises questions of the effectiveness of the

due process and whether all voices are being heard (Baruch, 1988). The relatively

low response rate to the mail surveys in this thesis (as well as submissions to

exposure drafts) is also another indication that not all parties take time to participate

in the process.

In addition, user responses from the phone survey reveal that the main reason for the

lack of participation in the debate is a belief that their input will not have any

239
influence (and not make a difference) over the whole standard setting process (see

Section 4.3.2.2 and Appendix G). It seems that this constituent, who is the final

benefactor of financial statements, has largely given up in participating in the due

process. It is disturbing that users believe that their voices or votes on accounting

standards do not count. Presumably, this situation would be further exacerbated for

Australian users under the post 2004 IFRS regime. However, in the international

arena, 28% of the comment letters (33 in total) received in response to the IASB ED

on the fair value option (issued April 2004) came from users of financial statements

of financial institutions. They are mostly regulators of the financial institutions

industry (which support the restriction on the fair value option) and accounting

professional associations. Interestingly, the majority of the respondents prefer the

unrestricted status quo in the 2003 revised IAS 39 (IASB, 2004b).

Given the status of IAS 39 (December 2003) with the fair value option and the

overall (albeit mild) support for fair value accounting shown by respondent users and

preparers in this study, it seems that the standard setters have (possibly) got the

accounting for financial instruments standard about right. Ultimately, it seems that

there is no position that would please all preparers and users on this very

controversial issue. One may argue that the views of users, as the final beneficiary of

financial statements, are paramount and so long as they believe the accounting

standard will result in relevant, reliable and comparable information, then rule-

makers are achieving a useful equilibrium.

8.3.2 Classic Concerns on Volatility, Reliability and

Comparability

Findings of this thesis also show that support for fair value accounting is influenced

by: 1) perceptions of comparability across entities, 2) concerns that volatility may be

240
misunderstood, and 3) perceptions that fair values of non-traded financial

instruments are unreliable. These seems especially reflected in the influence various

European regulators had over the IASB decision to issue the fair value option ED

proposing to restrict the use of the fair value option in the 2003 revised IAS 39.

The inclusion of the fair value option (in the 2003 revised IAS 39) is a reflection of

the IASB’s belief that fair value accounting is most relevant and reliable for all

financial instruments (IASB, 2004b). However, the IASB ED on the fair value

option is a direct response to European concerns that this option might be used

inappropriately and might result in increased (and unnecessary) volatility in earnings

that could ultimately lead to the demise of entities. The BASLE Committee on

Banking Supervision (2004) also voiced concern that such an option may reduce the

comparability of financial statements.

This crucial concern on reliability resulted in the ED on Fair Value Option restricting

its use based on a verifiability condition as defined in paragraph 25 of the Basis for

Conclusions (BC) as “meaning that the variability in the range of reasonable fair

value estimates made in accordance with IAS 39 is low”. BC 25 goes on to state

that: “Accordingly, if this proposal is adopted, fewer items will qualify for the fair

value option that are measured at fair value if classified as held for trading or

available for sale in accordance with IAS 39 requirements”.

The Australian Prudential Regulatory Authority (APRA), in their comment letter in

response to the IASB fair value option exposure draft believes that the proposals for

amending the fair value measurement option in IAS 39 should be specific and

directive to enable consistent application. They also called for consideration on how

the concept of “verifiability” relates to the IASB Framework. In the Australian

241
context, SAC 3: Qualitative Characteristics of Financial Information specifically

relates “verifiable” to the concept of “reliability” but the IASB Framework does not

specifically define that term. Clearly, clarification of key concepts is needed.

Subsequently, the APRA (2004) believed there are practical difficulties in applying

this test and a more robust fair value measurement model may be needed. APRA

also called for specific directive guidance to ensure that comparability and

understandability is not compromised with the introduction of “verifiability”. In

addition, reliable measurement has an element of verifiability but not at the expense

of relevance (see SAC 3 Qualitative Characteristics of Financial Information,

paragraph 23). As APRA (2004, p. 3) puts it “…there needs to be a balance between

the qualitative characteristics of financial information to enable general purpose

financial reports to be useful for users”.

The concern on the resulting volatility in earnings has been around for some time.

Even back in 1991, a letter (dated 1 November 1991) from the Australian Bankers’

Association to the Executive Director of the Australian Accounting Research

Foundation in response to the release of Discussion Paper 14 says “How relevant will

the accounts be to analysts and other users in comparing the accounts with those of

non financial institutions if the result of market value accounting will be greater

volatility in reported profits?”. Gray (2003) also notes that the resulting earnings

volatility can have adverse effects on companies. Similarly, user respondents are

also concerned with the resulting volatility in earnings from the use of fair value

accounting and the perception that this is not an issue increases their support for fair

value accounting (see Section 6.9). Arguably, the concern on earnings volatility

could be the ultimate driver behind the push from the banks and banking regulators

for the use of verifiability proposed in the fair value ED.

242
The comparability of financial statements across entities is also a predictor of

preference for fair value accounting. The lack of comparability is highlighted to be a

problem of fair value accounting by the JWGBA, Chalmers and Godfrey (2000) and

Hernandez (2003). In addition, this notion of comparability is implicitly linked to

underlying reliability or lack thereof for fair values (see Hernandez, 2004). In

arguing their cases, the JWGSS (2000) and the JWGBA (1999a) relied on the

qualitative characteristics set out in the IASB Framework of relevance, reliability,

comparability and understandability. However, user and preparer responses indicate

only moderate support for the assertions made by either JWGSS or the JWGBA.

Results show minimal differences in the opinions between users and preparers but

there is still variation within each group. The general linear models indicate that

when users and preparers perceive that fair value accounting information possess the

qualitative characteristics listed on the IASB framework, they tend to support fair

value accounting for all financial instruments. This trend should be comforting for

standard setters to the extent that the qualitative characteristics listed in the IASB

framework seem to be representative of useful information and thus, making the

framework highly relevant. Walker and Jones (2003, p. 361) even suggests that a

way of selecting among measurement attributes is to “refer to the qualitative

characteristics of information” and the preferred measurement basis should be

chosen based on for example, relevance, reliability and interpretability.

The IASB framework stated four principal qualitative characteristics including

understandability, relevance, reliability and comparability as attributes that make

financial information useful. All the attributes listed in the IASB framework that

were covered in the survey were found to be highly statistically correlated with the

overall measure of useful information according to user responses (see Table 6.30).

243
This confirms the role of the IASB framework, according to sophisticated users in

Australia and Singapore. The framework does provide guidance on attributes that

result in useful information. Arguable, the role of the framework is even more

crucial if the IASB becomes the sole global accounting standard setter (Jones and

Wolnizer, 2003). Its presence, as a coherent and rigorous body of accounting theory

to guide the development of accounting, is as (if not more) crucial than mere

technical requirements (Loftus, 2003; Gerboth, 1987; Solomons 1986; Horngren,

1981)

Overall, as per the findings of this study, the distinction between the user and

preparer needs seems less crucial than ensuring that the resulting financial statements

possess the qualitative characteristics listed in the IASB framework. Thus, to obtain

acceptance of IAS 39, the key goals for the standards setters should be to ensure that

the proposed accounting model can achieve the qualitative characteristics listed on

the IASB framework. This is potentially much more important than appeasing

interest groups.

8.3.3 Complexity of Topic and Need for Education

Another important finding of this thesis is a concern about the lack of education and

training on this very complex topic. Some highly experienced users admit to a lack

of understanding due to the complexity of fair value accounting and financial

instruments in general (see Section 4.3). However, respondents to the user survey

agree that fair value accounting in financial statements is understandable by users

with reasonable knowledge of business activities and accounting (see Chapter 6).

The European regulators concern on the possible misuse of the fair value option is

another signal that fair value accounting is very complex for preparers. This

244
complexity could be reflected by the findings that more experienced users have a

higher preference for fair value accounting. In addition, Betts and Wines (2004) also

emphasise principles and difficulties involved in operationalising the fair-value

concept, especially in the absence of an active market. Ryan, et. al. (2002) also

highlight substantial measurement issues in fair valuing financial assets on the

Balance Sheet and recognising fair value changes in the Income Statement. The

complexity of IAS 39 is also found to be one of the most significant impediments to

convergence in a survey conducted in 2002 by the (then) Big-6 accounting firms

(Larson and Street, 2004). The existence of respondents who voiced lack of

understandability due to the complexity of fair value accounting backs up these

literature assertions.

This further implies that topics that are too technical could potentially impede the

understandability of accounting standards and resulting financial statements. The

IASB Framework implicitly assumes users to have at least a reasonable (if not

sophisticated) knowledge of business and economic activities and accounting.

Evidence from this thesis casts some doubt about this assumption. Furthermore, this

brings into question the potential for optimal decision making especially if it impedes

the understandability to the average knowledgeable investor (also pointed out by

Eyes and Tabb, 1978). Such a lack of understanding has been highlighted since Lee

and Tweedie (1975). It might also impact on the effectiveness of the due process –

perhaps standards are just too complex. Findings in this study raise an urgent need

for simplifying IAS 39 and some other IFRSs in order to be able to achieve

understandability as per IASB framework. An important step in this area is the

IASB’s active project (available via the IASB website) to develop accounting

standards suitable (i.e. more simplified) for Small and Medium Enterprises. There is

245
also evidence of a “two standard” system in numerous European countries (Larson

and Street, 2004)

In addition, Section 7.3.4 shows that users’ preference for fair value accounting

increases more than preparers where there is an increase in experience. Another

obvious trend is the high support and call for education and training on this highly

complex topic which was echoed by various respondents in the interview and phone

survey phases (as per Sections 4.2 and 4.3). This evidence tentatively suggests that

increased training and education, inevitably increasing experience and awareness of

fair value accounting for financial instruments, will result in higher support for this

measurement attribute for financial instruments. Furthermore, this confirms the need

for the understandability characteristics listed in the IASB Framework (1989).

8.3.4 Other Implications

Another observation is that Singaporean preparer responses tend around the neutral

regardless of the type of financial institution while there is considerable variance in

Australian preparers’ views. Foreign banks in Australia have the lowest preference

for fair value accounting for all financial instruments whereas their counterparts in

Singapore showed the highest level of support. This may be explained by the

extensive Australian involvement in the international debate on the accounting for

financial instruments as compared to the mild more accepting reaction from

Singapore. Traditionally, accounting standards in Singapore tend to follow closely

with IFRSs (and the IASs) which may explain the lesser variation in views and

perhaps also the lesser opposition to the new proposed standard. Furthermore,

Singapore first adopted IAS 39 as Singaporean Accounting Standard 3341 for

41
This has since been renamed Financial Reporting Standard 39 (Chua and Chen, 2004).

246
implementation after 1 July 2001 but then postponed the implementation to 1

January 2004 and subsequently changed to 1 January 2005 (Chua, 2002; Chua and

Chen, 2004). This suggest that international accounting standards are seen to be

more acceptable and preferred in countries that do not actively participate in

accounting standard setting.

8.4 IMPORTANCE OF THIS RESEARCH

Kevin Stevenson, the Director of Technical Activities for the IASB, in his speech at

the 2004 annual conference of the Accounting and Finance Association of Australia

and New Zealand (AFAANZ) asserted that the accounting for financial instruments

is potentially the main driving force of a paradigm shift in the conceptual framework

away from historical cost towards fair value accounting. The overall support for fair

value accounting in this thesis is indicative that this move is likely to be preferred

(albeit with some trepidation) by users and preparers alike.

Although there have been various developments in the accounting for financial

instruments, the main premise and principles behind the initial IASB discussion

paper (1997) lives on in its successors; both from the 2000 draft standard to the 2003

revised IAS 39. This move towards fair value accounting continues but the four

exposure drafts released in 2004 alone prove that the topic is still difficult and

controversial.

This study concentrates on the actual views but not actions of constituents to this

debate and the resulting ongoing evolution of this highly complex and controversial

standard. Importantly, the views of smaller benefactors (such as credit unions and

247
building societies) are solicited to improve the level of insights into the issues

surrounding this debate.

Overall, the results signal support for fair value accounting and suggests that the

current IAS 39 caters for the concerns of all major parties and is essentially a

compromise between the standard setters’ views as well as major users’ (namely

European regulators) concerns. This middle ground is reflected by the numerous

neutral average responses from users and preparers to the mail survey. The extensive

and global-encompassing developments and publications on IAS 39 itself further

support the importance of research into this area.

In addition, currently highly topical issues (evidenced by coverage in prominent

domestic and international accounting conferences such as the European Accounting

Associations Annual Congress) tend to focus on the concept of harmonisation (and

convergence and internationalisation), corporate governance, earnings management

and auditing. All these topics signal one common denominator, the focus on user

needs and the related provision of accurate and useful accounting information.

More recently, the International Association of Accounting Education and Research

(IAAER), as part of their Reporting Financial Performance Research Program, called

for proposals for up to five research projects with funding of USD$20,000 into the

reporting for financial performance. Some of the main topics include how the

performance statement could display the effects of mixed attribute accounting and

whether realised and unrealised items should be distinguished and displayed

separately from each other. These two topics contain similar issues to the underlying

concerns about the fair value accounting model and serve as a further reminder on

the need for this research. Some even believe that the crucial issue in accounting for

248
financial instruments is performance measurement (Macve, 1999; Mumford, 2000;

Horton and Macve, 2000).

This move to fair value accounting is not just restricted to financial instruments; it is

part of a bigger tapestry of a slow but steady movement towards fair value

measurement in financial accounting and reporting. Standard setters are showing a

clear predisposition towards fair value measurement already (Betts and Wines, 2004;

Poon, 2004; Nissim, 2003) as evidenced by the accounting for agriculture and the

asset impairment standard. As various prominent researchers and players have

pointed out, accounting for financial instruments is indeed pushing a fundamental

change in measurement as we know it.

8.5 FUTURE SUGGESTIONS AND RESEARCH IDEAS

With current movements towards international convergence or harmonisation, further

research into the usefulness of accounting standards is called for. Considering that

the objective of financial statements is to provide information useful for users’

decision making purposes (as per IASB Framework and many others), current and

pending standards should be examined to ensure that they truly assist in the

generation of useful financial statements. The findings of this study signal the need

for more research on the actual information needs of users and the actual views of

preparers (and not just assume it is the same as the official industry representative’s

assertions).

As noted in Chapter Two, there are various interested parties (such as countries that

are active and prominent in the international accounting standard setting arena not

limited to the United States of America and countries in the European Union) with

249
active research and discussions on this proposed fair value accounting model for all

financial instruments (including the ECB and the JWGBA). Therefore, research on

preparer and user views from countries such as Europe, the US and UK, to name a

few, will add to the understanding on the preference for fair value accounting for all

financial instruments. Also, with the massive convergence towards IFRS by

Australia and the European Union as of 1 January 2005, it would be beneficial to

again talk to users (perhaps via phone surveys or focus groups) for their views now

with all the recent developments on the accounting for financial instruments. Such

ongoing dialogue is essential if harmony (consistency across countries) is to be

obtained.

A longitudinal study may also capture the changes (or lack thereof) in preparer and

user views over a longer period of time considering the changes undergone by IAS

39 from mixed attribute to fair value accounting to mixed attribute with an option to

use fair value accounting for all financial instruments. This is especially crucial in

light of international movements towards a single set of accounting standards.

An alternative approach to investigating the issues associated with fair value

accounting for financial instruments could be to utilise experimental research

methods to better test actual user understandability of the implications of fair value

accounting for financial instruments. Responses to a survey that follow this

experiment could capture the actual perception of the understandability of fair value

accounting in a controlled environment (Sarantakos, 1993).

In addition, there is a need for research into the information needs of different user

groups. For example auditors as compared to analysts and individual investors

versus institutional investors, and whether there are any significant differences

250
between the user groups. Also relevant is the examination into whether different

groups of users actually can understand fair value accounting. All this will have

important implications for accounting standard setting that assumes a reasonable

understanding of accounting from users of financial statements (see Harding and

McKinnon, 1997 for discussion on user involvement in the standard setting process).

8.6 FAIR VALUE ACCOUNTING - FINAL WORDS

This thesis makes an original contribution to the accounting literature on fair value

accounting for financial instruments in the financial institutions industry, exploring

the perceptions of preparers as well as users. In response to the two global

contrasting views of the JWGBA and the JWGSS, this thesis provides actual

empirical data on both preparer and user views on fair value accounting for all

financial instruments. In addition, this research contributes towards bridging the gap

in accounting literature in relation to users’ actual information needs and finds

‘representative’ preparer or trade groups may not always fully reflect all their

constituents’ views.

The examination and understanding of preparer and user views indicate general level

of support for fair value accounting for all financial instruments. Potential areas of

controversy and difficulties with the proposed fair value accounting model seem to

be the contentious reliability of the fair values for non-traded financial instruments as

well as problems with earnings volatility as a result of its use. The results highlight

that these issues need to be adequately dealt with in the international arena before

there will be full acceptance of fair value accounting for all financial instruments.

251
The findings result in a better understanding of the accounting for financial

instruments in financial institutions industry, a highly complex industry. Given that

this industry is usually one of the most influential and significant in any country, a

better understanding of the contentious issues surrounding fair value accounting is

likely to result in a more appropriate accounting standard for financial instruments

for both preparers and users alike. This may well enhance the stewardship and

accountability function of accounting.

This study provides insights into the accounting preference for two key financial

markets in the Asia-Pacific region, Australia and Singapore. Australia has now

adopted the IFRSs and Singapore has issued Financial Reporting Standards (FRSs)

that are almost identical to the current set of IFRSs42 (Delloite, 2003). These moves

signal a big win for international harmonisation/convergence advocates. The

complementary use of both qualitative and quantitative methods strengthened the

research methods adopted in this study. Interviews and phone surveys conducted in

the qualitative phase provided rich and deep insights into the world of fair value

accounting for financial instruments and resulted in a highly defined mail survey.

This research offers users, preparers and the accounting regulatory bodies a wider

picture of the acceptance and decision-usefulness of fair value accounting for all

financial instruments within the financial institutions industry. The lack of variation

the views of the two groups shows that the ensuring that fair value accounting results

in financial statements that possess the qualitative characteristics set out in the IASB

framework is far more important than trying to resolve different opinions. Thus, a

significant challenge this thesis has found, is that the need for accounting standards

42
Singapore accounting standards has an additional FRS 25 Accounting for Investments but
excluded IAS 40 Investment Property.

252
that result in information that has the qualitative characteristics set out in the IASB

framework for there to be less resistance from users and preparers alike. This move

towards fair value accounting is imminent; the need for reliability is imperative and

earnings volatility concerns still play a huge part in this ongoing controversy.

253
 REFERENCES
Ahmed, A. S., & Takeda, C. (1995). Stock market valuations of gains and losses on
commercial banks' investments securities: An empirical analysis. Journal of
Accounting and Economics, 20(2), 207-225.

AICPA. (1994). Improving business reporting - A customer focus: Meeting the

information needs of investors and creditors. Retrieved 6th January, 2000, from
http://accounting.rutgers.edu/raw/aicpa/.

AIMR. (2001). Comment Letter JWG Draft Standard and Basis for Conclusions.
Retrieved 27th November, 2004, from http://www.iasb.org/docs/fijwg/in122.pdf.

Aivazian, V. A., & Callen, J. L. (1983). Core theory and uniformity in accounting:
Rationalizing the accounting rulemaker. Journal of Accounting and Public Policy,
2(4), 225-237.

Alfredson, K. (2000). AASB Invites Comment on Proposals on the Measurement and

Recognition of Financial Instruments, AASB Media Release. Retrieved 18th October,
2003, from http://www.aasb.com.au/whatsnew/media_docs/aasb_mr_00.pdf.

Alfredson, K. (2003). Pathway to 2005 IASB Standards. Australian Accounting

Review, 13(1), 3-7.

American Accounting Association (AAA). (1966). A Statement of Basic Accounting

Theory. Illinois: AAA Committee to Prepare a Statement of Basic Accounting
Theory.

American Accounting Association Subcommittee on Establishing Materiality

Criteria of the Committee on Financial Accounting Standards. (1976). Response to
the Financial Accounting Standards Board's discussion memorandum entitled 'An
analysis of issues related to criteria for determining materiality'. The Accounting
Review, 51(Committee Reports: Supplement to Vol. XLXI 1976), 243-256.

Andrews, F. M. (1984). Construct validity and error components of survey method.

Public Opinion Quarterly, 48(2), 409-442.

Aram, J. D., Salipante, P. F. Jr., & Knauf, J. W. (1987). Human resource indicators
for hospital managers. Health Care Management Review, 12(2), 15-22.

Atkinson, P., & Silverman, D. (1997). Kundera's immortality: The interview society
and the invention of self. Qualitative Inquiry, 3(3), 304-325.

254
Australian Accounting Standards Board (AASB). (2001). SAC 2: Objectives of
general purpose financial reporting [Electronic version]. Retrieved 21st November,
2001, from http://www.aasb.com.au/public_docs/concept_statements/SAC2_8-90_
2001V.pdf.

Australian Financial Markets Association (AFMA). (2000). Accounting for Financial

Instruments - Survey 2000 Results. Sydney, Australia.

Australian Prudential Regulatory Authority (APRA). (2004). Comment letter to the

Exposure Draft of proposed amendments to IAS 39 Financial Instruments
Recognition and Measurement: The Fair Value Option. Retrieved 22nd November,
2004, from http://www.iasb.org/docs/ed-ias39fv/IAS39_FairValue_CL16.pdf.

Balmford, J. D. (1977). Australian Standards - Development and prospects since

1972. Australian Accountant, 47, 546-555.

Bank for International Settlements. (2002). International Banking and Financial

Market Developments. Basel, Switzerland: Monetary and Economic Department.

Bank for International Settlements. (2004, June 2004). The Basel Committee on
Banking Supervision. Retrieved 27th November, 2004, from http://www.bis.org/bcbs/
aboutbcbs.htm.

Barth, M. E. (1994). Fair value accounting: Evidence from investment securities and
the market valuation of banks. The Accounting Review, 69(1), 1-25.

Barth, M. E., Beaver, W. H., & Landsman, W. R. (1996). Value-relevance of banks'

fair value disclosures under SFAS No. 107. The Accounting Review, 71(4), 513-537.

Barth, M. E., Beaver, W. H., & Stinson, C. H. (1991). Supplemental data and the
structure of thrift share prices. The Accounting Review, 66(1), 56-66.

Barth, M. E., Beaver, W. H., & Wolfson, M. A. (1990). Components of earnings and
the structure of bank share prices. Financial Analysts Journal, 46(3), 53-60.

Barth, M. E., & Landsman, W. R. (1995). Fundamental issues related to using fair
value accounting for financial reporting. Accounting Horizons, 9(4), 97-107.

Barth, M. E., Landsman, W. R., & Wahlen, J. M. (1995). Fair value accounting:
Effects on banks' earning volatility, regulatory capital and value of contractual cash
flows. Journal of Banking and Finance, 19(3-4), 577-605.

Baruch, L. (1988). Toward a theory of equitable and efficient accounting policy. The
Accounting Review, 63(1), 1-22.

255
Baruch, Y. (1999). Response rate in academic studies: A comparative analysis.
Human Relations, 42(4), 421-438.

Basel Committee on Banking Supervision. (2004). Comments on the IAS 39 fair

value option proposal. Retrieved 22nd November, 2004, from http://www.iasb.org/
docs/ed-ias39fv/IAS39_FairValue_CL114.pdf.

Beaver, W. H., Eger, C., Ryan, S., & Wolfson, M. A. (1989). Financial reporting,
supplemental disclosures, and bank share prices. Journal of Accounting Research,
27(2), 157-178.

Benjamin, J., & Stanga, K. (1977). Differences in disclosure needs of major users of
financial statements. Accounting and Business Research, 8, 187-192.

Benston, G. J. (1974). Evaluation of the Securities Exchange Act of 1934. Financial

Executive, May, 28-42.

Bernard, V. L., Merton, R. C., & Palepu, K. G. (1995). Mark-to-market accounting

for banks and thrifts: Lesson from the Danish Experience. Journal of Accounting
Research, 33(1), 1-31.

Bernstein, D. A., Roy, E. J., Scrull, T. K., & Wickens, C. D. (1988). Psychology.
Boston, MA: Houghton-Mifflin.

Betts, J., & Wines, G. (2004). Application of the fair-value concept: Evidence from
Australian legal decisions. Australian Accounting Review, 14(1), 63-72.

Bianco, A. (1986). Playing with fire. Business Week, 16 September, 78-90.

Bonoma, T. V. (1985). Case research in marketing: Opportunities, problems, and a

process. Journal of Marketing Research, 22(2), 199-208.

Bradbury, M. E. (2000). Issues in the drive to measure liabilities at fair value.

Australian Accounting Review, 10(2), 19-25.

Bradbury, M. E. (2001). Survey concerning the reliability of fair values for financial
instruments. Unpublished Manuscript, Auckland, New Zealand: Financial
Instruments Joint Working Group.

Bradbury, M. E. (2002). Financial instruments update: Responses to the draft

standard on financial instruments. Chartered Accountants Journal, 81(3), 65-66.

Bradbury, M. E. (2003). Implications for the conceptual framework arising from

accounting for financial instruments. Abacus, 39(3), 388-397.

256
Burton, J. C. (1982). Intermediate accounting from a user perspective. In D. L.
Jensen (Ed.), The Impact of Rule-making on Intermediate Financial Accounting
Textbooks. Columbus, GA: College of Administrative Sciences.

Buzby, S. L. (1975). The nature of adequate disclosure. Journal of Accountancy,

April, 38-47.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and Quasi-experimental

Designs for Research. Chicago, IL: Rand McNally.

Cavana, R. Y., Delahaye, B. L., & Sekaran, U. (2001). Applied Business Research:
Qualitative and Quantitative Methods. Milton, Queensland: John Wiley and Sons
Australia.

Chalmers, K. G., & Godfrey, J. M. (2000). Practice versus prescription in the

disclosure and recognition of derivatives. Australian Accounting Review, 10(2), 40-
50.

Chang, L., & Most, K. (1985). The Perceived Usefulness of Financial Statements for
Investors' Decisions. Miami, FL: Florida International University Press.

Chapman, R. G. (1988). False Economies in Survey Research. Applied Marketing

Research, 28(1), 16-20.

Chastain, C. E. (1974). Accounting objectives and user needs: A behavioral view

part 1. The National Public Accountant, 19(6), 26.

Chiu, I., & Brennan, M. (1990). The effectiveness of some techniques for improving
mail survey response rates: A meta-analysis. Marketing Bulletin, 1, 13-18.

Chua, K. C. (2002). A need for preparation. Asia Risk, September, 35-37.

Chua, K. C., & Chen, V. H. (2004). IAS 39: The final countdown.
PriceWaterhouseCoopers: Banking Industry Update, April, 9-11.

Chua, W. F. (1986). Radical developments in accounting thought. The Accounting

Review, 61(4), 601-632.

Cohen, J., & Cohen, P. (1983). Applied Multiple Regression/Correlation Analysis for
the Behavioral Sciences. (Second Edition). Hillsdale, NJ: Erlbaum.

Collet, P., Godfrey, J., & Hrasky, S. (2001). International harmonization: Cautions
from the Australian experience. Accounting Horizons, 15(2), 171-182

Couch, C. (1987). Researching Social Processes in the Laboratory. Greenwich: JAI

Press.

257
Crawford, L. E., Wilson, A. C., & Bryan, B. J. (1997). Using and accounting for
derivatives: An international concern. Journal of International Accounting, Auditing
and Taxation, 6(1), 111-121.

Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests.

Psychological Bulletin, 52(2), 281-302.

Crotty, M. (1998). The Foundations of Social Research, Meaning and Perspective in

the Research Process. St. Leonards, NSW: Allen & Unwin.

Danos, P., Holt, D. L., & Imhoff, E. A. (1989). The use of accounting information in
bank lending decisions. Accounting, Organizations and Society, 14(3), 235-246.

Daugaard, D., & Valentine, T. (1995). Financial Risk Management: A Practical

Approach to Derivatives. Pymble, NSW: Harper Educational (Australia).

Davidson, K. (2002). An inquiry? Yes, into the Treasury. The Age, 18th November,
2002, Retrieved 10th October, 2003 from http://www.theage.com.au/articles/2002/
2011/2017/1037490050405.html.

Davis, D., & Cosenza, R. M. (1985). Business Research for Decision Making.
Boston, MA: PWS-Kent Pub. Co.

De Vaus, D. (1995). Surveys in Social Research. North Sydney, NSW: Allen and
Unwin.

Deloitte Touche Tohmatsu. (2003). The New Financial Reporting Framework in

Singapore. Singapore.

Deloitte Touche Tohmatsu. (2004). International Accounting Standards IAS 39:

Financial instruments: Recognition and measurement. Retrieved 24th September,
2004, from http://www.iasplus.com/standard/ias39.htm.

Dillman, D. A. (1978). Mail and Telephone Surveys: The Total Design Method. New
York, NY: Wiley-Interscience.

Dillman, D. A. (2000). Mail and Internet Surveys: The Tailored Design Method
(Second Edition). New York, NY: John Wiley & Sons, Inc.

Eccher, E. A., Ramesh, K., & Thiagarajan, S. R. (1996). Fair value disclosures by
bank holding companies. Journal of Accounting and Economics, 22(1-3), 79-117.

Edmondson, A. C. (1996). Group and Organizational Influences on Team Learning.

Unpublished doctoral dissertation, Harvard University, Boston, MA.

258
Epstein, M. J., & Pava, M. L. (1993). How useful are corporate annual reports?
Business Credit, 95(4), 18-19.

European Central Bank. (2004a). Comment Letter to Exposure Draft of proposed

amendments to IAS 39 Financial Instruments Recognition and Measurement: The
Fair Value Option, Retrieved 20th November, 2004, from http://www.iasb.org/docs/
ed-ias39fv/IAS39_FairValue_CL116.pdf.

European Central Bank. (2004b). The impact of fair value accounting on the
European banking sector - A financial stability perspective. ECB Monthly Bulletin,
February, 69-81.

European Financial Reporting Advisory Group (EFRAG). (2004). Comment letter to

the Exposure Draft of proposed amendments to IAS 39 Financial Instruments
Recognition and Measurement: The Fair Value Option. Retrieved 22nd November,
2004, from http://www.iasb.org/docs/ed-ias39fv/IAS39_FairValue_CL90.pdf.

Eyes, A. D., & Tabb, J. (1978). Bankers’ use of financial statements. The
Accountants Journal, April, 81-85.

Fargher, N. (2001). Management perceptions of fair-value accounting for all

financial instruments. Australian Accounting Review, 11(2), 62-72.

Feyerabend, P. (1987). Farewell to Reason, London, UK: Verso.

Filion, F. L. (1975). Estimating bias due to non-response in a mail survey. Public

Opinion Quarterly, 39(4), 482-492.

Financial Accounting Standards Board. (1978). Objectives of financial reporting by

business enterprises. Stamford, CT: FASB.

Financial Reporting Council (2002). Bulletin of the Financial Reporting Council

2002/4, 3 July 2002, Retrieved July 14, 2003, from Financial Reporting Council
website: http://www.frc.gov.au/content/bulletins/bull_2002_4.asp.

Fontana, A., & Frey, J. H. (2000). The interview: From structured questions to
negotiated text. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of Qualitative
Research (Second Edition, pp. 645-672). Thousand Oaks, CA: Sage Publications,
Inc.

Francis, J., & Schipper, K. (1999). Have financial statements lost their relevance?
Journal of Accounting Research, 37(2), 319-352.

Gaa, J. C. (1986). User primacy in financial reporting rulemaking: A social contract

approach. The Accounting Review, 61(3), 435-454.

259
Gay, L. R., & Diehl, P. L. (1992). Research Methods for Business and Management.
New York, NY: Macmillan Publishing.

Gerboth, D. (1987). The conceptual framework: Not definitions, but professional

values. Accounting Horizons, 1(3), 1-8.

Gray, R. P. (2003). Research note: Revisiting fair value accounting--measuring

commercial banks' liabilities. Abacus, 39(2), 250-261.

Griffis, S. E., Goldsby, T. J., & Cooper, M. (2003). Web-based and mail surveys: A
comparison or response, data, and cost. Journal of Business Logistics, 24(2), 237-
258.

Guba, E. G., & Lincoln, Y. S. (1994). Competing Paradigms in Qualitative Research.

In N. K. A. L. Denzin, Y. S. (eds) (Ed.), Handbook of Qualitative Research (pp. 105-
117.). Thousand Oaks, CA: Sage Publications.

Guerrera, F., Parker, A., & Pretzlik, C. (2003). As criticism of derivatives grows,
regulators push for tougher accounting rules: Financial Instruments: In hearings
starting today in Brussels, companies will protest that a new standard will limit their
ability to manage risk. Financial Times, London, UK, March 10, p. 19.

Guildford, J. P. (1965). Fundamental Statistics in Psychology and Education (Fourth

Edition). New York, NY: McGraw-Hill

Hague, I. (2002). Fair value for financial instruments - Where to next? Balance
Sheet, 10(1), 8-11.

Hair, J. F., Andersen, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate
Data Analysis. Englewood Cliffs, NJ: Prentice-Hall International.

Hancock, P. (1994). Accounting for financial instruments: An overview. Australian

Accounting Review, 4(2), 3-12.

Hancock, P. (1999). Financial Instruments. In K. Leo, C. Lambert & J. Sweeting

(Eds.), Financial Accounting Issues (2nd Edition). Brisbane, QLD: John Wiley and
Sons.

Harding, N., & McKinnon, J. (1997). User involvement in the standard-setting

process: A research note on the congruence of accountant and user perceptions of
decision usefulness. Accounting, Organizations and Society, 22(1), 55-67.

Haswell, S., & McKinnon, J. (2003). IASB standards for Australia by 2005: Catapult
or Trojan horse? Australian Accounting Review, 13(1), 8-16.

260
Hayne, S. C., & Pollard, C. E. (2000). A comparative analysis of critical issues
facing Canadian information systems personnel: A national and global perspective.
Information & Management, 38(2), 73-86.

Hernandez, F. G. (2003). Derivatives and the FASB: Visibility and transparency?

Critical Perspectives on Accounting, 14(8), 777-789.

Hernandez, F. G. (2004). Another step towards full fair value accounting for
financial instruments. Accounting Forum, 28(2), 167-179.

Hill, F. M. (1993). Research methodology and the management disciplines: The need
for heterogeneity. Irish Business and Administrative Research, 14(2), 46-51.

Hirschman, E. C. (1985). Scientific Style and the Conduct of Consumer Research.

Journal of Consumer Research, 12(2), 225-239.

Hodgkinson, G. P., & Payne, R. L. (1998). Graduate selection in three European

countries. Journal of Occupational and Organizational Psychology, 71(4), 359-365.

Holthausen, R. W., & Watts, R. W. (2001). The relevance of the value-relevance

literature for financial accounting standard setting. Journal of Accounting and
Economics, 31, 3-75.

Horngren, C. T. (1981). Uses and limitations of a conceptual framework. Journal of

Accountancy, 151(4), 86-95.

Horton, J., & Macve, R. (2000). 'Fair value' for financial instruments: How erasing
theory is leading to unworkable global accounting standards for performance
reporting. Australian Accounting Review, 10(2), 26-39.

Howieson, B., & Langfield-Smith, I. (2003). The FRC and accounting standard-
setting: Should I still call Australia home? Australian Accounting Review, 13(1), 17-
26.

Hu, H. T. C. (1993). Misunderstood derivatives: The cause of informational failure

and the promise of regulatory incrementalism. Yale Law Journal, 102(6), 1457-1513.

Husserl, E. (1931). Ideas: General Introduction to Pure Phenomenology. London,

UK: George Allen and Unwin.

International Accounting Standards Board. (1989). Framework for the Preparation

and Presentation of Financial Statements. London, UK: International Accounting
Standards Board.

261
International Accounting Standards Board. (2002). Financial instruments - Analysis
of responses to the December 2000 JWG Draft Standard. Information for Observers.
London, UK: International Accounting Standards Board.

International Accounting Standards Board. (2003a). ED Fair value hedge accounting

for a portfolio hedge of interest rate risk (2003). Retrieved 27th November, 2004,
from http://www.iasb.org/current/comment_letters.asp?showPageContent=no&xml=
16_30_79_19122003.htm.

International Accounting Standards Board. (2003b). International Accounting

Standards 2003. London, UK: International Accounting Standards Board.

International Accounting Standards Board. (2004a). Active Projects: Projects

Relating to Financial Instruments. Retrieved 22nd May, 2004, from
http://www.iasb.org/current/active_projects.asp?showPageContent=no&xml=16_99_
116_12102004.htm.

International Accounting Standards Board. (2004b). Amendments to IAS 39: The fair
value option. Unpublished manuscript, London, UK.

International Accounting Standards Board. (2004c). ED proposed amendments to

IAS 39 Financial Instruments: Recognition and Measurement: The Fair Value
Option. Retrieved 22nd November, 2004, from http://www.iasb.org/docs/projects/fi-
fvo-ps.pdf.

International Accounting Standards Committee. (1997). Accounting for financial

assets and financial liabilities. A Discussion Paper Issued for Comment by the
Steering Committee on Financial Instruments. London, UK: International
Accounting Standards Committee.

International Monetary Fund (IMF). (2004). Singapore: Selected Issues (IMF

Country Report No. 04/103) (Country Report). Washington, D. C.: International
Monetary Fund.

Joint Working Group of Banking Associations. (1999a). Accounting for Financial

Instruments for Banks.

Joint Working Group of Banking Associations. (1999b). Financial Instruments:

Issues Relating to Banks - Comments on the JWGSS Paper.

Joint Working Group of Standard Setters. (1999). Financial Instruments: Issues

Relating to Banks.

Joint Working Group of Standard Setters. (2000). Accounting for financial

instruments and similar items: An invitation to comment on the JWG draft standard,
Australian Accounting Standards Board.

262
Jonas, G. J., & Young, S. J. (1998). Bridging the gap: Who can bring user focus to
business reporting. Accounting Horizons, 12(2), 154-159.

Jones, M. (1995). Whatever happened to the corporate report? Management

Accounting, 76, 52-53.

Jones, S., & Wolnizer, P. W. (2003). Harmonization and the conceptual framework:
An international perspective. Abacus, 39(3), 375 - 387.

Judge, T. A., Thoresen, C. J., Bono, J.E., & Patton, G. K. (2001). The job
satisfaction-job performance relationship: A qualitative and quantitative review.
Psychological Bulletin, 127(3), 376-407.

Kanuk, L., & Berenson, C. (1975). Mail surveys and response rates: A literature
review. Journal of Marketing Research, 12, 440-453.

Kawaller, I. G. (2004). What analysts need to know about accounting for derivatives.
Financial Analysts Journal, 60(2), 24-30.

Kay, R. S. (1985). Accounting development. Journal of Accounting, Auditing and

Finance, 9(4), 67-73.

Kemp, S. (2004). Bank hit by $180m scandal. The Age, 14 January: p. 6.

Kerlinger, F. N. (1986). Foundations of Behavioral Research (3rd edition). New

York, NY: Holt, Rinehart and Winston.

KPMG. (1993). Fair Value of Financial Instruments - Disclosure and Reaction: A

Study of Bank Holding Company Annual Reports for 1992, KPMG.

KPMG Peat Marwick. (1992). Estimating Fair Values for Financial Instruments:
Disclosure and Beyond, a Study Prepared for the Association of Reserve City
Bankers. Washington, D.C., USA.

Kuhn, T. (1977). The Essential Tension: Selected Studies in Scientific Tradition and
Change. Chicago, IL: University of Chicago Press.

Lambert, D. M., & Harrington, T. C. (1990). Measuring non-response bias in

customer service mail surveys. Journal of Business Logistics, 11(2), 5-25.

Larson, R. K., & Street, D. L. (2004). Convergence with IFRS in an expanding

Europe: Progress and obstacles identified by large accounting firms' survey. Journal
of International Accounting, Auditing and Taxation, 13(2), 89-119.

Lavrakas, P. J. (1987). Telephone Survey Methods: Sampling, Selection and

Supervision. Newbury Park, CA: Sage Publications Inc.

263
Lee, T. A. (Ed.). (1988). Making Corporate Reports Valuable: The Literature
Surveys. Edinburgh, UK: The Institute of Chartered Accountants of Scotland.

Lee, T. A., & Tweedie, D. P. (1975). Accounting information: An investigation of

private shareholder understanding. Accounting and Business Research, 6(1), 3-17.

Lee, T. A., & Tweedie, D. P. (1981). The Institutional Investor and Financial
Information. London, UK: ICAEW.

Littleton, A. C. (1953). Statement of Accounting Theory. Evanston, IL: American

Accounting Association.

Loftus, J. A. (2003). The CF and accounting standards: The persistence of

discrepancies. Abacus, 39(3), 298-309.

Macve, R. (1999). One step forward, two steps back. Accountancy, 123(1269), 78.

Maiden, M. (2004). NAB board may have to air its dirty laundry. The Age, Retrieved
22nd November, 2004, from http://www.theage.com.au/articles/2004/2003/2025/
1079939785844.html.

Malhotra, N. K. (1993). Marketing Research: An Applied Orientation. Englewood

Cliffs, NJ: Prentice Hall.

Marsh, C. (1982). The Survey Method. London, UK: George Allen and Unwin.

Masel, L. (1983). The future of accounting and auditing standards. Australian

Accountant, 53(8), 541-549.

Matolcsy, Z. P., & Petty, J. D. (2001). Internal reporting of derivatives: Some

Australian evidence. Australian Accounting Review, 11(1), 26-33.

Mautz, R. K. (1963). Accounting as a social science. The Accounting Review, 38(2),

317-325.

Maynard, M. (1994). Methods, Practices and Epistemology: The Debate about

feminism and research. In M. a. P. Maynard, J. (Eds), Researching Women's Lives
from a Feminist Perspective (pp. 10-26). London, UK: Taylor and Francis.

McMonnies, P. N. (Ed.). (1988). Making Corporate Reports Valuable. London, UK:

Institute of Chartered Accountants of Scotland & Kogan Page.

McSweeney, B. (1984). Raising the standard-setting standards. Accountancy,

95(1089), 18.

264
Monetary Authority of Singapore. (2003). Financial Sector Overview. Retrieved 22nd
November, 2004, from http://www.mas.gov.sg/masmcm/bin/pt1Overview.htm#.

Mumford, M. J. (2000). Lessons for accounting measurement from transaction cost

economics. Australian Accounting Review, 10(2), 51-57

Neuman, W. L. (2000). Social research methods: Qualitative and Quantitative

Approaches (Fourth Edition). Boston, MA: Allyn and Bacon.

Nissim, D. (2003). Reliability of banks' fair value disclosure for loans. Review of
Quantitative Finance and Accounting, 20(4), 355-384.

Oppenheim, A. N. (1966). Questionnaire Design and Attitude Measurement.

London, UK: Heinemann Educational Books Ltd.

Osteraker, M. (2001). To put your cards on the table - collection of data through
silent interviews. Management Decision, 39(7), 578-582.

Pacter, P. (1999). Side by side: IAS 39 compared with FASB standards. Retrieved
19th May, 2000, from http://www.iasc.org.uk/news/cen8_142.htm.

Pakrul, H. A. (1977). A decision maker's perspective on how the accounting system

can meet user needs. Cost and Management, 51(5), 12-19.

Parker, A., & Pretzlik, C. (2004). Take it or leave it: New rules offer harmony for
Europe’s accounting standards, but who will play by them? The European Union
wants consistent guidelines for financial reporting to be adopted from next year but
many banks and insurers say the measures could adversely affect their balance
sheets. Financial Times, London, UK, 31 March 2004, p. 17.

Pearl, D. K., & Farley, D. (1985). Testing for the potential of non-response in sample
surveys. Public Opinion Quarterly, 49(4), 553-560.

Pernice, R. (1996). Methodological issues in unemployment research: Quantitative

and/or qualitative approaches? Journal of Occupational and Organizational
Psychology, 69(4), 339-349.

Perry, C., Alizadeh, Y., & Riege, A. (1997). Qualitative Methods in

Entrepreneurship Research. Proceedings of the Small Enterprise Association of
Australia and New Zealand 1997 Annual Conference, Novotel Opal Cove Resort,
Coffs Harbour, 21-24 September: 547-567.

Poon, W. W. (2004). Using fair value accounting for financial instruments. American
Business Review, 22(1), 39-41.

265
Popper, K. R. (1972). Objective Knowledge: An Evolutionary Approach. Oxford,
UK: Clarendon Press.

Pozzi, C. (2003). Changes afoot. CA Charter, 74(11), 54-55.

Puxty, A. G., & Laughlin, R. C. (1983). A rational reconstruction of the decision-

usefulness criterion. Journal of Business Finance and Accounting, 10, 543-559.

Rahman, A. R. (1991). Due process and user participation in standard setting.

Australian Accountant, 61(5), 28-34.

Ratneshwar, S., & Srinivasan, D. W. (1989). Non-response in mail surveys: An

integrative review. Applied Marketing Research, 29(3), 37-46.

Ravlic, T. (2002). Time bomb. CA Charter, 73(10), 58-60.

Riahi-Belkaoui, A. (1996). Accounting, A Multiparadigmatic Science. Westport, CT:

Quorum Books.

Riahi-Belkaoui, A. (1997). Research Perspectives in Accounting. Westport, CT:

Quorum Books.

Rice, C. D. (1973). Businessman's View of the Purpose of Financial Reporting. New

York, NY: Financial Executives Research Foundation.

Roberts, E. S. (1999). In defence of the survey method: An illustration from a study

of user information satisfaction. Accounting and Finance, 39(1), 53-77.

Ryan, B., Scapens, R. W., & Theobald, M. (1992). Research Method and
Methodology in Finance and Accounting. San Diego, CA: Academic Press Inc.

Ryan, S. G. (2002). Financial Instruments and Institutions: Accounting and

Disclosure Rules. New York, NY: John Wiley & Sons.

Ryan, S. G., Herz, R. H., Iannaconni, T. E., Maines, L. A., Palepu, K., Schrand, C.
M., Skinner, D. J., & Vincent, L. (2002). Reporting fair value interest and value
changes on financial instruments. Accounting Horizons, 16(3), 259-267.

Sanborn, R. H., & Atchison, M. (1985). A CPA's field guide to new financial
instruments. Journal of Accountancy, 160(4), 165-172.

Sarantakos, S. (1993). Social Research. Melbourne, VIC: Macmillan Education

Australia Pty Ltd.

Schipper, K. (1991). Analyst's forecasts. Accounting Horizons, 5(4), 105-121.

266
Scott, W. R. (2003). Financial Accounting Theory (Third Edition). Toronto, Ontario:
Pearson Education.

Shaffer, M. A., & Harrison, D. A. (2001). Forgotten partners of international

assignments: Development and test of a model of spouse adjustment. Journal of
Applied Psychology, 86 (2), 238-254.

Silverman, D. (1993). Interpreting Qualitative Data: Strategies for Analysing Talk,

Text and Interaction. London, UK: Sage Publications Ltd.

Singer, E., & Presser, S. (1989). Survey Research Methods. Chicago, IL: University
of Chicago Press.

Sirota Consulting. (1998). Investment Community Interest in Reporting the Fair

Values of Financial Instruments in Financial Statements - A Focus Group Summary:
Final Report. Retrieved 28th November, 2004, from http://accounting.rutgers.edu/
raw/fasb/project/sirota.pdf.

Smith Barney. (2004). Equity Research: Banks. Melbourne.

Soewarso, E., Tower, G., Hancock, P., & Taplin, R. (2003). Comparative study of de
jure and de facto disclosure between Australia and Singapore. Asian Review of
Accounting, 11(1), 18-47.

Solomons, D. (1986). The FASB's conceptual framework: An evaluation. Journal of

Accountancy, 161(6), 114-121.

Sproull, N. L. (1995). Handbook of Research Methods (Second Edition). London,

UK: The Scarecrow Press.

Staubus, G. (1961). A Theory of Accounting Investors. Berkeley, CA: University of

California Press.

Stevenson, K. (2004a). Personal Interview on Accounting for Financial Instruments.

Alice Springs, Northern Territory, Australia.

Stevenson, K. (2004b). Plenary Session 1: International Financial Reporting

Standards: Perspectives and Prognostications. Speech to the delegates of the 2004
Accounting and Finance Association of Australia and New Zealand (AFAANZ)
Annual Conference, Alice Springs, Northern Territory, Australia.

Stewart, J. E., & Neuhausen, B. S. (1986). Financial instruments and transactions:

The CPA's newest challenge. Journal of Accountancy, 162(2), 102-112.

267
Study Group on the Objectives of Financial Statements. (1973). Objectives of
Financial Statements. New York, NY: American Institute of Certified Public
Accountants.

Sutton, R. I., & Rafaeli, A. (1988). Untangling the relationship between displayed
emotions and organizational sales: The case of convenience stores. Academy of
Management Journal, 31(3), 461-487.

Swift, B. (1979). Design of Surveys. Unpublished manuscript, The Open University,

UK.

Tan, C. W., Taplin, R., Hancock, P., & Tower, G. (2003). Non-response bias in the
use of annual financial reports. Asian Review of Accounting, 11(2), 73-85.

Teagarden, M. B., Von Glinow, M. A., Bowen, D. E., Frayne, C. A., Nason, S., Huo,
Y. P., Milliman, J., Arias, M. E., Butler, M. C., Geringer, J. M., Kim, N. H., Scullion,
H., Lowe, K. B., & Drost, E. A. (1995). Toward a theory of comparative
management research: An idiographic case study of the best international human
resources management project. Academy of Management Journal, 38 (5), 1261-1287.

Tabachnick, B. G. & Fidell, L. S. (1996). Using Multivariate Statistics. New York,

NY: Harper Collins College Publishers.

Ticehurst, G. W., & Veal, A. J. (2000). Business Research Methods: A Managerial

Approach. NSW, Australia: Pearson Education Australia.

Tosen, G. (2003). The tainting dilemma. Accountancy SA, June, 13.

Tower, G. (1993). A public accountability model of accounting regulation. The

British Accounting Review, 25(1), 61-85.

Trevelyan, R. (2001). The paradox of autonomy: A case of academic research

scientists. Human Relations, 54(4), 495-525.

Van Goor, H., & Stuiver, B. (1998). Can weighting compensate for nonresponse bias
in a dependent variable? An evaluation of weighting methods to correct for
substantive bias in a mail survey among Dutch municipalities. Social Science
Research, 27(4), 481-499.

Walker, R. G. (1990). History repeats itself in standard-setting process. New

Accountant, 6(1), 22-23.

Walker, R. G. (1993). A feeling of deja vu: Controversies in accounting and auditing

regulation in Australia. Critical Perspectives on Accounting, 4(1), 97-109.

268
Walker, R. G. (2003). Objectives of financial reporting. Abacus, 39(3), 340-355.

Walker, R. G., & Jones, S. (2003). Measurement: A way forward. Abacus, 39(3),
356-374.

Weitz, J. (1990). Telephone interviewing. In R. Birn, P. Hague & P. Vangelder

(Eds.), A Handbook of Market Research Techniques. London, UK: Kogan Page.

Willis, D. W. (1998). Financial assets and liabilities - Fair value or historical?

Financial Accounting Series: Status Report 304, August, FASB.

Wilk, R. R. (2001). The impossibility and necessity of re-inquiring: Finding middle

ground in social science. Journal of Consumer Research, 28 (2), 308-312.

Wiseman, F., & MacDonald, P. (1979). Noncontact and refusal rates in consumer
telephone surveys. Journal of Marketing Research, 16(4), 478-484.

Young, J. J. (1996). Institutional thinking: The case of financial instruments.

Accounting, Organizations and Society, 21(5), 487-512.

Young, S. M. (1996). Survey research in management accounting: A critical

assessment. In A. J. Richardson (Ed.), Research Methods in Accounting: Issues and
Debates, Research Monograph 25 (pp.55-68). CGA Canada.

Zikmund, W. G. (1997). Exploring Marketing Research (Second Edition). Fort

Worth, TX: The Dryden Press.

Zikmund, W. G. (1999). Essentials of Marketing Research. Orlando, FL: The Dryden

Press.

269
 APPENDIX A:
AUSTRALIAN PREPARER SURVEY
 Financial Instruments Survey – Financial Institutions

For the purpose of this survey;

1) Fair value means:

i) Market Value, if there is a deep and liquid market for the instrument; or, if market
value is unavailable,
ii) An estimate of the value for which one could realise the asset (extinguish the liability)
in an arm’s length transaction.

2) Fair value accounting refers to the measurement of all financial instruments at fair value
and the recognition of changes in fair value as revenues or expenses in the statement of
financial performance (profit and loss statement) in the period in which they arise.

3) Banking book refers to the raising of funds and the investing of those funds in assets in
order to make a profit from the margin between the amount received on interest bearing
assets and the amount paid on interest bearing liabilities.

4) Trading book refers to transactions undertaken with the objective to profit from
fluctuations in market prices.

Section 1: Fair Value Accounting

1. For each of the following statements, please indicate how strongly you support or oppose it.

Strongly Strongly
Oppose Oppose Neutral Support Support
i. Do you support the concept of 1 2 3 4 5
marking financial instruments in the
banking book to fair value on the
statement of financial position
(balance sheet)?
ii. Do you support the concept of 1 2 3 4 5
marking financial instruments in the
trading book to fair value on the
statement of financial position
(balance sheet)?
iii. Do you support the concept of taking 1 2 3 4 5
changes in fair values of financial
instruments in the banking book to the
statement of financial performance
(profit and loss statement)?
iv. Do you support the concept of taking 1 2 3 4 5
changes in fair values of financial
instruments in the trading book to the
statement of financial performance
(profit and loss statement)?

2. The banking book is so fundamentally different from the trading book to warrant the use of
different accounting measurement bases.

Strongly Disagree Disagree Neutral Agree Strongly Agree

1 2 3 4 5

Page 1 of 4

A-1
 Financial Instruments Survey – Financial Institutions

3. For each of the following statements, please indicate how strongly you agree or disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. Fair value accounting brings financial
reporting in line with current financial 1 2 3 4 5
risk management policies
ii. Fair value accounting improves the
relevance of the information in the 1 2 3 4 5
accounts for users
iii. Fair value accounting promotes the
comparability of balance sheets 1 2 3 4 5
between organizations
iv. Fair value accounting introduces
volatility in reported profits that may 1 2 3 4 5
be misunderstood by users of the
accounts
v. Reliable and independent market
valuations are impossible to obtain for 1 2 3 4 5
some products
vi. The cost of obtaining fair value
information will be unacceptably high 1 2 3 4 5

4. To what extent do you currently use fair value accounting for each of the following
financial instruments?
Never Sometimes Always Not Applicable
i. Trading derivatives 1 2 3 4
ii. Trading securities 1 2 3 4
iii. Hedging derivatives 1 2 3 4
iv. Investments securities 1 2 3 4
v. Loans and receivables 1 2 3 4
vi. Other financial assets 1 2 3 4
vii. Other financial liabilities 1 2 3 4

Section 2: Reliability of Fair Values

5. In your estimation, what percentage of financial assets and liabilities would you be able to
determine a reliable fair value for:

i. Financial assets? ii. Financial liabilities?

0 to 10% ‰ 0 to 10% ‰
11 to 25% ‰ 11 to 25% ‰
26 to 50% ‰ 26 to 50% ‰
51 to 75% ‰ 51 to 75% ‰
76 to 90% ‰ 76 to 90% ‰
91 to 100% ‰ 91 to 100% ‰

Page 2 of 4

A-2
 Financial Instruments Survey – Financial Institutions

6. Of the two categories of financial instruments below, please circle the number that in your
opinion, best represents the reliability of the fair values for each type of financial
instrument.

Highly Unrealiable Neutral Reliable Highly

Unrealiable Reliable

i. Traded financial instruments 1 2 3 4 5

ii. Non-traded financial instruments 1 2 3 4 5

Section 3: General Information

7. Please tick the box which best describes your organisation.

i. One of the big four Australian banks ‰

ii. A non-big four Australian bank ‰
iii. A merchant bank ‰
iv. A foreign bank ‰
v. A corporate treasury ‰
vi. A government treasury ‰
vii. Other ____________________ ‰

8. Please tick the box corresponding to the total assets of your organisation, as reported in the
most recent statement of financial position.

i. Less than $1 million ‰

ii. Between $1 and $10 million ‰
iii. Between $10 and $100 million ‰
iv. Between $100 and $1 billion ‰
v. Between $1 and $10 billion ‰
vi. Between $10 and $100 billion ‰
vii. More than $100 billion ‰

Section 4: Your Information

9. How many years’ experience have you working in financial reporting?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

Page 3 of 4

A-3
 Financial Instruments Survey – Financial Institutions

The following information is strictly confidential and will only be referred to for purposes of
elimination in a second mail-out.

Organisation:

Page 4 of 4

A-4
 APPENDIX B:
AUSTRALIAN USER SURVEY
 Financial Instruments Survey – User

For the purpose of this survey;

A) Fair value means:
i) Market Value, if there is a deep and liquid market for the instrument; or
ii) If market value is unavailable, an estimate of the value for which one could
realise the asset (extinguish the liability) in an arm’s length transaction.

B) Fair value accounting refers to the measurement of all financial instruments at fair
value AND the recognition of changes in fair value as revenues or expenses in the
statement of financial performance (profit and loss statement) in the period in
which they arise.

Section 1: Fair Value Accounting

1. Do you support the use of fair value accounting for all financial instruments?
Strongly Oppose Oppose Neutral Support Strongly Support
1 2 3 4 5

2. For each of the following statements, please indicate how strongly you agree or
disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. Fair value accounting for all financial
instruments provides useful information 1 2 3 4 5
for economic decision making
ii. Fair value accounting ensures that
financial statements show a true and fair 1 2 3 4 5
view of the financial position,
performance and changes in financial
position of an entity
iii. Fair value accounting for all financial
instruments introduces volatility that 1 2 3 4 5
will be misunderstood by users
iv. The benefits of using fair value
accounting for all financial instruments 1 2 3 4 5
outweigh the costs
v. Fair value accounting in financial
statements is understandable by users 1 2 3 4 5
with reasonable knowledge of business
activities and accounting
vi. Fair value accounting provides
information that help users evaluate 1 2 3 4 5
past, present and/or future events
vii. The use of a combination of historical
cost and fair value measurement bases 1 2 3 4 5
that are separated in the financial
statements provides useful information

Page 1 of 4

B-1
 Financial Instruments Survey – User

Section 2: Qualitative Characteristics of Fair Value Accounting

3. For each of the following statements, please indicate how strongly you agree or
disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. The nature of the information resulting
from fair value accounting affects the 1 2 3 4 5
assessment of the risks and
opportunities facing the entity
ii. Information resulting from fair value
accounting is free from material error 1 2 3 4 5
and bias
iii. Fair value accounting represents the
substance and economic reality of the 1 2 3 4 5
underlying transaction
iv. The omission of fair values can cause
the financial statements to be false or 1 2 3 4 5
misleading
v. Fair value accounting enables users to
compare financial statements of an 1 2 3 4 5
entity through time
vi. Fair value accounting enables users to
compare financial statements of 1 2 3 4 5
different entities

Section 3: Trading and Banking Book

For the purpose of this section;
A) Banking book refers to the raising of funds and the investing of those funds in
assets in order to make a profit from the margin between the amount received on
interest bearing assets and the amount paid on interest bearing liabilities.

B) Trading book refers to transactions undertaken with the objective to profit from
fluctuations in market prices.

4. The banking book is so fundamentally different from the trading book to warrant
the use of different accounting measurement bases.
Strongly Disagree Disagree Neutral Agree Strongly Agree
1 2 3 4 5

5. To what extent would you use the fair values of financial instruments in your
analysis if it were available?
Never Sometimes Always
i. Banking book 1 2 3
ii. Trading book 1 2 3

Page 2 of 4

B-2
 Financial Instruments Survey – User

6. To what extent do you think that fair value accounting is relevant for each of the
following financial instruments?
Never Sometimes Always
i. Trading derivatives 1 2 3
ii. Trading securities 1 2 3
iii. Hedging derivatives 1 2 3
iv. Investments securities 1 2 3
v. Loans and receivables 1 2 3
vi. Deposits and payables 1 2 3

Section 4: Reliability of Fair Values

7. Of the two categories of financial instruments below, please circle the number that
in your opinion, best represents the reliability of the fair values for each type of
financial instrument.
Highly Unrealiable Neutral Reliable Highly
Unrealiable Reliable

i. Traded financial instruments 1 2 3 4 5

ii. Non-traded financial instruments 1 2 3 4 5

Section 5: Your Information

8. Please tick the box which best describes you.

i. Assurance and advisory ‰

ii. Bank analyst ‰
iii. Financial analyst ‰
iv. Fund/Trust manager ‰
v. Industry association ‰
vi. Investment analyst ‰
vii. Other _________________ ‰

9. How many years’ experience have you had in financial reporting?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

Page 3 of 4

B-3
 Financial Instruments Survey – User

The following information is strictly confidential and will only be referred to for
purposes of elimination in a second mail-out.

Organisation:

Page 4 of 4

B-4
 APPENDIX C:
SINGAPOREAN PREPARER SURVEY
 Financial Instruments Survey – Financial Institutions

For the purpose of this survey;

1) Fair value means:

i) Market Value, if there is a deep and liquid market for the instrument; or, if market
value is unavailable,
ii) An estimate of the value for which one could realise the asset (extinguish the liability)
in an arm’s length transaction.

2) Fair value accounting refers to the measurement of all financial instruments at fair value
and the recognition of changes in fair value as revenues or expenses in the profit and loss
statement in the period in which they arise.

3) Banking book refers to the raising of funds and the investing of those funds in assets in
order to make a profit from the margin between the amount received on interest bearing
assets and the amount paid on interest bearing liabilities.

4) Trading book refers to transactions undertaken with the objective to profit from
fluctuations in market prices.

Section 1: Fair Value Accounting

1. For each of the following statements, please indicate how strongly you support or oppose it.

Strongly Strongly
Oppose Oppose Neutral Support Support

i. Do you support the concept of

marking financial instruments in 1 2 3 4 5
the banking book to fair value on
the balance sheet?

ii. Do you support the concept of

marking financial instruments in 1 2 3 4 5
the trading book to fair value on
the balance sheet?

iii. Do you support the concept of

taking changes in fair values of 1 2 3 4 5
financial instruments in the
banking book to the profit and loss
statement?

iv. Do you support the concept of

taking changes in fair values of 1 2 3 4 5
financial instruments in the trading
book to the profit and loss
statement?

2. The banking book is so fundamentally different from the trading book to warrant the use of
different accounting measurement bases.

Strongly Disagree Disagree Neutral Agree Strongly Agree

1 2 3 4 5

Page 1 of 4

C-1
 Financial Instruments Survey – Financial Institutions

3. For each of the following statements, please indicate how strongly you agree or disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. Fair value accounting brings financial
reporting in line with current financial 1 2 3 4 5
risk management policies
ii. Fair value accounting improves the
relevance of the information in the 1 2 3 4 5
accounts for users
iii. Fair value accounting promotes the
comparability of balance sheets 1 2 3 4 5
between organizations
iv. Fair value accounting introduces
volatility in reported profits that may 1 2 3 4 5
be misunderstood by users of the
accounts
v. Reliable and independent market
valuations are impossible to obtain for 1 2 3 4 5
some products
vi. The cost of obtaining fair value
information will be unacceptably high 1 2 3 4 5

4. To what extent do you currently use fair value accounting for each of the following
financial instruments?
Never Sometimes Always Not Applicable
i. Trading derivatives 1 2 3 4
ii. Trading securities 1 2 3 4
iii. Hedging derivatives 1 2 3 4
iv. Investments securities 1 2 3 4
v. Loans and receivables 1 2 3 4
vi. Other financial assets 1 2 3 4
vii. Other financial liabilities 1 2 3 4

Section 2: Reliability of Fair Values

5. In your estimation, what percentage of financial assets and liabilities would you be able to
determine a reliable fair value for:

i. Financial assets? ii. Financial liabilities?

0 to 10% ‰ 0 to 10% ‰
11 to 25% ‰ 11 to 25% ‰
26 to 50% ‰ 26 to 50% ‰
51 to 75% ‰ 51 to 75% ‰
76 to 90% ‰ 76 to 90% ‰
91 to 100% ‰ 91 to 100% ‰

Page 2 of 4

C-2
 Financial Instruments Survey – Financial Institutions

6. Of the two categories of financial instruments below, please circle the number that in your
opinion, best represents the reliability of the fair values for each type of financial
instrument.

Highly Unrealiable Neutral Reliable Highly

Unrealiable Reliable

i. Traded financial instruments 1 2 3 4 5

ii. Non-traded financial instruments 1 2 3 4 5

Section 3: General Information

7. Please tick the box which best describes your organisation.

i. A local bank ‰
ii. A foreign bank ‰
iii. A merchant bank ‰
iv. A finance company ‰
v. A corporate treasury ‰
vi. A government treasury ‰
vii. Other ____________________ ‰

8. Please tick the box corresponding to the total assets of your organisation, as reported in the
most recent statement of financial position.

i. Less than S$1 million ‰

ii. Between S$1 and S$10 million ‰
iii. Between S$10 and S$100 million ‰
iv. Between S$100 and S$1 billion ‰
v. Between S$1 and S$10 billion ‰
vi. Between S$10 and S$100 billion ‰
vii. More than S$100 billion ‰

Section 4: Your Information

9. How many years’ experience have you working in financial reporting?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

Page 3 of 4

C-3
 Financial Instruments Survey – Financial Institutions

The following information is strictly confidential and will only be referred to for purposes of
elimination in a second mail-out.

Organisation:

Page 4 of 4

C-4
 APPENDIX D:
SINGAPOREAN USER SURVEY
 Financial Instruments Survey – User

For the purpose of this survey;

A) Fair value means:
i) Market Value, if there is a deep and liquid market for the instrument; or
ii) If market value is unavailable, an estimate of the value for which one could
realise the asset (extinguish the liability) in an arm’s length transaction.

B) Fair value accounting refers to the measurement of all financial instruments at fair
value AND the recognition of changes in fair value as revenues or expenses in the
profit and loss statement in the period in which they arise.

Section 1: Fair Value Accounting

1. Do you support the use of fair value accounting for all financial instruments?
Strongly Oppose Oppose Neutral Support Strongly Support
1 2 3 4 5

2. For each of the following statements, please indicate how strongly you agree or
disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. Fair value accounting for all financial
instruments provides useful information 1 2 3 4 5
for economic decision making
ii. Fair value accounting ensures that
financial statements show a true and fair 1 2 3 4 5
view of the financial position,
performance and changes in financial
position of an entity
iii. Fair value accounting for all financial
instruments introduces volatility that 1 2 3 4 5
will be misunderstood by users
iv. The benefits of using fair value
accounting for all financial instruments 1 2 3 4 5
outweigh the costs
v. Fair value accounting in financial
statements is understandable by users 1 2 3 4 5
with reasonable knowledge of business
activities and accounting
vi. Fair value accounting provides
information that help users evaluate 1 2 3 4 5
past, present and/or future events
vii. The use of a combination of historical
cost and fair value measurement bases 1 2 3 4 5
that are separated in the financial
statements provides useful information

Page 1 of 4

D-1
 Financial Instruments Survey – User

Section 2: Qualitative Characteristics of Fair Value Accounting

3. For each of the following statements, please indicate how strongly you agree or
disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. The nature of the information resulting
from fair value accounting affects the 1 2 3 4 5
assessment of the risks and
opportunities facing the entity
ii. Information resulting from fair value
accounting is free from material error 1 2 3 4 5
and bias
iii. Fair value accounting represents the
substance and economic reality of the 1 2 3 4 5
underlying transaction
iv. The omission of fair values can cause
the financial statements to be false or 1 2 3 4 5
misleading
v. Fair value accounting enables users to
compare financial statements of an 1 2 3 4 5
entity through time
vi. Fair value accounting enables users to
compare financial statements of 1 2 3 4 5
different entities

Section 3: Trading and Banking Book

For the purpose of this section;
A) Banking book refers to the raising of funds and the investing of those funds in
assets in order to make a profit from the margin between the amount received on
interest bearing assets and the amount paid on interest bearing liabilities.

B) Trading book refers to transactions undertaken with the objective to profit from
fluctuations in market prices.

4. The banking book is so fundamentally different from the trading book to warrant
the use of different accounting measurement bases.
Strongly Disagree Disagree Neutral Agree Strongly Agree
1 2 3 4 5

5. To what extent would you use the fair values of financial instruments in your
analysis if it were available?
Never Sometimes Always
i. Banking book 1 2 3
ii. Trading book 1 2 3

Page 2 of 4

D-2
 Financial Instruments Survey – User

6. To what extent do you think that fair value accounting is relevant for each of the
following financial instruments?
Never Sometimes Always
i. Trading derivatives 1 2 3
ii. Trading securities 1 2 3
iii. Hedging derivatives 1 2 3
iv. Investments securities 1 2 3
v. Loans and receivables 1 2 3
vi. Deposits and payables 1 2 3

Section 4: Reliability of Fair Values

7. Of the two categories of financial instruments below, please circle the number that
in your opinion, best represents the reliability of the fair values for each type of
financial instrument.
Highly Unrealiable Neutral Reliable Highly
Unrealiable Reliable

i. Traded financial instruments 1 2 3 4 5

ii. Non-traded financial instruments 1 2 3 4 5

Section 5: Your Information

8. Please tick the box which best describes you.

i. Assurance and advisory ‰

ii. Bank analyst ‰
iii. Financial analyst ‰
iv. Fund/Trust manager ‰
v. Industry association ‰
vi. Investment analyst ‰
vii. Other _________________ ‰

9. How many years’ experience have you had in financial reporting?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

Page 3 of 4

D-3
 Financial Instruments Survey – User

The following information is strictly confidential and will only be referred to for
purposes of elimination in a second mail-out.

Organisation:

Page 4 of 4

D-4
 APPENDIX E:
PROTOTYPE OF PREPARER SURVEY

Section 1: Reporting Requirements and Current Developments

1. Under what accounting regime/s does your organisation prepare financial records?
(Please tick all relevant standards).

i. Australian Accounting Standards ‰

ii. US Accounting Standards ‰
iii. International Accounting Standards ‰
iv. Other (Please Explain): ________________ ‰

2. Has your organisation researched the impact of international accounting developments

for financial instruments (FAS 133 or IAS 39) on your business?

YES ‰ NO ‰ UNKNOWN ‰

3. Please circle the number that best indicates your understanding of the impact of any of
the following standards on your business in the future:

FAS 133 “Accounting for Derivative Instruments and Hedging Activities”

Nil Trivial Limited Substantial Very Substantial

1 2 3 4 5

IAS 39 “Financial Instruments: Recognition and Measurement”

Nil Trivial Limited Substantial Very Substantial

1 2 3 4 5

Section 2: Traditional Banking Activities versus Trading Activities

4. Does your organisation differentiate traditional banking activities from trading

activities for financial reporting practices?

YES ‰ NO ‰

5. Does your organisation differentiate traditional banking activities from trading

activities for internal management purposes?

YES ‰ NO ‰

E-1
6. In your opinion, what are the differences between traditional banking activities from
trading activities? (Please tick all relevant answers).

i. There are no fundamental difference between the two ‰

ii. The income earning process for the two areas is completely different ‰
iii. Banking activities are based on long- term relationships ‰
iv. Banking activities relate to retail and commercial banking business ‰
v. The objective of banking activities is to profit from earning a margin
between the amount received on interest-earning assets and the amount
paid on interest-bearing liabilities ‰
vi. Trading activities entail actively dealing and operating in financial markets ‰
vii. The objective of trading activities is to profit from short-term fluctuations
in the financial markets ‰

7. To what extent are financial instruments in the banking books of your organisation
currently being measured at their fair values?

NONE ‰ SOME ‰ ALL ‰

8. To what extent are financial instruments in the trading books of your organisation
currently being measured at their fair values?

NONE ‰ SOME ‰ ALL ‰

9. Are changes in the fair values of financial instruments being taken to the current
year’s profit and loss statement?

NEVER ‰ SOMETIMES ‰ ALWAYS ‰

10. Are changes in the fair values of financial instruments being taken to equity in the
current year’s balance sheet?

NEVER ‰ SOMETIMES ‰ ALWAYS ‰

Section 3: Fair Value Accounting

In this section, fair value accounting refers to the measurement of all financial instruments at
fair value and the recognition of changes in fair value as revenues or expenses in the
statement of financial performance in the period in which they arise.

11. Do you support the concept of marking all financial instruments to fair value on the
balance sheet?

Strongly Oppose Oppose Neutral Support Strongly Support

1 2 3 4 5

E-2
12. The following statements describe some issues relating to the use of fair value
accounting for financial instruments. For each statement, please indicate how strongly
you agree or disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. It ensures that reported profits adequately
reflect changes in economic conditions 1 2 3 4 5
ii. It brings financial reporting in line with
current financial risk management policies 1 2 3 4 5
iii. It improves the relevance of the
information in the accounts 1 2 3 4 5
iv. It promotes the comparability of balance
sheets between organisations 1 2 3 4 5
v. It removes the reliance on management
intention to determine treatment 1 2 3 4 5
vi. The practice introduces volatility in
reported profits that may be 1 2 3 4 5
misunderstood by users of the accounts
vii. Reliable and independent market
valuations are impossible to obtain for 1 2 3 4 5
some products
viii. Fair value is not appropriate for financial
assets and liabilities intended to be held to 1 2 3 4 5
maturity
ix. The amounts stated may not be truly
realisable because an established 1 2 3 4 5
secondary market for some assets or
liabilities may not exist
x. The cost of obtaining the information will
be unacceptably high 1 2 3 4 5
xi. The increased volatility in reported
earnings arising from fair value
accounting will create increased difficulty 1 2 3 4 5
in projecting and meeting earnings
forecast
xii. The increased volatility in reported
earnings arising from fair value
accounting will cause problems due to the 1 2 3 4 5
reaction of investors to reported earnings
xiii. The increased volatility in reported
earnings arising from fair value
accounting will result in transient changes 1 2 3 4 5
in reported earnings not related to
underlying economic activity
xiv. The use of fair values leads to the
disclosure of commercially sensitive 1 2 3 4 5
information

E-3
13. What financial instruments do you currently value at fair value? (Please tick all
relevant instruments):

i. Trading derivatives ‰
ii. Trading securities ‰
iii. Hedging derivatives ‰
iv. Investments securities ‰
v. Loans and receivables ‰
vi. Other financial assets ‰
vii. Other financial liabilities ‰

14. For this question, please rank the following financial instruments in the order of those
you deem to be the easiest to determine a fair value for to those you deem to be
hardest to determine a fair value for. The most easily determined should be “1” while
the hardest should be “14”.

i. Loans and receivables _____

ii. Credit guarantees _____
iii. Trading securities _____
iv. Trading derivatives _____
v. Hedging derivatives _____
vi. Investment Securities _____
vii. Related party transactions _____
viii. Non-traded options _____
ix. Non-traded other derivatives _____
x. Non-traded equity instruments _____
xi. Non-traded short term financial assets _____
xii. Non-traded longer term financial assets _____
xiii. Non-traded short term financial liabilities _____
xiv. Non-traded longer term financial liabilities _____

15. For this question, please rank the following list of factors in the order of those you
deem to be the more important factors in the determination of fair value for financial
instruments to those you deem least important in the determination of fair value for
financial instruments. The most important factor should be “1” while the least
important should be “10”.

i. Illiquid markets _____

ii. Ongoing capital usage _____
iii. Future estimated transaction costs _____
iv. Ongoing credit usage _____
v. Future ongoing administration/servicing costs _____
vi. Basis risk _____
vii. Any market anomaly that exists at the time _____
viii. Volume or size of book relative to size of market _____
ix. Transaction costs incurred _____
x. Aging allowance if an asset is held in the portfolio
for an extended period _____

E-4
Section 4: Reliability of Fair Values

16. In your estimation, what percentage of financial assets and liabilities would you not
be able to determine a reliable fair value for:

i. Financial assets?
0 to 25% ‰
26 to 50% ‰
51 to 75% ‰
76 to 100% ‰

ii. Financial liabilities?

0 to 25% ‰
26 to 50% ‰
51 to 75% ‰
76 to 100% ‰

17. Of the list of financial instruments below, please circle the number that in your
opinion, best represents the reliability of the estimates of fair values for each type of
financial instrument.

Highly Unreliable Unreliable Neutral Reliable Highly Reliable

1 2 3 4 5

i. Loans and receivables 1 2 3 4 5

ii. Credit guarantees 1 2 3 4 5
iii. Related party transactions 1 2 3 4 5
iv. Trading derivatives 1 2 3 4 5
v. Trading securities 1 2 3 4 5
vi. Hedging derivatives 1 2 3 4 5
vii. Investments securities 1 2 3 4 5
viii. Non-traded options 1 2 3 4 5
ix. Non-traded other derivatives 1 2 3 4 5
x. Non-traded equity instruments 1 2 3 4 5
xi. Non-traded short term financial assets 1 2 3 4 5
xii. Non-traded longer term financial assets 1 2 3 4 5
xiii. Non-traded short term financial liabilities 1 2 3 4 5
xiv. Non-traded longer term financial liabilities 1 2 3 4 5

E-5
Section 5: General Information

18. Please tick the box which best describes your organisation.

i. One of the Big 4 Australian Banks ‰

ii. A non-Big 4 Australian Banks ‰
iii. A credit union ‰
iv. A merchant bank ‰
v. A government treasury ‰
vi. A foreign bank ‰
vii. Other ____________________ ‰

19. Please tick the box which best describes your organisation.

i. Australian based ‰
ii. A subsidiary of a foreign entity ‰
iii. A branch of a foreign entity ‰

20. Please indicate which category the size of your organisation belongs to.

i. Total Assets < $1 billion ‰

ii. Total Assets $1 billion to $10 billion ‰
iii. Total Assets $10 billion to $50 billion ‰
iv. Total Assets $51 billion to $100 billion ‰
v. Total Assets > $100 billion ‰

Section 6: Respondent Data

21. Please indicate your gender.

i. Male ‰
ii. Female ‰

22. Please indicate the age group that you belong to.

i. 18 to 24 years ‰
ii. 25 to 34 years ‰
iii. 35 to 44 years ‰
iv. 45 to 54 years ‰
v. 55 to 64 years ‰
vi. Over 65 years ‰

E-6
23. Please indicate your highest academic qualifications.

i. Doctorate ‰
ii. Masters ‰
iii. Undergraduate ‰
iv. Diploma ‰
v. None of the above ‰

24. How many years’ experience have you had in financial accounting and management?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

25. How many years’ experience have you had in the financial institutions’ industry?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

E-7
 APPENDIX F:
PILOT-TESTED PREPARER SURVEY
 Financial Instruments Survey - Preparer

For the purpose of this survey;

1) Fair value means:

i) Market Value, if there is a deep and liquid market for the instrument; or,
if market value is unavailable,
ii) An estimate of the value for which one could realise the asset (extinguish
the liability) in an arm’s length transaction.

2) Fair value accounting refers to the measurement of all financial instruments

at fair value and the recognition of changes in fair value as revenues or
expenses in the statement of financial performance (profit and loss statement)
in the period in which they arise.

3) Banking book refers to the raising of funds and the investing of those funds
in assets in order to make a profit from the margin between the amount
received on interest bearing assets and the amount paid on interest bearing
liabilities.

4) Trading book refers to transactions undertaken with the objective to profit

from fluctuations in market prices.

Section 1: Fair Value Accounting

1. Do you support the concept of marking financial instruments in the banking

book to fair value on the statement of financial position (balance sheet)?
Strongly Oppose Oppose Neutral Support Strongly Support
1 2 3 4 5

2. Do you support the concept of marking financial instruments in the trading

book to fair value on the statement of financial position (balance sheet)?
Strongly Oppose Oppose Neutral Support Strongly Support
1 2 3 4 5

3. Do you support the concept of taking changes in fair values of financial

instruments in the banking book to the statement of financial performance
(profit and loss statement)?
Strongly Oppose Oppose Neutral Support Strongly Support
1 2 3 4 5

4. Do you support the concept of taking changes in fair values of financial

instruments in the trading book to the statement of financial performance
(profit and loss statement)?
Strongly Oppose Oppose Neutral Support Strongly Support
1 2 3 4 5

Page 1 of 6

F-1
 Financial Instruments Survey - Preparer

5. For each of the following statements, please indicate how strongly you agree or
disagree.
Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. Fair value accounting ensures that reported
profits adequately reflect changes in 1 2 3 4 5
economic conditions
ii. Fair value accounting brings financial
reporting in line with current financial risk 1 2 3 4 5
management policies
iii. Fair value accounting improves the
relevance of the information in the accounts 1 2 3 4 5
for users
iv. Fair value accounting promotes the
comparability of balance sheets between 1 2 3 4 5
organizations
v. Fair value accounting removes the reliance
on management intention to determine the 1 2 3 4 5
accounting treatment
vi. Fair value accounting introduces volatility
in reported profits that may be 1 2 3 4 5
misunderstood by users of the accounts
vii. Reliable and independent market valuations
are impossible to obtain for some products 1 2 3 4 5
viii. Fair value is not appropriate for financial
assets and liabilities intended to be held to 1 2 3 4 5
maturity
ix. The cost of obtaining fair value
information will be unacceptably high 1 2 3 4 5
x. The increased volatility in reported earnings
arising from fair value accounting will
create increased difficulty in projecting and 1 2 3 4 5
meeting earnings forecast
xi. The increased volatility in reported earnings
arising from fair value accounting will
cause problems due to the reaction of 1 2 3 4 5
investors to reported earnings
xii. The increased volatility in reported earnings
arising from fair value accounting will
result in transient changes in reported 1 2 3 4 5
earnings not related to underlying economic
activity
xiii. The use of fair values leads to the
disclosure of commercially sensitive 1 2 3 4 5
information
xiv. Fair value information clarifies
information on cash flows 1 2 3 4 5

Page 2 of 6

F-2
 Financial Instruments Survey - Preparer

6. To what extent do you currently measure (and would prefer to measure) each
of the following financial instruments at fair value in the statement of
financial position (balance sheet)?

Current Practice Preferred method

None Some All Not None Some All

Applicable

i. Trading derivatives 1 2 3 4 1 2 3
ii. Trading securities 1 2 3 4 1 2 3
iii. Hedging derivatives 1 2 3 4 1 2 3
iv. Investments securities 1 2 3 4 1 2 3
v. Loans and receivables 1 2 3 4 1 2 3
vi. Other financial assets 1 2 3 4 1 2 3
vii. Other financial liabilities 1 2 3 4 1 2 3

7. To what extent do you currently take (and would prefer to take) movements
in fair value of each of the following financial instruments to the statement of
financial performance (profit and loss statement)?

Current Practice Preferred method

None Some All Not None Some All
Applicable

i. Trading derivatives 1 2 3 4 1 2 3
ii. Trading securities 1 2 3 4 1 2 3
iii. Hedging derivatives 1 2 3 4 1 2 3
iv. Investments securities 1 2 3 4 1 2 3
v. Loans and receivables 1 2 3 4 1 2 3
vi. Other financial assets 1 2 3 4 1 2 3
vii. Other financial liabilities 1 2 3 4 1 2 3

8. To what extent do you measure financial instruments in the banking books at

cost for financial reporting purposes but at fair value for internal management
purposes?
Never Rarely Sometimes Mostly Always
1 2 3 4 5

9. To what extent do you measure financial instruments in the trading books at

cost for financial reporting purposes but at fair value for internal management
purposes?
Never Rarely Sometimes Mostly Always
1 2 3 4 5

Page 3 of 6

F-3
 Financial Instruments Survey - Preparer

10. For each of the following statements in relation to the banking and trading
book, please indicate how strongly you agree or disagree.

Strongly Strongly
Disagree Disagree Neutral Agree Agree
i. There is no fundamental difference
between the banking and trading book in 1 2 3 4 5
relation to fair value accounting
ii. The income earning process for the
banking book is different to the trading 1 2 3 4 5
book
iii. Banking book activities are based on
long- term relationships 1 2 3 4 5
iv. Banking book activities relate to retail and
commercial banking business 1 2 3 4 5
v. The objective of the banking book is to
profit from earning a margin between the 1 2 3 4 5
amount received on interest-earning assets
and the amount paid on interest-bearing
liabilities
vi. Trading book activities entail actively
dealing and operating in financial markets 1 2 3 4 5
vii. The objective of the trading book is to
profit from short-term fluctuations in the 1 2 3 4 5
financial markets

Section 2: Reliability of Fair Values

11. In your estimation, what percentage of financial assets and liabilities would
you be able to determine a reliable fair value for:

i. Financial assets? ii. Financial liabilities?

0 to 10% ‰ 0 to 10% ‰
11 to 25% ‰ 11 to 25% ‰
26 to 50% ‰ 26 to 50% ‰
51 to 75% ‰ 51 to 75% ‰
76 to 90% ‰ 76 to 90% ‰
91 to 100% ‰ 91 to 100% ‰

12. Of the two categories of financial instruments below, please circle the
number that in your opinion, best represents the reliability of the fair values
for each type of financial instrument.
Highly Unrealiable Neutral Reliable Highly
Unrealiable Reliable

i. Traded financial instruments 1 2 3 4 5

ii. Non-traded financial instruments 1 2 3 4 5

Page 4 of 6

F-4
 Financial Instruments Survey - Preparer

Section 3: General Information

13. Please tick the box which best describes your organisation.

i. One of the big four Australian banks ‰

ii. A non-big four Australian bank ‰
iii. A credit union ‰
iv. A merchant bank ‰
v. A government treasury ‰
vi. A foreign bank ‰
vii. Other ____________________ ‰

14. Please tick the box which best describes your organisation.

i. Australian based ‰
ii. A subsidiary of a foreign entity ‰
iii. A branch of a foreign entity ‰

15. Please tick the box corresponding to the total assets of your organisation, as
reported in the most recent statement of financial position.

i. Less than $1 million ‰

ii. Less than $10 million ‰
iii. Less than $100 million ‰
iv. Less than $1 billion ‰
v. Less than $10 billion ‰
vi. Less than $100 billion ‰
vii. More than $100 billion ‰

Section 4: Your Information

16. How many years’ experience have you working in accounting and financial
reporting?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

Page 5 of 6

F-5
 Financial Instruments Survey - Preparer

17. How many years’ experience have you had working in the financial
institutions’ industry?

i. Less than 2 years ‰

ii. 2 to 5 years ‰
iii. 6 to 10 years ‰
iv. 11 to 15 years ‰
v. More than 15 years ‰

Page 6 of 6

F-6
 APPENDIX G:
LAST PHONE SURVEY QUESTION

One final question in the second phone survey asked for the user’s thoughts on the

low response rate for a separate mail survey sent to sophisticated users even though it

provides them with the opportunity for them to make their voices heard. They were

given six different choices of possible reasons43 with a seventh option of “other”.

Table G1 lists the reasons and the number of respondents who picked each one44.

Table G1: Phone Surveyed Users’ Views on the Reasons for Low Response
Rate
Reason/choice Number of respondents
Lack of knowledge 2
Too busy 4
Don’t think their input has any influence over the process 5
Don’t care – it is not relevant to them 0
Will worry about it when time comes 3
Other 0

The evidence in Table 9 shows a general consensus among the respondents that most

users are too busy with professional commitments to have time to respond to

surveys. Two respondents also think that there is a lack of adequate knowledge on

fair value accounting to be able to complete a survey. Some respondents surmise

that most users will worry about the full fair value accounting model when it

becomes mandated. The biggest concern is that users do not perceive their input to

have an influence over the standard-setting process. This is despite the supposedly

user-oriented approach that an accounting standard-setting process is supposed to

adopt.

43
These reasons were volunteered by the respondents to the first phone survey.
44
Respondents could choose more than one reason.

G-1
 APPENDIX H:
THE PILOT STUDY

H1. INTRODUCTION

For the purposes of pilot-testing, the preparer survey was sent to 212 chief financial

officers in the big-4 Australian banks, 20 non big-4 banks, 2 foreign banks, 168

credit unions and 18 building societies in Australia. Of the 29 (13.6% response rate)

surveys that were received there were two big four Australian banks, two non-big

four Australian banks, 21 credit unions and four building societies. This means that

responses were received from 50% of the big four banks, 10% of the other banks,

12.5% of the credit unions and 22% of the building societies. The respondents were

all Australian based financial institutions.

Among the 29 respondent financial institutions, two had assets of between 1 and 10

million, while the majority (20 financial institutions) had assets between 100 million

and 1 billion. The remaining seven had assets in excess of 1 billion. Virtually all the

chief financial officers that responded had extensive accounting and banking

experience. All but one (preparer with between 2 and 5 years experience) had at

least 6 to 10 years of experience and thirteen respondents (45%) had more than 15

years experience.

This 13.6% response rate is low and thus, complete generalisation to the financial

industry is inappropriate. However, as the big four banks make up a significant

voice in the JWGBA, this study’s strength lies in the ability to provide insights into

the views of all other financial institutions in Australia (especially the credit unions)

H-1
to assess whether the JWGBA view is indeed universal. Overall, the results provide

insight of the perception of the Australian financial institutions’ community.

H2. RESULTS

H2.1 Current Practice and Preference for Fair Value Accounting

In the survey, chief financial officers were asked to indicate the extent that fair value

accounting is used and preferred for different financial instruments. Survey

responses indicate no difference in current and preferred accounting practice for

financial instruments. The evidence indicates that financial institutions prefer to use

the accounting method that is currently applied. Table H.1 shows the frequencies of

survey responses documenting their practices of measuring different types of

financial instruments at fair value in the Statement of Financial Position. As the

sample size indicates, respondents that do not have that particular type of financial

instrument were excluded.

Table H.1: The extent that financial instruments are measured at fair value in
the Statement of Financial Position.

Category n Type of Financial Instruments None Some All

Trading 6 Trading Derivatives 2 1 3

11 Trading Securities 3 1 7

Banking 9 Hedging Derivatives 5 3 1

22 Investment Securities 11 2 9

29 Loans and receivables 13 5 11

28 Other Financial Assets 12 7 9

28 Other Financial Liabilities 14 6 8

H-2
As shown in Table H.1 not all financial instruments in the trading book are currently

measured at fair value. This contradicts expectations as the literature suggests that

fair value is used. There are a few chief financial officers who stated that none of the

trading derivatives and/or trading securities are measured at fair value. Conversely,

the results indicate that as many as 50% of the respondents use fair value for the

financial instruments in the banking book (in this case, loans and receivables).

Table H.2 shows the respondents’ tendency to take resultant changes in fair values of

the different types of financial instruments to the Statement of Financial

Performance. This is the other half of the requirement of the proposed fair value

accounting model.

Table H.2: The extent that changes in fair value of each of the financial
instruments is taken to the Statement of Financial Performance.

Category n Type of Financial Instruments None Some All

Trading 4 Trading Derivatives 0 1 3

8 Trading Securities 1 1 6

Banking 4 Hedging Derivatives 1 2 1

11 Investment Securities 2 4 5

16 Loans and receivables 3 5 8

16 Other Financial Assets 3 7 6

14 Other Financial Liabilities 3 6 5

The results in Table H.2 show that the majority of respondents indicate that fair value

changes for some if not all financial instruments are recognised in the Statement of

Financial Performance. This applies to all instruments irrespective of whether

classified as trading and banking. This result is somewhat surprising as there is a

general expectation, and this is supported by the views expressed by the JWGBA

H-3
(JWGBA 1999b), that changes in the fair values of financial instruments in the

banking book would not be immediately recognised in the Statement of Financial

Performance.

Responses showed that financial institutions are more likely to use fair value

accounting for trading financial instruments as compared to those in the banking

book. However, the level of agreement as expressed by respondents to this survey, is

not as extreme as asserted the members of the JWGBA.

H2.2 Preparer Support for Fair Value Accounting

The respondents are asked four questions to determine the level of support for fair

value accounting. Their responses are analysed as follows. Firstly, the responses of

strongly oppose, oppose, neutral, support, and strongly support are replaced with the

values 1, 2, 3, 4 and 5 respectively to quantify the strength of support for each

question. Then, a one-sample t-test is used to compare the means to the midpoint

position (3 on the 5-point scale). Significant p-values indicate that the means are

significantly different from 3 (i.e. a neutral response). This statistical technique is

also used in Tables H.5 and H.6.

As shown in Table H.3, there is general support (means of 3.7 and 3.8, highly

significant at the .01 level) for fair value accounting in trading book but opposition

(means of 2.3 and 2.2, again highly significant at the .01 level) to its use for financial

instruments in the banking book. The number of managers supporting or strongly

supporting are also significantly different to the number of managers opposing or

strongly opposing. This is achieved by testing whether the proportion of managers

strongly supporting or supporting was significantly different to 0.5 after removing

H-4
neutral managers. The p-values for these proportion tests in Table H.3 are 0.006,

0.002, 0.029 and 0.029 respectively. Since these two approaches result in similar

conclusions, only means and associated p-values from a t-test for a mean of 3

(neutral) are presented here and for future analyses.

Table H.3: Level of support for fair value accounting.

Strongly

Strongly
Support

Support

p-value
Neutral

Means
(n=29)
Oppose

Oppose
Question

Do you support the concept of

marking financial instruments in 3 3 2 12 9 3.7*** .006
the trading book to fair value on
the Statement of Financial
Position (Balance Sheet)?

Do you support the concept of

taking changes in fair values of 3 2 2 13 9 3.8*** .002
financial instruments in the
trading book to the Statement of
Financial Performance (Profit and
Loss Statement)?

Do you support the concept of

marking financial instruments in 10 9 3 6 1 2.3*** .004
the banking book to fair value on
the Statement of Financial
Position (Balance Sheet)?

Do you support the concept of

taking changes in fair values of 11 8 3 6 1 2.2*** .003
financial instruments in the
banking book to the Statement of
Financial Performance (Profit and
Loss Statement)?

Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.

* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

A cross tabulation shows a significant correlation between the responses to the first

two questions (r = 0.94, p = 0.000) as well as the responses to the second pair of

question (r = 0.9, p = 0.000). However, the results indicate that at least for the

respondents to the survey, the support for the use of fair value accounting in the

H-5
trading book and the opposition to its use in the banking book is not unanimous.

Statistically, the respondents’ support for fair value accounting in the trading book

and in the banking book is not strongly related (correlation, r = 0.237). Mann

Whitney test for non-response bias reveals insignificant differences between the

preferences for fair value accounting from the first 40% of respondents and last 40%

of respondents to the survey.

An independent sample’s t-test is conducted to determine if size was a determinant

for the respondents’ level of support for fair value accounting. In this capacity, the

respondents are split into two groups based on the size of their assets. The seven

financial institutions with more than 1 billion (essentially the larger banks) are

grouped together and smaller institutions made up the second group. Results of the

analyses are shown in Table H.4.

Table H.4: Independent samples t-test results on the significance of size on

respondents’ support for fair value accounting.

Means (n=29)
Question p-value
Big Small
Firms Firms
Do you support the concept of marking financial 4.3 3.5 .171
instruments in the trading book to fair value on the
Statement of Financial Position (Balance Sheet)?
Do you support the concept of taking changes in fair 4.3 3.6 .222
values of financial instruments in the trading book to
the Statement of Financial Performance (profit and
loss statement)?
Do you support the concept of marking financial 1.4 2.5 .003***
instruments in the banking book to fair value on the
Statement of Financial Position (Balance Sheet)?
Do you support the concept of taking changes in fair 1.3 2.5 .001***
values of financial instruments in the banking book
to the Statement of Financial Performance (profit and
loss statement)?
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

H-6
There is a clear difference between the answers given by chief financial officers of

big and small financial institutions (p-values of .003 and .001) for the banking book.

The big firms oppose the use of fair value accounting for the banking book more than

the smaller ones. For additional analyses, the financial institutions are also divided

based on a big-4 and non big-4 category. A big-4 versus non big-4 split indicates

that the big-4 banks are on the extreme, as expected.

The survey also asks the extent to which respondents agree (or disagree) with the

assertion that trading and banking books are fundamentally different. Table H.5 lists

the questions asked, their respective means and whether the means are significantly

difference to 3 (neutral). On the asserted difference between the banking and trading

books, the JWGBA’s logic is substantiated (means ranging from 3.7 to 4.2).

However, a mean of 2.4 (on a 5-point scale) shows weaker support for the

fundamental difference argument.

Table H.5: Level of support for the JWGBA reasons for the fundamental
difference between the trading and banking books
Strongly

Strongly
Support

Support
Neutral
Oppose

Oppose

p-value
Means
(n=29)

Question

There is no fundamental difference

between the banking and trading book 5 11 9 4 0 2.4* .002
in relation to fair value accounting.
The income earning process for the
banking book is different to the trading 0 1 6 18 14 3.9* .000
book.
Banking book activities are based on 0 2 7 19 1 3.7* .000
long- term relationships.
Banking book activities relate to retail 0 0 7 21 1 3.8* .000
and commercial banking business.
The objective of the banking book is to
profit from earning a margin between 0 0 2 19 8 4.2* .000
the amount received on interest-
earning assets and the amount paid on
interest-bearing liabilities.

H-7
Table H.5: Level of support for the JWGBA reasons for the fundamental
difference between the trading and banking books (continued)

Strongly

Strongly
Support

Support
Neutral
Oppose

Oppose

p-value
Means
(n=29)
Question

Trading book activities entail actively

dealing and operating in financial 0 1 3 21 4 4.0* .000
markets.
The objective of the trading book is to
profit from short-term fluctuations in 1 0 9 17 2 3.7* .000
the financial markets.
Legend: 1 = Strongly Oppose; 2 = Oppose; 3 = Neutral; 4 = Support; 5 = Strongly Support
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

In addition, financial institutions do not usually (mean = 2.2) measure financial

instruments in the banking books at cost for financial reporting purposes but at fair

value for internal management purposes. This response is an anomaly as a strong

never response (choice 1), is expected, based on the past literature. Therefore,

contrary to JWGBA assertions, it seems that fair values are known and used for

financial instruments in the banking book for internal management purposes.

Further t-tests indicate that the means are very similar between big and small

financial institutions and thus, not significantly different between the two groups.

Again, when the big-4 versus non big-4 split is used, the big-4 banks sit on the

extreme.

H2.2.1 Why financial institutions think that way?

The resistance towards fair value accounting is further examined with questions that

assess the level of support for the reasons against fair value accounting as enunciated

by the JWGBA.

H-8
Table H.6: Level of support for the JWGBA views on the reasons against fair
value accounting.

Disagree
Disagree
Strongly

Strongly

p-value
Means
(n=29)
Neutral
JWGBA Views

Agree

Agree
Fair value accounting introduces
volatility in reported profits that may 0 2 3 14 10 4.1* .000
be misunderstood by users of the
accounts.
Fair value accounting is not
appropriate for financial instruments 1 0 5 17 6 3.9* .000
in the banking book.
Reliable fair values are impossible to 1 4 2 16 6 3.8* .001
obtain for some products.
Fair value accounting impedes the
relevance of the information in the 0 6 8 10 5 3.5** .017
accounts for users.
The cost of obtaining fair value 0 5 12 9 3 3.3** .048
information will be unacceptably
high.
Fair value accounting does not 1 7 7 9 5 3.4 .115
promote the comparability between
organisations.
The use of fair values leads to the
disclosure of commercially sensitive 3 13 9 4 0 2.8 .380
information.
Legend: 1 = Strongly Disagree; 2 = Disagree; 3 = Neutral; 4 = Agree; 5 = Strongly Agree
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Respondents tend to agree with the JWGBA logic with most means ranging from 3.3

to 4.1 (see Table H.6). The only contradiction is that respondents are neutral in

relation to the purported commercial sensitivity of fair value information (mean =

2.8) as well as the comparability of fair value accounting financial statements (mean

= 3.4).

Similarly, another independent sample’s t-test is conducted to determine if size (and

big-4 or non big-4 voices) was a determinant for the responses. Table H.7 depicts

these results.

H-9
Table H.7: Independent samples t-test results on the significance of size on
respondents’ support for JWGBA views.

Means (n=29)
Question p-value
Big Small
Firms Firms
Fair value accounting introduces volatility in 4.7 3.9 .006*
reported profits that may be misunderstood by users
of the accounts.
Fair value accounting is not appropriate for financial 4.1 3.9 .268
instruments in the banking book.
Reliable fair values are impossible to obtain for 4.4 3.5 .009*
some products.
Fair value accounting impedes the relevance of the 4.4 3.2 .000*
information in the accounts for users.
The cost of obtaining fair value information will be 3.3 3.4 .828
unacceptable high.
Fair value accounting does not promote the 4.3 3.0 .027*
comparability between organisations.
The use of fair values leads to the disclosure of 2.9 2.9 .984
commercially sensitive information.
Legend: 1 = Strongly Disagree; 2 = Disagree; 3 = Neutral; 4 = Agree; 5 = Strongly Agree
* Moderately Significant at the 0.10 level
** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

The bigger financial institutions (those with assets in excess of $1 billion) indicate

that they strongly believe that the volatility introduced by fair value accounting may

be misunderstood and their opinion proved statistically different from the views of

their smaller counterparts. Bigger financial institutions feel that reliable fair value

information is impossible to obtain (mean = 4.4) but the smaller firms are more

moderate (mean = 3.5). There is also a clear difference between the answers given

by chief financial officers of big and small financial institutions in relation to the

relevance and comparability of fair value accounting. The seven big firms regard

fair value accounting as providing irrelevant financial information and impede

comparability while the remaining respondents were neutral on both issues. Again,

H-10
further t-test shows big-4 banks’ views are more aligned with the JWGBA’s (i.e. the

more extreme views).

H2.2.2 Reliability of fair values

Contrary to JWGBA assertions that fair value information is highly subjective and

their reliability questionable, respondents are of the opinion that on average the fair

value of 51% to 75% of financial assets and liabilities could be reliably determined.

Furthermore, approximately a third of the respondents stated that over 90% could be

reliably determined. This result contradicts another reason put forward by the

JWGBA against fair value accounting.

The respondents are then asked to rate (on a scale of 1 to 5, 1 being highly unreliable,

3 neutral and 5 highly reliable) the reliability of the fair values of traded and non-

traded financial instruments. Responses show that the fair values of traded

financial instruments are generally reliable (mean = 3.6). However, it was

predicted from past literature that these fair values should be highly reliable as they

are traded. In relation to the non-traded financial instruments, a mean of 2.9

indicated that respondents are relatively neutral about the non-reliability of non-

traded financial instruments. This contradicts the past literature because according to

the JWGBA, fair values for non-traded financial instruments is difficult and

problematic to obtain and highly unreliable at best.

H3. PILOT STUDY CONCLUSIONS

The results of this study indicate that in general, Australian financial institutions are

using (and prefer to use) fair value accounting for the trading book but not for the

banking book. In addition, they prefer current practice signaling a reluctance to

H-11
depart from familiarity. There is general agreement that the banking and trading

books are somewhat different. The results also tend to support the view of the

JWGBA in that fair value accounting introduces unacceptable volatility in reported

earnings, is inappropriate for the banking book, is difficult and costly of obtaining

reliable fair values, impedes comparability and does not improve relevance. It seems

that views remain moderate since Fargher (2001) and have not substantially changed

despite the time lag and active international movements towards fair value

accounting for financial instruments.

The evidence points toward some significant contradictions from the official

JWGBA and JWGSS positions. The study results show that there are financial

institutions that currently employ fair value accounting for some of the financial

instruments in the banking book whereas the JWGBA (1999b) assert that fair value

accounting is not used for the banking book. They also argue that the trading and

banking books are fundamentally different but this is not the case as the responses

indicate weaker support for that logic. Another interesting finding is that (contrary to

JWGBA assertions) fair values of financial instruments in the banking book can be

reliably determined and are used by management. The respondents also believe that

information about the fair values of financial instruments is not commercially

sensitive, in contrast to the JWGBA position.

The results of this study further show that the views of the JWGBA tend to be much

more supported by the larger financial institutions more than the smaller ones (the

credit unions). This is perhaps not surprising given that the big-four Australian

banks dominate the JWGBA. Therefore, all financial institutions, particularly the

smaller ones, do not necessarily share the views expressed by the JWGBA.

H-12
In presenting their cases, both the JWGBA and the JWGSS used arguments based on

their assessment of what financial institutions want or believe. Their assertions are

generally not supported by empirical evidence and yet the conclusions they derived

regarding the fair value accounting model are partly based on these assumptions

concerning financial institutions. Therefore, the arguments of the JWGBA and the

JWGSS need to be reassessed in the light of the empirical evidence presented in this

paper and Fargher (2001). For example, the JWGBA argues that fair value

accounting will result in the disclosure of commercially sensitive information. The

results of this study do not support this argument.

Subsequently, the preparer survey to be sent out was shortened and unnecessary

questions deleted to improve the response rate. The difference between the two can

be seen by comparing Appendix F (six page survey) and Appendix A (three page

survey). First up, Question 5 of the pilot survey is shortened from 14 questions to six

crucial ones. Then, question 6 and 7 are combined into one question on fair value

accounting in its entirety without the column for “preferred method”. Questions 8, 9,

14 and 17 are deleted while question 10 shortened to just one question gauging the

perception of the distinction between the trading and banking book (Question 2 in the

new survey). The rest of the questions remain as is for the survey to be sent to

preparers and users.

H-13
 APPENDIX I:
TESTING FOR NON-RESPONSE BIAS

I1 INTRODUCTION

Lambert and Harrington (1990) state that non-response bias is the difference between

the answers or data from respondents and non-respondents to assess whether non-

respondents differ systematically from respondents. Possible non-response bias is a

persistent concern of researchers since the 1800s (Pearl and Farley, 1985; Tan,

Taplin, Hancock and Tower, 2003). There are a variety of methods to minimise this

problem as suggested by Dillman (1978) but there are also various ways to estimate

the extent of the possible bias such as detailed by Van Goor and Stuiver (1998).

A common method to assess potential non-response bias is the comparison of early

and late respondents (Ratneshwar and Stewart, 1989). Filion (1975) and Wiseman

and Macdonald (1979) find non-response bias to be a great concern for marketing

researchers while also acknowledging cost and time pressure cause the general

neglect of this issue. There is also a problem with a lack of information from

existing external resources (Ratneshwar and Stewart, 1989). Chapman (1988) labels

the making of the assumption that non-response bias does not exist as one of the

seven deadly sins of survey research.

The order the preparer and user surveys were received is included in the general

linear model for preference for fair value accounting to test for non-response bias.

Here, the first survey received from preparers (or users) from Australia and

Singapore will be labeled one. The remaining surveys are ordered in increments of

one as their measure of order.

I-1
I2 TESTING NON-RESPONSE BIAS IN PREPARER
SURVEYS

I2.1 On Preparer Preference for Fair Value Accounting

General linear models are fitted as per Section 5.9.3 but with an additional covariate

of order. Table I.1 depicts the first and final general linear model after backward

elimination. These results suggest no non-response bias.

Table I.1: Regression Results for Overall Preparer Preference for Fair Value
Accounting to Test for Non-response Bias

Initial Model
Variables in the equation Level of Significance
Order 0.507
Country 0.977
Type of Financial Institution 0.992
Size 0.957
Level of Experience 0.669
Country * Type of Financial Institution 0.034**
Country * Size 0.700
Country * Level of Experience 0.550
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.162
Support for Reliability of Traded Financial Instruments 0.416
Support for Reliability of Non-traded Financial Instruments 0.001***
Support for Comparability 0.060*
Volatility will not be Misunderstood 0.598
R-square = 0.754; Adjusted R-square = 0.689; p = 0.000
Final Model
Country 0.481
Type of Financial Institution 0.675
Country * Type of Financial Institution 0.012**
No Distinction between Banking and Trading Books 0.000***
Support for Reliability of Non-traded Financial Instruments 0.000***
Support for Comparability 0.000***
R-square = 0.737; Adjusted R-square = 0.704; p = 0.000

* Moderately significant at the 0.10 level

** Significant at the 0.05 level
*** Highly significant at the 0.01 level

I-2
I2.2 On Preparer Experience

The inclusion of the order variable in a regression testing for possible influence on

preparer experience reveals that for every 1-unit increase in order, experience

decreases by 0.02492 (p-value = 0.047, adjusted r-square = 0.036).

This suggests that more experienced preparers are more likely to respond quickly,

and hence, non-response bias may exist in that the survey is more likely to capture

experienced preparers.

I3 TESTING NON-RESPONSE BIAS IN USER SURVEYS

I3.1 On User Preference for Fair Value Accounting

Similarly, general linear models are fitted as per Section 6.9.1 but with an additional

covariate of order. Table I.2 depicts the first and final general linear model after

backward elimination.

Table I.2: Regression Results for User Preference for Fair Value Accounting
for all Financial Instruments to Test for Non-response Bias

Initial Model
Variables in the equation Level of Significance
Order 0.580
Country 0.289
Type of User 0.843
Level of Experience 0.154
Country * Type of User 0.934
Country * Level of Experience 0.376
No distinction between trading and banking book 0.008***
Support for Relevance 0.208
Support for Comparability 0.843
Support for Reliability of Traded Financial Instruments 0.837
Support for Reliability of Non-traded Financial Instruments 0.059*
Volatility will not be misunderstood 0.002***
R-square = 0.615; Adjusted R-square = 0.507; p = 0.000

I-3
Table I.2: Regression Results for User Preference for Fair Value Accounting
for all Financial Instruments to Test for Non-response Bias
(continued)

Final Model
Level of Experience 0.006***
No distinction between trading and banking book 0.001**
Support for Reliability of Non-traded Financial 0.017**
Instruments
Volatility will not be misunderstood 0.000***
R-square = 0.570; Adjusted R-square = 0.541; p = 0.000

* Moderately significant at the 0.10 level

** Significant at the 0.05 level
*** Highly Significant at the 0.01 level

Once again, no non-response bias is detected.

I3.2 On User Experience

This order variable is also tested for possible influence on the level of experience of

users. The regression results indicate that for every 1-unit increase in response,

experience decreases by 0.04918 (p-value = 0.003, adjusted r-square = 0.114).

Therefore, as with preparers, non-response bias may exist in that the survey is more

likely to capture experienced users.

I-4
 APPENDIX J:
TRADING VERSUS BANKING BOOK

J1 INTRODUCTION

In the preparer survey, four questions are asked to ascertain the level of preference

for fair value accounting from the perspective of financial institutions. As mentioned

in Section 5.10.1, there is general support for fair value accounting in trading book

but neutrality to its use for financial instruments in the banking book. Cross

tabulations shows that respondents tend to favour the marking of financial

instruments to fair value in the statement of financial position whether they are from

the banking or trading book (see Table J.1). There are instances (purple shading)

where respondents show support for fair value measurement in the trading book but

high opposition to it in the banking book and vice versa (green shading).

Table J.1: Cross Tabulation of Preparer Preference for Fair Valuing

Financial Instruments on the Balance Sheet

Do you support the concept of marking financial

instruments in the trading book to fair value on the Balance
Sheet?
Strongly Oppose Neutral Support Strongly
Oppose Support
Do you support Strongly
1 1 5
the concept of Oppose
marking
Oppose 5 1 12 4
financial
instruments in Neutral 3 4 5
the banking Support 2 18 7
book to fair
value on the Strongly 15
Balance Sheet? Support
Total 1 7 4 35 36

One of the strongest reasons put forward by the JWGBA is the reluctance to take the

changes in fair values to the Income Statement. Cross tabulations of preparer

J-1
support for the two aspects of fair value accounting for the banking and trading book

respectively are shown in Table J.2 and Table J.3. Respondents tend to either favour

fair value accounting, both the measurement of financial instruments at fair value

and the taking of fair value changes to the Income Statement, as a whole or not at all.

There are a few exceptions with preparers who oppose the taking of changes in fair

value to the Income Statement as per JWGBA assertions.

Table J.2: Cross Tabulation of Preparer Preference for Fair Value

Accounting in the Banking Book.

Do you support the concept of taking changes in fair values of

financial instruments in the banking book to the Income
Statement?
Strongly Oppose Neutral Support Strongly
Oppose Support
Do you Strongly
5 2
support the Oppose
concept of
Oppose 1 20 1
marking
financial Neutral 1 8 2 1
instruments in Support 1 2 1 23
the banking
book to fair Strongly
2 13
value on the Support
Balance Sheet?
Total 7 25 9 28 14

The tendency to support fair value accounting as a whole or not at all is also shown

in Table J.3. Here respondents who support fair value measurement for the banking

(trading) book also support the taking of the changes in fair value to the statement of

financial performance and vice versa. There are no outliers in terms of preference

for fair value accounting for the trading book. Although the JWGBA argue that fair

value accounting is highly relevant for the trading book, there are still a handful of

preparers who oppose its use.

J-2
Table J.3: Cross Tabulation of Preparer Preference for Fair Value
Accounting in the Trading Book.

Do you support the concept of taking changes in fair values

of financial instruments in the trading book to the Income
Statement?
Strongly Oppose Neutral Support Strongly
Oppose Support
Do you support the Strongly
Oppose 1
concept of marking
financial Oppose 7
instruments in the Neutral 1 2 1
trading book to fair Support 1 33 1
value on the
Strongly 1 2 33
Balance Sheet?
Support
Total 1 8 4 36 34

Respondents do not oppose the use of fair value accounting for the banking book

contrary to expectations. They also show general support for fair value accounting in

the trading book but the literature suggests that financial institutions strongly support

it for financial instruments in the trading book.

Regressions (via general linear models) are carried out to determine if any of the

independent variables are determinants of the preference for fair value accounting.

Although four questions are asked to glean the level of preference for fair value

accounting, the highly correlated questions are considered in tandem. Here,

responses to the two questions related to the banking book are averaged and resulted

in a measure of the preference for fair value accounting in the banking book. This is

repeated for the two questions on the trading book. Two separate set of regressions

are then conducted on these two different measures.

J-3
J2 PREFERENCE FOR FAIR VALUE ACCOUNTING IN
THE BANKING BOOK

All four independent variables of country (Australia versus Singapore), size (small,

medium, large), type of financial institution and level of experience are included in

each regression. In addition, the two-way interactions between country and the other

three variables are also included. Similar to Section 5.10.3.1 preparer responses for

the questions related to the distinction between banking and trading book, qualitative

characteristics and problems with volatility of reported earnings are also included as

covariates in the initial general linear model fitted. Following this initial model, the

least significant variable is removed for each subsequent in a backward elimination

model as a further check. Table J.4 shows the regression results for the first pair of

questions that relate to the use of fair value accounting in the banking book.

Table J.4: Regression Results for Preparer Preference for Fair Value
Accounting in the Banking Book
Initial Model
Variables in the equation Level of Significance
Country 0.968
Type of Financial Institution 0.825
Size 0.931
Level of Experience 0.869
Country * Type of Financial Institution 0.083*
Country * Size 0.440
Country * Level of Experience 0.430
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.121
Support for Reliability of Traded Financial Instruments 0.619
Support for Reliability of Non-traded Financial Instruments 0.001***
Support for Comparability 0.113
Volatility will not be Misunderstood 0.894
R-square = 0.689; Adjusted R-square = 0.619; p = 0.000

J-4
Table J.4: Regression Results for Preparer Preference for Fair Value
Accounting in the Banking Book (Continued)
Final Model
Variables in the equation Level of Significance
Country 0.901
Type of Financial Institution 0.598
Country * Type of Financial Institution 0.035**
No Distinction between Banking and Trading Books 0.000***
Support for Relevance 0.000***
Support for Reliability of Non-traded Financial Instruments 0.001***
R-square = 0.719; Adjusted R-square = 0.684; p = 0.000
* Moderately significant at the 0.10 level
** Significant at the 0.05 level
*** Highly significant at the 0.01 level

The regression models show that the interaction between country and type of

financial institution, perception on the banking and trading distinction, the perceived

relevance of fair value accounting and reliability of fair values for non-traded

financial instruments are statistically significant predictors.

J2.1.1 Country by Type of Financial Institution Effect

Figure J.1 shows country by type of financial institution effect..

Figure J.1: Australian and Singaporean Preparers’ Perception on Preference

for Fair Value Accounting in the Banking Book

5
Average Response

3 Australia

2 Singapore

0
Local Bank Foreign Bank Credit Union Other Financial
Institution
Type of Financial Institution

J-5
The relationship between foreign banks and their preference for fair value accounting

in the trading book in Australia is significantly different from the relationship

between foreign banks in Singapore and their respective preference.

Table J.5: Preparer Preference for Fair Value Accounting by Country by

Type of Financial Institution
Country n Type of Financial Institution Mean* Std. Error
Australia 5 Local Bank 3.21 .348
3 Foreign Bank 2.44 .428
23 Credit Union 3.17 .153
10 Other Financial Institution 3.24 .244
Singapore 11 Local Bank 3.10 .224
21 Foreign Bank 3.31 .160
10 Other Financial Institution 2.58 .232
Legend: Strongly Oppose = 1; Oppose = 2; Neutral = 3; Support = 4; Strongly Support = 5.
* Estimated marginal means when significant covariates are held constant at their mean
values: No distinction between the banking and trading book = 2.80, support for relevance
= 3.78 and reliability of fair values for non-traded financial instruments = 2.95.

Overall, the low mean scores indicate lack of support for fair value accounting in the

banking book. Similar to the findings of Section 5.10.3.1, foreign banks in Australia

had the lowest preference for fair value accounting for all financial instruments

whereas their counterparts in Singapore show the highest level of support.

Unfortunately, due to the anonymous nature of the surveys no further information is

available on these foreign banks to further examine possible.

J2.1.2 No Distinction between Banking and Trading Book Effect

It is contended that respondents who perceive that the trading and banking books are

not sufficiently different will tend to prefer fair value accounting. Statistically, a

regression coefficient of 0.531 confirms that for every 1-unit increase in the

perception that the trading and banking book are not different, the estimated marginal

J-6
means for preference for fair value accounting in the banking book increases by

0.531. In other words, preparers who do not perceive the two books to be different

support fair value accounting in the banking book more.

J2.1.3 Relevance of Fair Value Accounting

It is also professed that respondents who perceive that fair value accounting in

relevant will tend to support fair value accounting. This is confirmed by a regression

coefficient of 0.419 indicating that for every 1-unit increase in the perception on its

relevance, the estimated marginal means for preference for fair value accounting in

the banking book increases by 0.419.

J2.1.4 Reliable Non-traded Financial Instrument Fair Values Effect

It is also professed that respondents who perceive that the fair values of non-traded

financial instruments are reliable will tend to support fair value accounting. This is

confirmed by a regression coefficient of 0.314 indicating that for every 1-unit

increase in the perception on the reliability of those fair values, the estimated

marginal means for the overall preference for fair value accounting increases by

0.314. Thus, when preparers have faith in the reliability of the fair values of non-

traded financial instruments, support for fair value accounting increases.

J3 PREFERENCE FOR FAIR VALUE ACCOUNTING IN

THE TRADING BOOK

As per Section 2, the same 13 variables are included in the initial model and then a

backward elimination model conducted. Table J.5 shows the regression results.

J-7
Table J.6: Regression Results for Preparer Preference for Fair Value
Accounting in the Trading Book
Initial Model
Variables in the equation Level of Significance
Country 0.212
Type of Financial Institution 0.207
Size 0.602
Level of Experience 0.360
Country * Type of Financial Institution 0.407
Country * Size 0.251
Country * Level of Experience 0.214
No Distinction between Banking and Trading Books 0.052*
Support for Relevance 0.121
Support for Reliability of Traded Financial Instruments 0.056*
Support for Reliability of Non-traded Financial Instruments 0.837
Support for Comparability 0.081*
Volatility will not be Misunderstood 0.157
R-square = 0.534; Adjusted R-square = 0.421; p = 0.000
Final Model
Variables in the equation Level of Significance
No Distinction between Banking and Trading Books 0.070*
Support for Reliability of Traded Financial Instruments 0.000***
Support for Comparability 0.000***
R-square = 0.404; Adjusted R-square = 0.381; p = 0.000
* Moderately significant at the 0.10 level
** Significant at the 0.05 level
*** Highly significant at the 0.01 level

Preparers who perceive fair values for traded financial instruments are reliable and

fair value accounting result in comparable information show more support for fair

value accounting in the trading book (parameter estimates of 0.460 and 0.417

respectively). Although, the perception that the trading and banking books are

different was moderately significant, it was in the opposite direction to expectations.

The results show that for every 1-unit increase in the perception of no difference

between the two books, support for fair value accounting decreases by 0.15. There

are no other predictors according to the general linear model.

J-8
J3.1.1 No Distinction between Banking and Trading Book Effect

It is contended that respondents who perceive that the trading and banking books are

not sufficiently different will tend to prefer fair value accounting overall.

Surprisingly, a regression coefficient of -0.150 indicated that for every 1-unit

increase in the perception that the trading and banking book are not different, the

estimated marginal means for preference for fair value accounting in the trading

book decreases by 0.15. However, this variable is only moderately significant and

the regression coefficient also signals the small magnitude of its effect.

J3.1.2 Reliable Traded Financial Instrument Fair Values Effect

It is also professed that respondents who perceive that the fair values of non-traded

financial instruments are reliable will tend to support fair value accounting. As

expected, a regression coefficient of 0.460 indicates that for every 1-unit increase in

the perception on the reliability of fair values for traded financial instruments,

preference for fair value in the trading book increases by 0.460. Again, if preparers

perceive the fair value of traded financial instruments to be reliable, they tend to

support fair value accounting more.

J3.1.3 Support for Comparability Effect

Finally, it is proposed that respondents who perceive fair value accounting results in

comparable information will tend to prefer fair value accounting more than their

counterparts. The regression coefficient for this variable is 0.417; thus for every 1

unit increase in the support for comparability, the estimated mean increases by 0.417.

J-9
J4 CONCLUSION

There is a high level of agreement between the predictors for preparer perception as

measured by the minimum score of responses to the four questions on fair value

accounting and the minimum score for each pair of question related to the trading

and banking book respectively (see Table J.7).

Table J.7: Comparison of Significant Predictors of Preparer Preference for

Fair Value Accounting Overall, and in the Banking and Trading
Books
Predictor Overall Banking Trading
Preference Book Book
Country No No No
Type of Financial Institution No No No
Size No No No
Level of Experience No No No
Country * Type of Financial Yes Yes No
Institution
Country * Size No No No
Country * Level of Experience No No No
No distinction between trading and Yes Yes Yes
banking books
Relevance of fair value accounting No Yes No
Reliability of fair values for traded No No Yes
financial instruments
Reliability of fair values for non- Yes Yes No
traded financial instruments
Comparability of fair value Yes No Yes
accounting
Volatility will not be misunderstood No No No
by users

J-10