Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. A Study of Early and Late Adopters of International Financial Reporting Standards in New Zealand A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy in Accounting at Massey University, Albany, New Zealand Warwick Stent 2011 i ABSTRACT This study investigates accounting choice relating to the timing of adoption of International Financial Reporting Standards (IFRS) in New Zealand i.e., the choice to voluntarily early adopt IFRS or to defer adoption of IFRS until it became mandatory. Results for 40 early adopters are measured against those of a control group of 40 late adopters. The study includes an examination of the impact of IFRS on financial information (IFRS differences), as well as analyses of qualitative information obtained from discretionary narratives in annual reports, questionnaires and interviews. Significant IFRS differences are found for most financial statement elements and ratios for both early and late adopters. However, when IFRS differences for early adopters are compared to those for late adopters, the difference-in-differences are not found to be significant. Hence, IFRS differences result in incentives which may influence adoption timing, but these incentives are not significantly different for early and late adopters. Content analysis of discretionary narratives in annual reports reveals significant differences for all four of the measures used to assess the extent of disclosures, with early adopters providing typically twice as much disclosure as late adopters. Further analysis relating to the nature of disclosures reveals three major themes: ‘informing of importance’, ‘potency’ and ‘evaluative’. For the first two of these themes, significant differences are found and early adoption persists as a significant explanatory variable, after controlling for other incentives for voluntary disclosure, such as firm size, auditor and industry. ‘Evaluative’ disclosures are made by relatively few firms; are predominantly negative regarding IFRS adoption and no significant differences between early and late adopters are found for this theme. Disclosure findings reflect that early adopters attach a higher level of importance to IFRS adoption than late adopters. Survey data reveals significant differences for one of six measures of costs of IFRS as well as for a constructed ‘overall benefits’ score and three of nine individual benefits assessed. Also, three further themes emerge from content analysis of responses regarding motivations for adoption timing, namely ‘activity’, ‘manageability’ and ‘accounting choice’. Overall, the findings triangulate to suggest that ‘accounting choice’ has less explanatory power, with regard to adoption timing decisions, than ‘other factors’ which are unrelated to the impact of IFRS on accounting information. Examples of such ‘other factors’ include the level of importance which firms attach to IFRS adoption; evaluations of the consequences of IFRS (predominantly neutral or negative) and perceptions as to the manageability of IFRS adoption (unexpected factors influence deferral). Adoption timing decisions of both early and late adopters are found to be predominantly ‘dynamic’ (proactive) rather than ‘static’ (passive) activity. ii ACKNOWLEDGEMENTS I am pleased to have this opportunity to formally thank those who have assisted me along the path towards completion of this thesis. It has been a real privilege to work with my two supervisors, Professors Jill Hooks and Mike Bradbury. Their generous helpings of time, constructive criticism, support, ideas, knowledge and humour have made a huge difference to my progress and to my personal growth. I have been very fortunate to have the complementary skills of these two talented and positive people just down the passage from me. I am grateful too to Professor Fawzi Laswad for his support and encouragement over the years. I could not have asked for more from a head of department. It makes a difference too, to love going to work. The superb facilities, calibre of people and collegiality at the School of Accountancy and Massey University make it a very stimulating environment to work in. The long road to a PhD has much less uphill in a place like this and I am very grateful for that. Particular thanks go to Donna Perreau, Muralikrishna Prasad and Siata Tavite for their assistance with data capture and to Professor Asheq Rahman for helpful comments. I would also like to express my gratitude to: • the Accounting & Finance Association of Australia and New Zealand, for the research grant awarded in 2008, and • the New Zealand Institute of Chartered Accountants, for the scholarships and travel grant awarded in 2008 and 2010. Funds awarded contributed significantly to the progress and quality of this research. Finally, my thanks too to my good friend Rob Jackson for the opportunities, guidance and inspiration that set me on this path; to my wife Gail for all that she has done to support me over the years; and to my children Tyler and Hayley, as well as my sisters Carol and Alison and my mother and father – this achievement would be empty if I could not share it with you. Approval for the research was obtained from the Massey University Human Ethics Committee: Northern, Application MUHECN 07/008. Any concerns relating to the conduct of the research may be directed to Associate Professor Ann Dupuis, Acting Chair, Massey University Human Ethics Committee: Northern, telephone 09 414 0800 x9054, email humanethicsnorth@massey.ac.nz. iii TABLE OF CONTENTS ABSTRACT …………………………………………………………………………………....…..i ACKNOWLEDGEMENTS .…………………………………………………………………......ii LIST OF TABLES ..…………………………………………………………………..................vi LIST OF FIGURES …...…………………………………………………………………...........vii LIST OF APPENDICES ….…...……………………………………………………………......viii LIST OF ABBREVIATIONS .…………………………………………………………………...ix CHAPTER 1:  INTRODUCTION ........................................................................................... 1  1.1  STUDY OVERVIEW .............................................................................................................................. 1  1.2  RESEARCH QUESTIONS ..................................................................................................................... 3  1.3  MOTIVATION FOR THE STUDY AND CONTRIBUTIONS .......................................................... 4  1.4  MAIN FINDINGS ................................................................................................................................... 6  1.4.1  IMPACT OF IFRS ADOPTION ON FINANCIAL INFORMATION ................................................. 6  1.4.2  IFRS-RELATED DISCRETIONARY NARRATIVE DISCLOSURES ........................................... 7  1.4.3  COSTS AND BENEFITS ARISING FROM IFRS ADOPTION AS WELL AS OTHER SURVEY INSIGHTS REGARDING ADOPTION TIMING ............................................................................................. 8  1.5  FRAMEWORK OF THESIS ................................................................................................................. 9  1.6  PUBLICATIONS ..................................................................................................................................... 9  CHAPTER 2:  BACKGROUND TO IFRS ADOPTION ....................................................... 10  2.1  INSTITUTIONAL BACKGROUND ................................................................................................... 10  2.2  MOTIVATION FOR ADOPTING IFRS IN NEW ZEALAND ........................................................ 12  2.3  CHAPTER SUMMARY ....................................................................................................................... 17  CHAPTER 3:  THEORY AND LITERATURE ..................................................................... 18  3.1  INTRODUCTION ................................................................................................................................. 18  3.2  ACCOUNTING CHOICE .................................................................................................................... 19  3.2.1  OVERVIEW ....................................................................................................................................... 19  iv 3.2.2  STUDIES RELATING TO ADOPTION TIMING DECISIONS ..................................................... 22  3.3  STUDIES RELATING TO THE IMPACTS OF IFRS ...................................................................... 27  3.3.1  INTRODUCTORY COMMENTS AND EARLY STUDIES.......................................................... 27  3.3.2  RECENT STUDIES RELATING TO CONSEQUENCES OF IFRS ............................................. 28  3.3.3  RECENT STUDIES RELATING TO IFRS DIFFERENCES ......................................................... 32  3.4  CHAPTER SUMMARY ....................................................................................................................... 35  CHAPTER 4:  RESEARCH DESIGN, DATA AND SAMPLE SELECTION ........................ 36  4.1  OVERVIEW OF RESEARCH METHODOLOGY ........................................................................... 36  4.2  IMPACT OF IFRS ADOPTION ON FINANCIAL INFORMATION ............................................. 41  4.2.1  RESEARCH QUESTION AND HYPOTHESES ............................................................................... 41  4.2.2  METHOD............................................................................................................................................ 42  4.2.3  DATA ................................................................................................................................................. 43  4.3  IFRS-RELATED DISCRETIONARY NARRATIVE DISCLOSURES ....................................... 45  4.3.1  RESEARCH QUESTION AND HYPOTHESES ............................................................................... 45  4.3.2  METHOD AND DATA ...................................................................................................................... 46  4.4  COSTS AND BENEFITS OF IFRS ADOPTION ............................................................................ 53  4.4.1  RESEARCH QUESTIONS AND HYPOTHESES ............................................................................. 53  4.4.2  METHOD AND DATA ...................................................................................................................... 54  4.5  SAMPLE SELECTION ........................................................................................................................ 58  4.6  CHAPTER SUMMARY ....................................................................................................................... 60  CHAPTER 5:  RESULTS .................................................................................................... 61  5.1  INTRODUCTION ................................................................................................................................. 61  5.2  IMPACT OF IFRS ADOPTION ON FINANCIAL INFORMATION ............................................. 62  5.2.1  INTRODUCTION............................................................................................................................... 62  5.2.2  IMPACT OF IFRS ON FINANCIAL STATEMENT ELEMENTS ................................................... 62  5.2.3  IMPACT OF IFRS ANALYSED BY ACCOUNTING STANDARDS ............................................. 67  5.2.4  EFFECT OF IFRS ON KEY RATIOS................................................................................................ 75  5.2.5  SUMMARY ........................................................................................................................................ 79  5.3  IFRS-RELATED DISCRETIONARY NARRATIVE DISCLOSURES ........................................ 80  v 5.3.1  INTRODUCTION............................................................................................................................... 80  5.3.2  FORM ORIENTED ANALYSIS ........................................................................................................ 80  5.3.3  MEANING ORIENTED ANALYSIS ................................................................................................ 82  5.3.4  SUMMARY ........................................................................................................................................ 89  5.4  COSTS AND BENEFITS ARISING FROM IFRS ADOPTION AS WELL AS OTHER SURVEY INSIGHTS ........................................................................................................................................................... 91  5.4.1  INTRODUCTION............................................................................................................................... 91  5.4.2  COSTS OF IFRS ADOPTION ........................................................................................................... 92  5.4.3  BENEFITS OF IFRS ADOPTION ..................................................................................................... 98  5.4.4  ADDITIONAL INSIGHTS AS TO POTENTIAL ADOPTION TIMING MOTIVATIONS ..... 103  5.5  CHAPTER SUMMARY ..................................................................................................................... 106  CHAPTER 6:  CONCLUSIONS ........................................................................................ 109  6.1  SUMMARY AND IMPLICATIONS OF FINDINGS ...................................................................... 109  6.2  REFLECTIONS ON RESULTS AS LINKED TO THE LITERATURE ...................................... 111  6.3  LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH ........................................... 114  6.3.1  Limitations ........................................................................................................................................ 114  6.3.2  Suggestions for further research ........................................................................................................ 115  REFERENCES ................................................................................................................................................. 117  vi LIST OF TABLES Page Table 1 Framework for content analysis 48 Table 2 Effect of sample selection criteria 59 Table 3 Descriptive Statistics for sample under OLD GAAP ($000) 63 Table 4 Impact of IFRS on financial statement elements 64-65 Table 5 Impact of IFRS by financial statement element and accounting standard 68-69 Table 6 Significance of difference in changes arising from individual accounting standards: Early adopter change compared to late adopter change 70 Table 7 Descriptive statistics for effect of IFRS on key ratios 76-77 Table 8 Form oriented analysis (count data per firm) 81 Table 9 Meaning oriented analysis (themes) 83 Table 10 Mann-Whitney test statistics relating to themes 85 Table 11 Ordered logistic regression on disclosures score 88 Table 12 Descriptive statistics of sample relating to cost of IFRS implementation 93-94 Table 13 Descriptive statistics of sample relating to benefits of IFRS implementation 99 Table 14 Thematic analysis of interview responses regarding adoption timing motivations 104 vii LIST OF FIGURES Page Figure 1 Sequential explanatory strategy used for mixed methods research approach 2; 39 viii LIST OF APPENDICES Page App. 1 Example of most informative type of IFRS reconciliation 122 App. 2 Example of less informative type of IFRS reconciliation 124 App. 3 Themes used for coding of IFRS related extracts from discretionary narrative disclosures (descriptions and examples) 125 App. 4 Sample questionnaire 128 App. 5 Sample covering letter 135 App. 6 Adapted questionnaire with interview prompts 136 ix LIST OF ABBREVIATIONS ASRB - Accounting Standards Review Board EA - Early Adopters of IFRS EU - European Union FMA - Financial Markets Authority FRSB - Financial Reporting Standards Board GAAP - Generally Accepted Accounting Practice GPFR - General Purpose Financial Reports IAS - International Accounting Standards IFRS - International Financial Reporting Standards IFRS for SMEs - IFRS for Small to Medium Enterprises IOSCO - International Organisation of Securities Commissions IPSAS - International Public Sector Accounting Standards LA - Late Adopters of IFRS MED - Ministry of Economic Development NZ - New Zealand UK - United Kingdom US - United States XRB - External Reporting Board 1 CHAPTER 1: INTRODUCTION 1.1 STUDY OVERVIEW This study draws on accounting choice literature to investigate one of the most significant events in world accounting history, namely the adoption of International Financial Reporting Standards (IFRS). The New Zealand setting provides a unique opportunity to study timing choice. It became mandatory for listed companies in New Zealand to adopt IFRS for periods beginning on or after 1 January 2007, but unlike the European Union (EU), Australia and many other countries, New Zealand companies were allowed the option of early adopting IFRS for periods beginning from 1 January 2005. Mixed methods research (quantitative and qualitative) is employed to conduct an in- depth investigation of the adoption timing choices of listed companies in New Zealand. Results for 40 early adopters are measured against those of a control group of 40 late adopters (i.e. listed companies that elected to delay adoption of IFRS until it became mandatory to do so as from 1 January 2007). This investigation includes an examination of the impact of IFRS on financial information, as well as analyses of qualitative information obtained from discretionary narratives in annual reports, questionnaires and interviews. Figure 1 (over page) provides a diagrammatic overview of the study. This diagram is explained in more detail in Chapter 4, 2 Figure 1 Sequential explanatory strategy used for mixed methods research approach       Theoretical framework of ‘Accounting Choice’ (Ch. 3)  Quantitative: • Capture OLD GAAP & IFRS values (e.g. assets, expenses etc.)  • Capture values($) and source (standard) of IFRS adjustments  • Compute ratios using OLD GAAP & IFRS values (e.g. leverage)  Motivations for IFRS adoption timing:  Early Adopters v Late Adopters? 1. COLLECT DATA FROM FINANCIAL STATEMENTS Qualitative:  Extract narrative descriptions for IFRS adjustments 2. ANALYSE DATA Quantitative:  • Descriptive statistics for IFRS differences  • Significance testing of OLD GAAP/IFRS and EA/LA differences  • Most frequent causes of IFRS differences (standards)  Qualitative: Summary of narratives for most frequent of IFRS adjustments  3. INTERPRET INTERPRET RESULTS 4. COLLECT DATA FROM ANNUAL REPORTS Quantitative:  • Count number words used in IFRS related disclosures  • Measure space devoted to IFRS related disclosures    5. ANALYSE DATA Quantitative: • Form oriented content analysis (count data)  • Descriptive statistics (word count, space usage, theme score)  • Significance of EA/LA differences  • Multivariate regression on importance of IFRS for EA  Entire analysis of financial  statements data:   Impact of IFRS on financial  information (EA v LA)?  6. INTERPRET Entire content analysis of annual  report discretionary narratives:  Attitude to IFRS adoption          (EA v LA)?  7. CONDUCT SURVEYS (QUESTIONNAIRES & INTERVIEWS)  Quantitative: Obtain values relating to costs & benefits of IFRS (questionnaires)  8. ANALYSE DATA Quantitative:  • Descriptive statistics on IFRS costs and benefits  • Significance of EA/LA differences  Qualitative: Meaning oriented content analysis (themes)  Qualitative: Summary of adoption timing insights  Qualitative: Extract/prepare discretionary narrative disclosures (outside of  financial statements) related to adoption of IFRS   Qualitative: Obtain adoption timing insights (semi‐structured interviews)  9. INTERPRET Entire analysis of survey data:  Costs/benefits of IFRS (EA v LA)?  Other adoption timing insights?  CONCLUSION (Ch. 6) Ch. 4.2 Ch. 5.2 Ch. 4.3 Ch. 5.3 Ch. 4.4 Ch. 5.4 3 1.2 RESEARCH QUESTIONS The aim of this thesis is to investigate what motivates entities to adopt IFRS early or late. To seek a better understanding of this research problem, three research questions are formed to explore potential motivating factors: 1. What impact does IFRS adoption have on the financial information of early as opposed to late adopters? 2. What do discretionary narrative disclosures in annual reports reveal about early and late adopter firms’ attitudes to IFRS adoption? 3. What are the costs and perceived benefits of IFRS adoption for early as opposed to late adopters? In addressing each of these research questions, there are two main sub-questions which are investigated. The first sub-question is quantitative, while the second is qualitative in nature: a) Are there significant differences between early and late adopters? b) What is the nature of significant differences (if any) that are identified? All research procedures are therefore designed to clearly distinguish results of early adopters from those of late adopters. The first research question is addressed by investigating the impact of IFRS adoption on financial statements and some common financial ratios. The impact on financial statements is analysed by financial statement components (e.g., assets, liabilities) as well as by the specific accounting standards concerned (e.g., financial instruments, income taxes). In addition, five key ratios that may influence analysts are selected for analysis. These ratios reflect the main ratios in the Du Pont analysis. Quantitative analysis provides descriptive statistics which report on the nature, frequency, extent and significance of IFRS differences. Qualitative analysis provides further detail regarding the nature of the most frequent IFRS adjustments and the potential for these adjustments to provide incentives which may influence adoption timing decisions. The second research question examines IFRS-related discretionary narrative disclosures which are outside of the financial statements. Content analysis is applied to investigate the extent (quantitative) and nature (qualitative) of such disclosures in annual reports for the year in which IFRS is first adopted. The purpose of these procedures is to 4 provide evidence regarding the main themes communicated by those responsible for management and governance, as well as the importance that they attach to IFRS adoption. These themes may provide further insights regarding influences for IFRS adoption timing decisions. Quantitative evidence provided includes descriptive statistics related to word counts and space devoted to IFRS related disclosures. Qualitative evidence provided relates to the identification and explanation of the main themes communicated by management. These measures as well as derived scores for the IFRS themes identified are then tested for significant differences between EA and LA. Multivariate analysis is also conducted to control for firm specific factors, identified in prior literature, which may influence disclosures (e.g., industry, firm size and auditor influence). The third research question considers evidence relating to the costs and benefits of IFRS adoption. Much of this evidence is not found in publicly available documents, so it is obtained by means of questionnaires and semi-structured interviews. Information gathered through the questionnaires includes quantitative measures of the cost and level of effort involved in implementing IFRS, as well as opinions of senior financial executives regarding expected benefits of adopting IFRS. Quantitative analysis provides descriptive statistics related to these measures and the significance of differences between EA and LA for these measures. This is supplemented with qualitative evidence gathered through comments in questionnaires and through the semi-structured interviews, which may provide further insights regarding IFRS adoption timing choices. 1.3 MOTIVATION FOR THE STUDY AND CONTRIBUTIONS This study is motivated by two key opportunities to contribute to prior literature. First, IFRS adoption is a recent and ongoing global phenomenon and there is growing recognition in the literature of the importance of such research. Second, accounting choices are traditionally considered in relation to ad hoc changes of individual accounting standards. The two year window for voluntary adoption of IFRS in New Zealand prior to 1 January 2007, presents a unique opportunity to examine the motivations of early adopters in the face of multiple/sweeping changes in accounting standards. 5 Daske, Hail, Leuz, & Verdi (2008) comment on the adoption of IFRS by over 100 countries as being one of the most significant changes in world accounting history.1 They also note that empirical evidence on the consequences of mandatory IFRS at that time was in its infancy and emphasise the need for further evidence. While a number of IFRS studies have emerged since then, research on the consequences of IFRS is still at an early stage. This study extends several contributions to the literature regarding the consequences of mandatory IFRS, which are noted in the Stent, Bradbury & Hooks (2010) study. First, it complements and extends early analyses of the financial statement effects of adopting IFRS in New Zealand (e.g. Kabir, Laswad, & Islam, 2010; Stent et al., 2010). Similar to these studies and a Hung & Subramanyam (2007) study of German firms, detailed financial statement effects of adopting IFRS are reported, but with an emphasis on distinguishing between effects for early and late adopters. This contribution may be useful to regulators and policy makers who are currently reviewing the financial reporting framework in New Zealand. Second, Stent et al. (2010) note, in a study related to this thesis, the importance of considering impacts of IFRS on key financial ratios as well as financial statement elements. The study points out that the impact of IFRS on ratios may impact analysts’ decisions regarding share value (value relevance) and credit worthiness, as well as some contracting decisions by firms that incorporate financial ratios (e.g., debt covenants, compensation contracts). This thesis extends the Stent et al (2010) study, which included only 16 early adopters as part of a representative sample of listed New Zealand companies. The expanded sample of 40 early adopters for this thesis allows for a more comprehensive assessment of the impact of IFRS on early adopters for both financial statements and some common financial ratios. Third, most studies examine the adoption of IFRS where previous GAAP was known to differ significantly from the international standards, whereas old NZ GAAP is perceived to be relatively similar to the international standards. Contrary to expectations, the analysis of effects in such a setting (the UK) indicates that IFRS information still conveys new information (Christensen, Lee, & Walker, 2009), although Goodwin, Ahmed & Heaney (2008) find otherwise for Australia. As noted in the Stent et al (2010) study, understanding the impact of IFRS is an important first step in seeking a deeper understanding of the motivations of early and late adopters of IFRS in New Zealand. 1 International Accounting Standards (IAS) were significantly developed and renamed after 2001 to become IFRS. For convenience, references to IFRS will include both IAS and IFRS. When the context is more specific IAS will be used. 6 Narrative disclosures are considered to be an important part of ‘quality’ corporate reporting as they offer useful insights and explanations concerning quantitative measures (Beattie, McInnes, & Fearnley, 2004; Beretta & Bozzolan, 2004; Clatworthy & Jones, 2003; Smith & Taffler, 2000). In particular, Stanton & Stanton (2002) identify that the ‘front half’ of the corporate annual report provides a tool for management to signal its response to matters of importance, hence an expectation arises that this should include management signals regarding IFRS. The second phase of this study therefore examines the nature and extent of the voluntary narrative disclosures made by governing bodies of corporate entities regarding the importance and effects of switching to IFRS. This results in the following contributions to the literature. First, although annual reports have been found to be the most important document for an entity conveying information to the public, no prior research has been identified which considers the importance and effects of IFRS or early adoption accounting choices, by extending analyses beyond the financial statements to the other sections of the annual report. Second, much of the literature to date on accounting choices relating to the motivations for, and effects of, adopting IFRS, is restricted to quantitative analysis of financial statements and capital markets data. This study provides depth, richness and a broader perspective in its evidence, as well as the opportunity to triangulate the findings with those from more traditional sources and perspectives. The depth and detail provided in this study is further enhanced by evidence gathered using questionnaires and interviews. This evidence provides unique insight into the extent of effort and cost required to implement IFRS, in relation to the benefits which arise as a result. Insights regarding choice of adoption timing are also collected. 1.4 MAIN FINDINGS 1.4.1 IMPACT OF IFRS ADOPTION ON FINANCIAL INFORMATION The results relating to the impact of IFRS adoption on financial information provide evidence of significant changes to most financial statement elements and ratios for both early and late adopters. These changes may provide incentives for adoption timing decisions. However, no evidence is found of statistically significant differences when 7 the changes for early adopters are compared to those for late adopters. For most firms a specific standard of IFRS has no impact, but a specific standard can be very material for a small number of firms. The most frequent sources of IFRS adjustments are those relating to income taxes, financial instruments and business combinations. These adjustments affect at least 50 percent of early and/or late adopter sample firms. In general IFRS tends to increase values of financial statement elements and these effects are more widespread for early than for late adopters. These findings contribute to the literature by clarifying that adoption timing choices do not appear to be influenced by incentives alone, but rather by the interaction between incentives and other factors, such as firm characteristics. 1.4.2 IFRS-RELATED DISCRETIONARY NARRATIVE DISCLOSURES Significant differences are found for all four measures of the extent of discretionary narrative disclosures related to IFRS adoption (e.g., the number of words and the amount of space devoted to these disclosures). The extent of IFRS adoption disclosures for early adopters is typically more than double that of late adopters. There is, however, relatively little narrative disclosure overall (e.g., none at all for fourteen late adopters and seven early adopters and generally less than a quarter of a page of the annual report for remaining firms). Analysis of the nature of disclosures identifies three major themes: ‘importance’, ‘potency’ and ‘evaluative’. ‘Informing of importance’ of IFRS adoption is the most dominant sub-theme. Differences between early and late adopters as regards the ‘Importance’ and ‘Potency’ themes are statistically significant. Multivariate analysis provides evidence that early adoption persists as a significant explanatory variable, after controlling for other incentives for voluntary disclosure, such as firm size, auditor and industry. ‘Evaluative’ narrative disclosures are made by relatively few firms and are found to be predominantly negative regarding IFRS adoption. There is no statistically significant difference between early and late adopters for this theme. Qualitative information such as theme descriptions and examples of disclosures are provided to afford deeper insight into the nature of the significant differences between early and late adopters. 8 Findings from the content analysis of discretionary narrative disclosures therefore reflect that early adopters attach a higher level of importance to IFRS adoption than late adopters and that firm attitudes in this regard influence the priority given to adoption timing. 1.4.3 COSTS AND BENEFITS ARISING FROM IFRS ADOPTION AS WELL AS OTHER SURVEY INSIGHTS REGARDING ADOPTION TIMING A significant difference between early and late adopters is found for only one of six measures of costs of IFRS adoption, but for three of nine expected benefits as well as for an ‘overall benefits’ score. Qualitative data in questionnaires and from interviews suggests that late adopters incur lower costs, because they have the advantage of being able to leverage off the experience of early adopters. Late adopters’ opinions regarding expected benefits of IFRS are generally more negative than those of early adopters. Qualitative data in this regard indicates that late adopters have concerns that IFRS is a detrimental change, while early adopters are more inclined to the view that there is disappointingly little change as a result of IFRS, in view of the level of cost and effort required. The survey also reveals additional insights regarding motivations for adoption timing, which emerge as three main themes, namely ‘activity’, ‘manageability’ and ‘accounting choice’. Little difference is found between early and late adopters regarding the ‘activity’ theme with approximately twice as many responses indicating ‘dynamic’ as opposed to ‘static’ approaches to IFRS adoption by firms. This provides important evidence to refute the possible misconception that early adopters are ‘dynamic’ and late adopters are ‘static’. Responses relating to the ‘manageability’ theme suggest that, while most early adopters were confident about what to expect from IFRS, late adopters had concerns regarding the ‘unexpected’ (e.g., lack of a stable platform of IFRS standards). ‘Accounting choice’ responses indicate that ‘other factors’ are twice as likely to influence adoption timing than the impact of IFRS on accounting information. These survey findings therefore suggest that significant differences, which exist 9 between early and late adopters as regards the costs and perceived benefits of IFRS adoption, may affect adoption timing decisions. Furthermore, the decision to late adopt IFRS is more likely to be the result of a dynamic evaluation than a static deferral of action, including concerns regarding unexpected matters which may affect the manageability of adoption. Finally, ‘accounting choice’ is found to have less explanatory power with regard to adoption timing decisions than ‘other factors’ which are unrelated to the impact of IFRS on accounting information. 1.5 FRAMEWORK OF THESIS The rest of the thesis proceeds as follows. Chapter 2 provides background information on the New Zealand adoption of IFRS. Chapter 3 reviews the literature relevant to accounting choice and IFRS adoption. Chapter 4 describes the research design for the study, including development of research questions, research methods, data and sample selection. Chapter 5 describes and discusses the results. Chapter 6 summarises and concludes the study as well as considering its limitations and suggestions for future research. 1.6 PUBLICATIONS One paper has been published which draws on data used for this thesis: Stent, W. J., Bradbury, M., & Hooks, J. (2010). IFRS in New Zealand: Effects on Financial Statements and Ratios. Pacific Accounting Review 22(2), 92-107. A second paper has been presented and published in terms of conference proceedings at the Sixth Asia Pacific Interdisciplinary Research in Accounting (APIRA) 2010 conference in Sydney, Australia: Stent, W. J., Hooks, J., & Bradbury, M. (2010). New Zealand’s switch to IFRS: Beyond the Financial Statements - a Qualitative Analysis of Annual Reports. http://apira2010.econ.usyd.edu.au A revised version of this paper is currently under review for publication in the British Accounting Review. 10 CHAPTER 2: BACKGROUND TO IFRS ADOPTION 2.1 INSTITUTIONAL BACKGROUND The influence of international accounting standards in New Zealand has been evident since 1974 (e.g., SSAP 1 Disclosure of Accounting Policies included the International Accounting Standards Committee crest), although these standards were modified (Bradbury, 1998). The Accounting Standards Review Board (ASRB) was the statutory body in New Zealand that had legal authority to review and approve financial reporting standards. The Financial Reporting Standards Board (FRSB), a board of the New Zealand Institute of Chartered Accountants (NZICA), was the body responsible for the development and consultation processes relating to new standards submitted to the ASRB. In principle, proposed standards could be submitted to the ASRB by entities other than NZICA, but the FRSB was traditionally the only entity to do so (Bradbury & van Zijl, 2006). In 1997, the FRSB decided to base new accounting standards on International or Australian accounting standards. However, these standards too were modified to ensure sector neutrality and consistency with other New Zealand pronouncements (Bradbury & van Zijl, 2006). These arrangements effectively resulted in New Zealand having its own version of generally accepted accounting practice, or ‘NZ GAAP’. On 19 December 2002, the ASRB announced that adoption of IFRS was to be mandatory for reporting entities in New Zealand for periods beginning on or after 1 January 2007. This announcement followed an earlier proposal on 21 October 2002, that listed issuers in New Zealand should adopt IFRS. The ASRB also allowed the option of early adoption for periods beginning on or after 1 January 2005, a point of distinction in relation to the European Union, Australia and many other countries which opted simply for mandatory adoption from 2005 (Bradbury & van Zijl, 2006). Since the ASRB’s announcements in 2002, there has been much commentary by various authors, professional bodies, accounting firms and commercial entities in New Zealand about the impact that IFRS is expected to have (e.g. Deloitte, 2003; Dunstan, 2002; Ernst & Young, 2004; Hickey, Spencer, van Zijl, & Perry, 2003; KPMG, 2006; NZICA, 11 2006; PriceWaterhouseCoopers, 2006; Texeira & Pickens, 2004). It is generally accepted that adopting IFRS could have significant and wide ranging implications for the adopting entities themselves as well as the financial markets and the countries within which they operate. In response to growing concerns as to the appropriateness of IFRS for smaller firms, the ASRB announced on 12 September 2007, that it had decided to delay mandatory adoption of IFRS for small companies that met specified criteria, pending a government review of financial reporting requirements in New Zealand. The government review culminated in the release of two discussion documents in September 2009, namely ‘The Statutory Framework for Financial Reporting’ (Ministry of Economic Development, 2009) and ‘Proposed Application of Accounting and Assurance Standards under the Proposed New Statutory Framework for Financial Reporting’ (Accounting Standards Review Board, 2009). The proposals put forward in these discussion documents began to take effect during 2011 with the establishment of a number of new statutory bodies including the Financial Markets Authority (FMA) and External Reporting Board (XRB). These discussion documents are not yet finalised at the time of writing this thesis but are expected to bring about significant changes2. Prior studies have identified that IFRS effects vary considerably with the strength of regulatory enforcement, firm governance, and reporting incentives (e.g., Barth, Landsman, & Lang, 2008; Byard, Li, & Yu, 2011; Daske et al., 2008; Lee, Walker, & Christensen, 2008). Kabir et al., (2010) describe New Zealand as a common law, Anglo- Saxon accounting country with an investor-oriented approach to standard setting similar to IFRS. They provide support from prior literature for claims that it has one of the strongest investor protection regimes and that it heavily enforces accounting standards (Kabir et al., 2010). Dunstan (2002) makes similar observations, noting that legislation (e.g. the Financial Reporting Act) and the interaction between the ASRB, FRSB and New Zealand Stock Exchange provide a well-developed financial reporting framework that compares favourably with international best practice. She also notes that rankings of 8th out of 49 countries for the quality of its accounting standards in a study by la Porta, Lopez-de-Silanes, Shleifer, & Vishny (1998) and 9th out of 81 countries in the 2 These changes include replacement of the ASRB by the XRB; and proposals that most small and medium-sized companies would no longer be required by statute to prepare General Purpose Financial Reports (GPFR). Reporting requirements for remaining entities are likely to specify IFRS or variations thereof, such as International Public Sector Accounting Standards (IPSAS) or IFRS for Small to Medium Enterprises (IFRS for SMEs). 12 Milken Capital Access Index provide independent support for the strength of New Zealand’s institutional arrangements. The framework described above is further strengthened by the activities and status of the New Zealand Securities Commission.3 Amongst their activities is the ‘Financial Reporting Surveillance Programme’, which comprises regular cyclical reviews to encourage high quality financial reporting. A number of these reviews placed particular emphasis on compliance with IFRS requirements. The Securities Commission was a member of the International Organisation of Securities Commissions (IOSCO) and was lead until recently by Jane Diplock, who also chaired IOSCO. Such factors are recognition of the status and international standing that the FMA inherits from the Securities Commission. 2.2 MOTIVATION FOR ADOPTING IFRS IN NEW ZEALAND The adoption of IFRS in Australia and New Zealand came with some surprise (Bradbury & van Zijl, 2006). On 3 July 2002, the Australian Financial Reporting Council (FRC) issued a directive to the Australian Accounting Standards Board (AASB) for adoption of IFRS in Australia from 2005, which was unexpected by the AASB as well as by New Zealand’s FRSB and ASRB. It is this directive and political motives for fostering close ties with Australia, which provided the catalyst for New Zealand’s decision to adopt IFRS (Bradbury & van Zijl, 2006). Hickey et al (2003) claim that the ASRB and FRSB conducted extensive consultations and that these consultations showed strong support for adopting IFRS. They go on to provide certain high level insights as to where this support was derived from including: • the trend internationally towards adoption of IFRS; • the decision taken in Australia on 3 July 2002 to adopt IFRS; • the risk to the credibility of financial reporting in New Zealand if these factors were ignored; • New Zealand’s commitment, over a number of years prior to the ASRB announcement, to work towards harmonisation with international standards in any case. 3 The Securities Commission was replaced by the Financial Markets Authority (FMA) in May 2011 (extracted with reference to http://www.nbr.co.nz/article/financial-markets-authority-established-nk- 91998 on 17 September 2011) 13 Bradbury & van Zijl (2006) note that a survey by PriceWaterhouseCoopers in 2003 corroborated the level of support for the adoption of IFRS (88% in favour). A slide used at a Deloitte presentation (Financial Reporting Update in October 2003) provides an apt summary of sentiment at the time: “Why adopt IFRS? – Everybody’s doing it!” It is acknowledged that not all parties shared this sentiment, notably, the United States. The United States has however, been committed for some years to advancing a converged set of standards as a joint project between the International Accounting Standards Board (IASB) and the Financial Accounting Standards Board (FASB)4. Texeira & Pickens (2004) provide further insights as to perceived motivations for adopting IFRS, explaining that international standards provide a common basis for measuring cash flows, assets and liabilities and that the more entities that prepare their financial information in the same way, the easier it should be for investors to compare investment options. They argue that both investors and entrepreneurs benefit as competition for global funds as well as investment opportunities increase, which in turn leads to lower agency costs and stronger market disciplines on entities to perform. Other benefits claimed by these authors include clearer identification of risk to guide resource allocation, reduced costs of accounting and financial infrastructure for capital markets and increased transferability of accounting personnel between countries. The consequences of a decision to adopt IFRS in New Zealand are considered in a briefing paper (Dunstan, 2002) prepared on behalf of the New Zealand Securities Commission, one of the parties with whom the FRSB and ASRB consulted on this matter. The briefing paper considers prior research relating to the impact of IFRS and, after presenting a cost/benefit analysis, concludes that adoption of IFRS will result in long term net benefits for New Zealand. This analysis and conclusion are particularly pertinent to this study and are therefore considered in more detail in the paragraphs below5. Expected Benefits of adopting IFRS Dunstan (2002) notes that a number of potential benefits for New Zealand had already 4 Extracted with reference to the IASB’s Project Agenda and Timetable at http://www.iasplus.com/agenda/agenda.htm on 22 June 2006 5 The third phase of this study uses questionnaires and semi-structured interviews to gather evidence relating to the cost and level of effort involved in implementing IFRS, as well as opinions of senior financial executives regarding expected benefits of adopting IFRS. This phase may therefore assist in accumulating early evidence towards assessment of long-term net cost/benefit for New Zealand of adopting IFRS. 14 been identified by the FRSB in ED 92 when making their earlier commitment to convergence/harmonisation of financial reporting standards: • improved quality of our financial reporting; • increased comparability of our financial reports with those of similar companies internationally; • reduced costs of financial analysis that we perform; • removal of barriers that we previously encountered with respect to international capital flows; • reduced financial reporting costs, such as requirements to reconcile New Zealand’s Generally Accepted Accounting Practice (GAAP) to a foreign country’s GAAP, as well as requirements to achieve compliance with the listing and reporting requirements in foreign countries; • greater alignment with preferences of our institutional investors and analysts; and • greater confidence in our financial information, as indicated by users. The major international accounting firms describe the benefits of convergence/ harmonisation of financial reporting standards as follows (Andersen, BDO, Deloitte Touche Tohmatsu, Ernst & Young, Grant Thornton, KPMG, PriceWaterhouseCoopers, 2001, p.4): “High quality global accounting standards are needed to improve the ability of investors to make informed financial decisions, thereby leading to a reduction in risk for investors and ultimately a reduction in the costs of capital. Equally important, global standards can improve access to capital markets and reduce costs and the complexity for international companies by reducing or eliminating some of the multiple reporting obligations6.” Expected Costs of adopting IFRS Dunstan (2002) notes that there are a number of direct compliance costs of adopting IFRS. She considers these costs as consisting of two main components: • Costs that will be initially incurred by preparers, users, auditors and regulators in re-training, planning and preparing for the transition; • Economic consequences for entities of changing accounting policies and the effects this may have on contracts (e.g., debt contracts, management 6 Quote also used by Dunstan (2002, p.16). 15 remuneration contracts and investment decisions). Particular mention is made of the significance of the impact that IAS 38 Intangible Assets will have on some entities. In view of the FRSB’s earlier commitment to harmonisation with international accounting standards, Dunstan (2002) points out that these costs were inevitable even prior to the decision to adopt IFRS.7 This thesis does not attempt to differentiate between likely ‘inevitable’ costs of the former harmonisation process and increased costs that may have arisen specifically as a result of the 2002 ASRB decision to adopt IFRS (e.g., due to increased pace and scale of change). As to the extent of costs, they could arguably have been expected to be relatively insignificant as a result of New Zealand’s early alignment with international accounting standards. However, publications issued prior to adoption of IFRS by the accounting profession suggest otherwise and Ernst & Young went as far as to say “Our overriding comment is ‘Do not underestimate the enormity of the change management process required in conversion from NZ GAAP to NZ IFRS’” (Ernst & Young, 2004).8 Examples drawn from such publications, which elaborate on the costs described above include: • Project teams to plan for and manage the implementation of IFRS. These costs will depend on the size and seniority of those appointed to such teams; the scope and duration of the project (i.e. hours spent on IFRS preparation and implementation) and the extent to which use of consultants is necessary. • Changes to financial and accounting systems, processes and reporting. These costs will depend on the extent to which changes to hardware and software are necessary to ensure compliance with additional information needs, disclosure and reporting requirements under IFRS (e.g. financial instruments disclosures). • Training and recruiting and/or outsourcing. These costs will increase due to skill shortages and competition for appropriately qualified and experienced staff, especially for certain aspects of IFRS related issues, such as fair value 7 The terms ‘convergence’, ‘harmonisation’ and ‘adoption’ have similar but slightly different meanings. In this thesis, ‘convergence’ is used to describe processes such as the joint project between the IASB and FASB to align two separate sets of standards with each other. ‘Harmonisation’ is used to describe processes such as the FRSB’s decision to base new accounting standards on International or Australian accounting standards. ‘Adoption’ is used to describe processes such as New Zealand’s implementation of IFRS, which involve the introduction of an entire new set of accounting standards in place of the previous standards (GAAP) used in a particular jurisdiction. It is acknowledged that others may attach different meanings to these terms (e.g. the NZ process of adoption was similar to that of Australia, but Zeff & Nobes, 2010, conclude that Australia 'fully converge' rather than adopt IFRS). 8 It is acknowledged that the publications referred to above may have promotional objectives and hence be designed in part to emphasise costs and work required with a view to attracting new client work/ engagements. 16 measurement. Other costs and issues are considered by Dunstan (2002) but are institutional rather than firm specific (e.g., standard setting, audit and enforcement issues, New Zealand’s commitment to sector neutrality, New Zealand’s predominantly small entity size and the need to minimise compliance costs, and the New Zealand Australia Closer Economic Relations Trade Agreement 1983). As an indication of the extent of these costs, Li (2010) notes that the average compliance cost for adopting IFRS is estimated to be around £360,000 for UK companies, rising to £625,000 for companies with a market value between £1 billion and £2 billion. Long-term net benefits of adopting IFRS The importance of Dunstan’s (2002) conclusion (i.e., that adoption of IFRS will result in long term net benefits) is emphasised in excerpts from a memorandum from the chair of the FRSB to the chair of the ASRB in December 2002. The memorandum concerns consultations between their joint consultative group and interested parties on the adoption of IFRS: “There is strong support for the principle that the benefits of reporting must exceed the associated costs” (Bradbury & van Zijl, 2006). The IASB expresses similar sentiments: “The benefits derived from information should exceed the cost of providing it” (IASB, 2004, para. 44 of 1989 Framework), although there have been some amendments to this wording since then. While there is little empirical evidence available to date regarding the costs and benefits of implementing IFRS in New Zealand, some indications are provided by Griffin, Lont, & Sun (2009) and a PriceWaterhouseCoopers (2006) survey. Griffin et al. (2009) find an increase in audit fees over 2002 – 2006, which they associate reliably with IFRS adoption. Somewhat surprisingly, they also document a decrease in non-audit fees over the same period, but suggest that stricter independence requirements provide a potential explanation. The PriceWaterhouseCoopers (2006) survey of senior executives indicated that only 14% of respondents were convinced that benefits would arise as a result of IFRS adoption in New Zealand (mainly as regards comparability/consistency), whereas 38% were negative (mainly due to concerns that IFRS was inappropriate to their firm or industry or to New Zealand and added considerable cost and complexity for very little extra benefit). Remaining respondents were undecided. 17 2.3 CHAPTER SUMMARY This chapter provides the context within which this study is set, by describing the background to the adoption of IFRS in New Zealand. The chapter begins with a brief overview of the statutory reporting framework in New Zealand. Key considerations for this study, which are described in the overview, include an appreciation that: • Accounting standards in New Zealand have been closely aligned with international accounting standards since the 1970’s; • Proposals defining new reporting requirements for New Zealand entities, which are likely to be based on IFRS, are being considered by regulators and policy makers at the time of writing this thesis; and • New Zealand is accepted as a jurisdiction which offers a high quality reporting environment with strong enforcement of accounting standards. The second part of this chapter considers the motivation for adopting IFRS in New Zealand, including discussion of the expected costs and benefits of doing so. The literature reveals an expectation that long-term net benefits should arise as a result of adopting IFRS. This thesis contributes further empirical evidence relating to costs and benefits, which may provide additional insight as to whether the benefits of IFRS exceed the costs. The next chapter proceeds with a review of the literature relevant to adoption timing choice and the impact of IFRS. 18 CHAPTER 3: THEORY AND LITERATURE 3.1 INTRODUCTION The literature relevant to this study is reviewed in two main parts. The first part (3.2) provides an overview of accounting choice theory as the framework within which timing choice for the adoption of IFRS is investigated. Studies relating to adoption timing decisions are then considered in greater depth. These decisions tend to be explained in terms of either the type of ‘news’ implications that adoption brings i.e. good (bad) news; or the firm characteristics of the adopters (e.g. size, leverage). The second part (3.3) reviews studies which examine the impact of IFRS adoption. For the purposes of this thesis, these studies are categorised into two broad groups: (1) Consequences of IFRS: There are two main streams of literature here, namely capital market consequences of IFRS (e.g. market liquidity, cost of capital, value relevance) and accounting quality consequences of IFRS (e.g. earnings management, timeliness of loss recognition); (2) IFRS differences in financial information as a result of the changes from previously used GAAP (e.g. impact on figures in financial statements, financial ratios). The first of these broad groups concerns IFRS consequences, which may act as incentives for those making accounting choices regarding adoption of IFRS. These consequences are presumed to follow on as a result of the changes to financial information brought about by adoption of IFRS. The second group concerns the changes themselves (i.e., the nature and extent of differences to financial information brought about by adoption of IFRS). The greater the IFRS differences, the more likely that adoption of IFRS should result in capital market consequences or accounting quality consequences. A focus of this thesis is to document IFRS differences in order to contribute to an improved understanding of the extent to which IFRS adoption may have capital market or accounting quality consequences. 19 3.2 ACCOUNTING CHOICE 3.2.1 OVERVIEW The broad definition of accounting choice proposed by Fields, Lys & Vincent (2001, p. 256) provides a useful point of reference for this thesis: “An accounting choice is any decision whose primary purpose is to influence (either in form or substance) the output of the accounting system in a particular way, including not only financial statements published in accordance with GAAP, but also tax returns and regulatory filings.” Fields et al. (2001) remind us that, in a world of complete and perfect markets, there would be little need for accounting, accounting disclosures or accounting regulation, as all information relevant to our markets would be disseminated immediately and efficiently without the need for such accounting interventions. These accounting interventions are therefore seen as potential solutions to market imperfections such as the agency problem, information asymmetry, and even flow of appropriate information to affected external parties who do not have the same right to information as owners or other parties who have contractual relationships with the entities concerned. Barth et al. (2008) note, with reference to Watts & Zimmerman (1986), that accounting choices can be used in contrasting ways. They may be used to reveal private information about the firm (which may promote efficient contracting by reducing information asymmetry).9 However, accounting choices may also be used opportunistically to promote wealth transfers to self-interested principals or agents at the expense of less informed parties (by exploiting information asymmetry for private advantage and possibly misleading others about the firm’s economic performance in the process). The accounting choice literature also considers questions such as whether, how and how much accounting choice matters. This thesis considers these questions in relation to timing choice: whether the choice of early or late adoption of IFRS matters, how it matters and how much it matters. 9 See (Scott, 2006) for discussion of circumstances where this ‘disclosure principle’ breaks down (e.g., where release of private or proprietary information would result in excessive costs, this would not be in the interests of efficient contracting). 20 Fields et al. (2001) review the accounting choice literature up until the 1990’s. They refer to three main categories in discussing motivations for accounting choice: contracting, asset pricing and influencing external parties. As regards ‘contracting’, they note that, in general, research findings indicate that efficient contracting incentives work i.e., managers select accounting methods to increase their compensation (the ‘bonus hypothesis’) and to avoid breaching debt covenants (the ‘debt hypothesis’). They suggest, however, that the literature leaves some uncertainty as to whether these choices are made opportunistically or for value- maximising purposes. This implies that accounting choices assist in addressing the agency problem, but fall short of resolving it. ‘Asset pricing’ concerns information asymmetry and certain economic consequences of accounting information, namely the effect it has on share prices (i.e., market value of equity) and cost of capital. Holthausen (1990) suggests that accounting choices may be used by managers to impart private information about expected future cash flows (i.e., reducing information asymmetry in the interests of efficient contracting). Fields et al. (2001) also note the possibility of self-interested managers using accounting choice as a means to exploit information asymmetry opportunistically. Fields et al. (2001) describe findings concerning accounting choices being made for their effect on share prices as generally unconvincing mainly because of competing hypotheses, such as market efficiency and those mentioned above for contracting. They note mixed results regarding studies as to whether increased levels of disclosure result in decreased costs of capital. Fields et al. (2001) conclude that research in accounting and in other disciplines does not yet provide clear evidence as to whether markets are efficient or not, and that it is therefore difficult to draw strong inferences concerning the effect of accounting choices on asset prices. The third category, ‘influencing external parties’, considers two main hypotheses, namely that accounting choices are made to reduce/defer taxes and to minimise political costs. Fields et al. (2001) note evidence from this research supporting the tax hypothesis, but note that findings regarding the stock market effects of these actions is mixed. As regards political costs/regulation, they note that overall, the literature suggests that accounting choices are made to increase shareholder wealth (i.e., avoid regulation which will transfer wealth to external parties). Watts (1992) adds that accounting choices may also be made to encourage regulation which will transfer wealth in the other direction - from external parties (e.g., managers’ use of income 21 decreasing accounting methods to strengthen the case of firms who sought government protection through increased import restrictions). It is argued that such restrictions are likely to increase the wealth of the protected firms at the expense of consumers, as they reduce competition. Holthausen & Leftwich (1983) note two persistent associations with accounting choice: size as a proxy for political visibility and leverage as a proxy for agency costs of lending agreements. They note, however, that despite these associations, there is little evidence to show that these firm characteristics are suitable proxies. Watts (1992) adds that, while size does appear to be associated with income-reducing accounting choices, it can proxy for variables other than political costs. Calls such as these for more careful consideration of proxies have been answered by a number of studies which focus on associations between various firm characteristics and accounting choices. Watts (1992) provides further insight as to the importance of the theory of the firm and firm characteristics in explaining and predicting variations in accounting choice. He emphasises the role that accounting plays in organisational arrangements (e.g., use of cost centres as opposed to profit centres) and contractual arrangements (e.g., unusual sales agreements guaranteeing sale of all produce). He notes that these arrangements vary across firms (hence so do firm characteristics) and illustrates that optimal accounting choices vary accordingly (e.g., profit centres allow more decision rights to managers than cost centres, which in turn affects how managers are evaluated/compensated and hence has accounting implications). To conclude this overview, it is noted that accounting choice includes choices made at firm/manager level as well as at industry, country and international level. Watts (1992) makes the link between the accruals process and accounting choice and notes the trade- off between timeliness and reliability, which is involved. He stresses that accepted procedures are conservative, presumably to constrain managements’ bias towards inflating performance. He concludes by warning against the Securities Exchange Commission’s (then) interest in current value accounting. He comments that, where market prices are not readily observable and reliable estimates of value cannot be observed, this will interfere with the use of accounting earnings numbers in their primary function and reduce the value of earnings as a performance measure. These comments appear relevant to IFRS and this thesis in light of the increasing trend for companies to refer to “underlying performance/earnings” (see Chapter 5). 22 3.2.2 STUDIES RELATING TO ADOPTION TIMING DECISIONS ‘Good News’/‘Bad News’ and Adoption Timing Amir & Ziv (1997a) provide a useful summary of the findings associated with many adoption timing decisions, where regulators allow extended windows of time for adoption of new standards. Their model predicts that firms with ‘good news’ in relation to new standards will early adopt new standards; those for whom the impact of new standards will be neutral will choose to disclose this in the notes to the financial statements; and those with ‘bad news’ will delay adoption until it becomes mandatory. They argue that adoption timing signals managers’ private information about the impact of a new standard and they therefore anticipate a positive market reaction to early adoption. They note that they obtain results consistent with these predictions in a companion study (Amir & Ziv, 1997b), where they find that the average liability is smallest for firms early-adopting SFAS 106, larger for early-disclosing firms and largest for late adopters10. They also find positive market reaction to early adoption announcements. While ‘good news’ (‘bad news’) discussed above concerns a liability decreasing (increasing) effect, Verrecchia (1983) notes acceptance in the accounting literature that ‘good news’ (‘bad news’) denotes an income increasing (decreasing) effect contributing to a favourable (unfavourable) difference between actual earnings reported and the market’s expectation of earnings. The model and findings discussed above build on earlier studies of voluntary disclosure and adoption timing. Amir & Ziv (1997a) and others (e.g., Dye, 1985; Milgrom, 1981) note the influence of the adverse-selection argument that investors will discount their estimates of firm valuations where they know that managers are not disclosing private information. Managers therefore have an incentive to disclose so that they distinguish their firms from worse affected firms. Verrecchia (1983) and Dye (1985) refine this argument by pointing out that disclosure-related costs influence the incentive to disclose. This may result in delayed or ‘less-than-full’ disclosure. This is an important consideration in relation to this thesis which investigates adoption timing, as well as the nature and extent of voluntary narrative disclosures. 10 Statement of Financial Accounting Standards (SFAS) No. 106: Employers’ Accounting for Post- Retirement Benefits Other Than Pensions 23 Both Verrecchia (1983) and Dye (1985) find evidence to support the prediction that revelation of private information constitutes good news. Trueman (1990) adds insight as to the timing of disclosure. He notes research findings that associate early (late) release of earnings reports with stock price rises (declines) and that market reaction is more positive the earlier the reports are released. He finds evidence supporting two possible explanations for these findings, namely that firms with unfavourable earnings may require a delay in order to: (1) engage in earnings management; or (2) observe other firms’ earnings to assess the extent to which they will engage in earnings management. Adoption Timing and Income Effects A number of prior studies reveal a consistent relationship between adoption timing and income effects. Sweeney (1994) finds that managers of firms approaching violations of accounting-based debt covenants are more likely to early adopt income-increasing standards and defer income-decreasing standards where new accounting standards allow a transitional period. Scott (1991), Langer & Lev (1993) and Ali & Kumar (1994) study the adoption timing choices made by firms regarding a new pension accounting standard (SFAS No.87), which they describe as an ‘income-increasing’ standard.11 Scott (1991) finds that three political variables are significant predictors of late adoption, namely size, regulation and potential legal sanctions for unfair pricing practices. He also finds that management compensation contracts, the magnitude of the income effect of adoption, whether the firm was a ‘bath’ firm and the earnings position of the firm relative to the prior year are all associated to some degree, with the choice available to firms to adopt the new standard in 1986 or 1987.12 Langer & Lev (1993) find in their study of adoption timing motives for SFAS No. 87 that only one of eight proxies examined, namely increased reported earnings, consistently discriminates between early and late adopters. Ali & Kumar (1994) add to these findings by considering interactions between firm characteristics and the magnitudes of the financial statement effects of adoption of the standard. They find statistically significant coefficients supporting the capital market consequence explanations advanced in the prior studies. They note that their results enhance the ability to explain accounting decisions and mitigate the problem of omitted variables. For example, they suggest that firm size may proxy for reasons other than political cost in explaining why large firms early adopt, such as their being more familiar with the standard, or because advisors influencing the new pension 11 Statement of Financial Accounting Standards (SFAS) No. 87: Employers’ Accounting for Pensions 12 Scott refers to a phenomenon recognised in the financial press/academic journals as the ‘big bath’, where firms experiencing financial difficulties often endeavour to incorporate as much negative news as possible into one annual report so as to improve future annual reports. 24 accounting requirements favoured the needs of larger firms. Various studies regarding the introduction of SFAS No.52 reveal similar findings.13 Benjamin, Grossman & Wiggins (1986) find that early adoption of this standard was largely motivated by a favourable impact on income and earnings per share. Ayres (1986) notes that only one of 103 firms that adopted this standard early, (i.e., in the first year of a possible three year period) reported decreased earnings as a result of adoption. Brown’s (1985) findings that there are significant differences between early and late adopters of this standard has important implications for this thesis. She finds that early adopters reported increased profits and decreased debt-to-equity ratios and notes that her results are consistent with those of a Coopers & Lybrand study, which show an average increase in net income of 11% for early adopters.14 She concludes that her results imply that managers follow the mechanistic hypothesis (i.e., they believe that investors are not able to distinguish between economic profits and accounting profits). Brown & Brandi (1986) note an average income increasing effect of $13.78 million for early adopters of SFAS No.52. They suggest that the majority of prior studies substantiate the hypothesis of semi-strong market efficiency and that investors are not naive and are therefore able to distinguish substantive changes in financial reports from changes based merely on accounting method choices (i.e., the ‘no effects’ hypothesis). However, they find that security price reactions to the new standard during the first year of its three-year adoption period (1981 – 1983) suggest that allowing a transition period for the adoption of new standards may confuse investors and lead to capital market inefficiencies. Abnormal returns for early adopters are compared to those for late adopters of SFAS No.52 and they find significant differences which begin two weeks prior to year-end and continue for five months. They therefore question the value of allowing transition periods for the adoption of new standards which could significantly affect financial statements. Significant IFRS differences identified in this thesis may indicate similar policy implications (for future research) in respect of New Zealand’s decision to allow a two year voluntary adoption period for IFRS, prior to making it a mandatory requirement for listed companies. Adoption Timing and Firm Characteristics Many studies consider associations between firm characteristics and accounting choice. 13 Statement of Financial Accounting Standards (SFAS) No. 52: Foreign Currency Translation 14 Coopers & Lybrand, Foreign Currency Translation: An Implementation Study (New York, Coopers & Lybrand, 1982) 25 Hung & Subramanyam (2007) control for self-selection bias in their study of German firms which voluntarily adopt IAS by predicting, on the basis of prior studies (e.g., Harris & Muller, 1999; Leuz, 2003), that voluntary adoption is a function of financial performance, leverage, size, financing needs (cross-listing and increases in debt or equity) and industry. Iatrides & Joseph (2006) examine characteristics of UK firms in relation to the timing of their adoption of SSAP 20, a standard relating to foreign currency translation which became effective in April 1983.15 They conclude that early adopters tend to be larger firms and that growth options, profitability, leverage and management payout have strong predictive power in relation to adoption timing decisions. Christie (1990) aggregates test statistics across studies of contracting and size hypotheses and finds six variables which are associated with accounting choices at statistically significant levels: managerial compensation, leverage, size, interest cover and dividend constraints and risk.16 Earlier studies specifically examining firm characteristics/ attributes of early adopters include Ayres (1986) and Brown (1985). Ayres, in contrast to Iatrides & Joseph (2006), finds that early adopters in the USA of SFAS No. 52 (a foreign currency standard with similar requirements to that of SSAP 20) are smaller. She also finds that early adopters are manager-controlled as opposed to owner-controlled (lower shareholdings by management), have lower earnings increases prior to adoption and are closer to debt and dividend constraints than late adopters. While the findings regarding size of early adopters by Iatrides & Joseph (2006) and Ayres (1986) appear contradictory, a comparison of the two studies reveals that the US dollar strengthened while the UK pound weakened over the period of each of these adoption studies. Ayres (1986) notes that early adoption of this standard would ordinarily have an income-increasing effect over the period concerned: a strengthening dollar meant net translation losses, but these translation losses flowed to the balance sheet rather than through the income statement in terms of the new standard, hence the income increasing effect. She explains her findings regarding size in terms of the political cost hypothesis: larger entities prefer to defer income-increasing accounting choices. The effect in the UK study was the opposite (income-decreasing), hence larger entities there preferred to early adopt. Interestingly, Iatrides & Joseph (2006) make only passing reference to the income- 15 UK Statement of Standard Accounting Practice (SSAP) No. 20 Foreign Currency Translation. 16 While this study is not limited to adoption timing choice, the aggregated results across a number of accounting choice studies confirm the explanatory power of many variables associated with adoption timing choice. 26 decreasing effect as a political cost motivation for larger firms to early adopt. They suggest a number of other reasons that early adopters are likely to be large firms, with support from literature published after the Ayres (1986) study, such as a desire to avoid political and regulatory scrutiny by demonstrating prompt compliance (e.g., D'Souza, 1998; Kim & Kross, 1998); better alignment with analysts and investors’ expectations (e.g., Kasznik, 1999; Lambert, 2001) and greater resources to allow them to meet costs of implementing new standards, particularly where these costs are high (e.g., Sami & Welsh, 1992). Similarly, Li (2010) claims that prior research suggests that voluntary IFRS adopters are likely to be larger, have greater financing needs, be cross-listed and have better long-run market performance. Similar to the earlier studies by Verrecchia (1983) and Dye (1985), Renders & Gaeremynck (2007) argue that IFRS leads to increased disclosure and reduced accounting choices and hence loss of private benefits for firm insiders (e.g., fewer opportunities to manage earnings reduces insiders’ ability to divert assets from the firm, influence bonuses, buy assets from the firm below market value etc.). They find that IFRS is more likely to be adopted early in countries with strong investor protection laws and corporate governance, as the cost (in lost private benefits) is smaller. 27 3.3 STUDIES RELATING TO THE IMPACTS OF IFRS 3.3.1 INTRODUCTORY COMMENTS AND EARLY STUDIES Recent studies relating to the impacts of IFRS adoption focus on capital market consequences and accounting quality consequences. Chapter 2 identifies some capital market consequences (e.g. reduced costs of capital) and improved accounting quality as expected benefits of adopting IFRS. Furthermore, the review of accounting choice literature in the previous section indicates that capital market consequences and accounting quality consequences may be potential incentives for IFRS adoption timing choices. The next part of this section (3.3.2) therefore begins with a brief overview of recent empirical evidence regarding the consequences of IFRS. Recent studies which analyse the IFRS differences in financial information, as a result of the changes from previously used GAAP, are much rarer. However, it is the nature and extent of these differences which are likely to be the drivers of any consequences which flow from IFRS adoption. Similarly, the more significant these IFRS differences are, the more likely they are to play a role in adoption timing decisions. IFRS differences (3.3.3) are therefore a focus for this thesis and are reviewed in more detail than the consequences of IFRS (3.3.2). As regards earlier studies which are of relevance to the impacts of IFRS, Hung & Subramanyam (2007) contend that studies of the effects of IAS based on pre-1998 financial statements are unlikely to be representative. The core IAS standards were completed in 1998 and removed the choice of partial adoption (i.e., ‘cherry picking’) of IAS by requiring full implementation of IFRS. Furthermore, Dunstan (2002) considers research up to and including 2002, that relates to the impact of IFRS/ IAS, with specific reference to likely impacts on New Zealand. Consideration of these earlier studies is therefore limited to the brief summary below of Dunstans’s conclusions. Dunstan (2002) comments that although two studies find that value relevance of accounting information did improve when firms switch to IAS (Auer, 1996; Barth, Clinch, & Shibano, 1999), various other studies find either that different accounting systems were equally informative (Ashbaugh & Olsson, 2002; Chan & Seow, 1996; 28 Leuz, 2003), or that differences were due to institutional factors like differences in infrastructure and enforcement rather than the accounting systems themselves (Alford, Jones, Leftwich, & Zmijewski, 1993; Ball, Kothari, & Robin, 2000). Dunstan (2002) therefore concludes that there is little evidence that IFRS will impact on value relevance. She also considers other factors that affect access to capital and economic growth. She notes research findings that link improved financial information/disclosure to greater confidence in the quality and transparency of that information; this in turn is linked to decreased uncertainty and hence lower risk assessments and cost of capital (Botosan, 1997; Choi, 1973). However, she contends that studies such as Leuz & Verrecchia (2000) and Barth et al. (1999) establish positive consequences of adopting IAS for developing or low quality reporting countries. Dunstan (2002) therefore contends that these findings cannot be generalised to New Zealand and that there is no support for an argument that adoption of IFRS in New Zealand will contribute to improved access to capital or economic growth. 3.3.2 RECENT STUDIES RELATING TO CONSEQUENCES OF IFRS Capital Market Consequences Daske et al. (2008) make similar observations to those noted above by Dunstan (2002), citing various studies which suggest IFRS should improve liquidity, decrease cost of capital and increase firm value, but note counter arguments in other studies which suggest that capital-market effects of IFRS adoption could be negligible in relation to other factors, such as firms’ reporting incentives, concurrent efforts to strengthen enforcement regimes and recent governance and auditing reforms in many countries. Daske et al. (2008) conduct an extensive examination of the capital market consequences of mandatory IFRS adoption for a large sample of firms across 26 countries covering fiscal years ending on, or after, 1 January 2001 through to 31 December 2005. Their findings indicate that IFRS adoption appears to be associated with positive consequences for market liquidity, cost of capital and firm value. These findings are most pronounced for voluntary adopters, both in the year when they switch and again later, when IFRS becomes mandatory. They caution that the initial impact is likely to be due to self-selection and that the mandatory impact will be affected by omitted variables. Further support for the view that factors other than IFRS contribute significantly to capital market effects is evident in other studies (e.g., Barth et al., 2008; Byard et al., 2011; Lee et al., 2008; 29 Li, 2010). Li (2010) studies the effect of mandatory adoption of IFRS on cost of capital. A decrease of 47 basis points is found for mandatory adopters, but only in countries with strong legal enforcement. Contrary to Daske et al. (2008), there is no significant decrease found in cost of capital for voluntary adopters, after the date IFRS becomes mandatory. Lee et al. (2008) find, across a sample of 17 European countries, that the cost of capital only reduced in countries with high reporting incentives and enforcement. Covrig, DeFond, & Hung (2007) test the assertion that voluntary adoption of IFRS enhances the ability of firms to attract foreign capital. They find that average foreign mutual fund ownership is significantly higher among IAS adopters. Goodwin et al. (2008) find that the reconciliation note for the earnings adjustments arising from IFRS adoption in Australia contained no new information. However, a later study by Christensen, Lee & Walker (2009) reports contrasting findings for IFRS reconciliations in the UK These are both single country studies where previous GAAP is perceived to be relatively similar to IFRS, unlike many of the cross-country EU studies mentioned above where differences between IFRS and local GAAP are more pronounced. A point of distinction in the study by Christensen et al. (2009), however, is a focus on debt contracting. The authors note that the focus of prior studies risks attributing IFRS market reactions to information about future cash flows rather than the likelihood of violating debt covenants. Christensen et al. (2009) find evidence of market reactions to IFRS reconciliations, which are more pronounced for firms with a greater likelihood and costs of covenant violation and for early announcements. They conclude that their findings suggest that IFRS reconciliations do convey new information to the market and that, in the UK, this is more likely due to debt-contracting implications and consequent wealth transfers between lenders and shareholders. They argue that this is so, because the quality of UK GAAP is similar to that of IFRS and therefore unlikely to convey new information about future operating cash flows, whereas IFRS is likely to affect technical compliance with debt covenants. In support of this argument, they note that if IFRS did force firms to reveal more private information, timing of disclosures should not affect market reaction. The findings of Christensen et al. (2009) are of particular relevance to this thesis, because of the similarity in settings: both are concerned with only one country where there is strong enforcement and where previous GAAP is perceived to be similar to IFRS (see Kabir et al., 2010; 30 and Stent et al., 2010 with respect to the New Zealand setting). Wang & Welker (2011) note that there is broad support for the concept that IFRS reconciliations convey new information in prior literature. They document circumstances in their study where market returns are associated with the reconciliation of net income between local GAAP and IFRS. The focus of their study is to investigate whether adoption of IFRS impacts upon equity financing decisions in Australia and Europe. They find that the likelihood of equity issuance and the size of equity issues for the three years prior to IFRS reconciliation disclosures are negatively associated with the unexpected change in net income arising from adoption of IFRS. They suggest that their findings support the premise that firms exploit information asymmetry to opportunistically decide on the timing and size of equity issues/ repurchases. Overall, recent studies find mixed results for value relevance of IFRS. Barth et al. (2008), find more value relevance and Hung & Subramanyam (2007) find no evidence that IFRS improves value relevance. Daske et al. (2008) note an increase in equity valuations but only if they account for the possibility that the effects occur prior to the official adoption date. This echoes comments by Fields et al. (2001) and Dunstan (2002) that findings related to the share price effect of accounting choices are generally unconvincing. In assessing the value relevance of IFRS, a pre- requisite is that the adoption of IFRS should have a significant impact on financial statements and common financial statement ratios. Hence, the objective of the first phase of this thesis is to provide descriptive evidence of the impact of IFRS on financial information. No published New Zealand specific studies on capital market consequences were found at the time of writing this thesis. Accounting Quality Consequences Barth et al. (2008) find that adoption of IFRS is associated with higher accounting quality. They study firms from 21 countries applying IAS between 1994 and 2003 and find that there is less evidence of earnings management, more timely loss recognition and more value relevance of accounting information for their sample firms than for a matched sample of firms applying non-US domestic GAAP. Byard et al. (2011) study the effect of mandatory IFRS adoption on financial 31 analysts, using early (voluntary) adopters as a control group to counter confounding events such as improved enforcement in the EU over the test period. They find over four years (two years pre- and two years post-mandatory IFRS) that analysts’ absolute forecast errors and forecast dispersion decrease, but only for IFRS adopters from countries whose domestic GAAP differs substantially from IFRS. Where these countries have strong enforcement regimes they find statistically and economically significant decreases in the above variables for IFRS adopters; where these countries have weak enforcement regimes, they find larger decreases in the variables for firms with stronger reporting incentives (e.g., growth opportunities, smaller closely held shareholdings, Big 4 auditors). Before partitioning their sample, they find no significant differences between mandatory IFRS adopters and the control group. A study of market reaction to 16 events associated with the likelihood of IFRS adoption in the EU by Armstrong et al. (2010) yields results consistent with those of Byard et al. (2011). They find an incrementally positive reaction for firms with lower quality pre-adoption information and higher pre-adoption asymmetry i.e., similar to the findings of Byard et al. (2011) regarding firms with substantial differences between domestic GAAP and IFRS. They also find an incrementally negative reaction for firms from code law (weak enforcement) countries. New Zealand specific studies on Accounting Quality Consequences Balshaw & Lont (2010) find that disclosure of operating expenses improves after the adoption of IFRS, although they note continued inadequacies which detract from transparency. They compare the 2002 – 2007 disclosures of early adopters of IFRS to those of companies who continued to report under old New Zealand GAAP until IFRS became mandatory. Cheong et al. (2010) find that analysts’ earnings forecast errors decrease in post- IFRS periods in Australia, Hong Kong and New Zealand. They also find that intangibles capitalised under new IFRS requirements are negatively associated with analysts’ earnings forecast errors. They suggest that these findings indicate that IFRS improves the quality of financial reporting and value relevance. Kabir et al. (2010) find an increase in absolute discretionary accruals under IFRS and hence, in contrast with Cheong et al. (2010), suggest that their results indicate lower quality earnings under IFRS. Kabir et al. (2010) also find no significant 32 difference between pre- and post-IFRS as regards the ability of earnings to predict one-year ahead cash flows. They note that while these findings differ from other studies in European countries, they are consistent with the findings of Goodwin et al. (2008) in Australia and offer a number of potential explanations relating to differences between the European and Australasian settings. Austin (1997) notes deficiencies, weaknesses and potential information gaps in the reporting of intangible assets both before and after adoption of IAS 38 Accounting for Intangible Assets (e.g., loss of relevant information on internally generated assets that this standard requires should be de-recognised). 3.3.3 RECENT STUDIES RELATING TO IFRS DIFFERENCES The study that is most relevant to the first phase of this thesis is Hung & Subramanyam (2007). They examine the financial statement effects of adopting IAS during 1998 – 2002 by direct comparison of financial statements prepared under both IAS and German GAAP (referred to as Handelsgesetzbuch or ‘HGB’). They find that the adoption of IAS resulted in “… widespread and significant changes to deferred taxes, pensions, property, plant and equipment, and loss provisions” (Hung & Subramanyam, 2007, p. 625). Overall, total assets and book value of equity were found to be significantly larger under IAS than under HGB, while variations in book value and net income were found to be significantly higher. Hung & Subramanyam (2007) make a case for a country-specific approach which considers the direct effects of adopting IAS for the same set of firm years. Such an approach would help to overcome problems associated with comparing across countries with different institutional arrangements, as well as controlling for time- series differences. The first phase of this thesis makes use of the mandatory IFRS/local GAAP reconciliations (now an IFRS 1 requirement), while their study must rely on voluntary IAS/local GAAP reconciliation disclosures. Hung & Subramanyam (2007) are only able to observe ‘book value’ reconciliation information (i.e., equity adjustments) for 57 firms; and “net income” reconciliation information (i.e., profit at end of prior period adjustments) for 31 firms in their sample of 80 firms. Hence, this thesis shares the advantages of a country specific approach noted in the Hung & Subramanyam (2007) study, but avoids an important source of self selection bias present in their study. 33 There are two further contributions which this thesis makes, which add to those of Hung & Subramanyam (2007): (1) The investigation period for this thesis is more current than theirs. As noted earlier, this is important in view of the significant, and on-going amendments to IFRS since the ‘stable platform’ was achieved in 2005. (2) Hung & Subramanyam (2007) consider the adoption of IAS where German GAAP was known to differ significantly from IAS, whereas old New Zealand GAAP is perceived to be relatively similar to IFRS. Christensen et al. (2009) find that in spite of IFRS being relatively similar to UK-GAAP, the IFRS reconciliations contained information that analysts considered relevant for firm valuation and that firms opportunistically tended to delay unfavourable reconciliations. New Zealand specific studies on IFRS Differences A number of earlier studies focussed on IFRS differences arising from individual standards rather than full adoption (e.g., Austin, 1997; Bishop, Bradbury, & van Zijl, 2005; Wong & Wong, 2005). Austin (1997) notes, inter alia, substantial decreases in net assets of some companies after adoption of IAS 38 Accounting for Intangible Assets, as a result of the requirement to derecognise self-generated intangibles. Bishop et al. (2005) find that the reporting of convertible financial instruments in accordance with NZ IAS 32 Financial Instruments: Presentation would result in higher amounts for liabilities and higher interest, as this standard has a broader definition of liabilities than was applied under previous GAAP. Wong & Wong (2005) analyse New Zealand companies with intangible assets and find that the mean enterprise value to earnings before interest and tax (EV/EBIT) and price to earnings (PE) multiples with amortization of intangibles of 12.403 and 13.586, respectively, decrease to 10.971 and 12.346, respectively, without amortization of intangibles. Kabir et al. (2010) find that total assets, total liabilities and net profit were significantly higher under IFRS than pre-IFRS GAAP. Income increases were attributed mainly to adjustments for goodwill, other intangibles and investment property; while income decreases were attributed mainly to employee benefits and share-based payments. Additional analysis (for both IFRS differences and consequences of IFRS) indicates that results for early adopters are similar to those for the full sample. Stent et al. (2010), in a study related to this thesis, examine the impact of IFRS on 34 financial statement elements and key financial ratios. Findings are generally consistent with those of Kabir et al. (2010), but Stent et al. (2010) offer richer detail and consider also revenue and additional common financial ratios reflected in the Du Pont analysis. While revenue is not found to differ significantly under IFRS, it is noted that the net effect of IFRS changes is a decrease in equity, as the largest and most significant impact is increases in liabilities. Significant changes are found for all ratios (all increases except for asset turnover), which has implications for financial analysis, valuation, credit decisions and contracting agreements. Small firms are generally found to be less significantly affected by the adoption of IFRS. In contrast to Kabir et al. (2010), results for early adopters are found to be quite different from those for late adopters. This is most evident in the impact of IFRS on ratios, which is generally found to be more strongly significant for late adopters than for early adopters (e.g. change in leverage is found to be insignificant for early adopters but significant at the 1% level for late adopters). These contrasting results may be due to sampling differences or comparison methods: Kabir et al. (2010) compare early adopter results with those of the full sample while Stent et al. (2010) obtain a more precise comparison by comparing early adopter results to those of late adopters. As noted in Chapter 1, this thesis extends the Stent et al. (2010) study by expanding the sample of early adopters from 16 to 40 firms. This allows for a more comprehensive comparison of early and late adopters in this thesis, whereas the sample selected for the related Stent et al. (2010) study was intended to be representative of the overall population of listed companies in New Zealand. 35 3.4 CHAPTER SUMMARY This chapter provides an overview of the theory of accounting choice in relation to IFRS adoption. One type of accounting choice concerns adoption timing of new accounting standards where regulatory bodies allow firms a discretionary period for adoption, before the standard(s) become(s) mandatory. There is a significant body of literature concerning the early adoption of new accounting standards especially since 1973, when the Financial Accounting Standards Board (FASB), became the designated organization in the US’s private sector for establishing standards of financial accounting and reporting (refer http://www.fasb.org/facts/ ). A review of this literature, in the first part of this chapter, reveals that most researchers have examined such choices in relation to the change of a single accounting standard. This thesis investigates firms’ adoption timing choices in the face of sweeping change, exploiting a rare opportunity offered by the institutional arrangements for the adoption of IFRS in New Zealand. The second part of this chapter reviews literature relating to the impacts of adopting IFRS. As can be expected, there is common ground between these studies and those relating to early adoption decisions. Understanding the impacts of IFRS is an important part of the process of investigating the motivations behind adoption timing decisions. Hickey et al (2003, p.4) make the following comment in relation to New Zealand’s early adopters: “It is likely that many of the early adopters will be cross-listed, have significant subsidiaries in Australia or Europe, or themselves be subsidiaries of a parent based in Australia or Europe. But some other entities may elect to adopt early as a means of demonstrating commitment to high-quality financial reporting.” This is a practical comment, which suggests efficient contracting motives. Wang & Welker (2011), on the other hand suggest that firms exploit information asymmetry to opportunistically decide on the timing and size of equity issues/ repurchases. The accounting choice literature reviewed in this chapter provides many other perspectives and competing hypotheses; a rich framework within which to investigate potential motivations for adoption timing decisions. The chapter which follows describes how this investigation is to be conducted. 36 CHAPTER 4: RESEARCH DESIGN, DATA AND SAMPLE SELECTION 4.1 OVERVIEW OF RESEARCH METHODOLOGY The aim of this thesis is to investigate what motivates entities to adopt IFRS early or late. To seek a better understanding of this research problem, three research questions are formed to explore potential motivating factors: 1. What impact does IFRS adoption have on the financial information of early as opposed to late adopters? 2. What do discretionary narrative disclosures in annual reports reveal about early and late adopter firms’ attitudes to IFRS adoption? 3. What are the costs and perceived benefits of IFRS adoption for early as opposed to late adopters? In addressing each of these research questions, there are two main sub-questions which are investigated. The first sub-question is quantitative, while the second is qualitative in nature: a) Are there significant differences between early and late adopters? b) What is the nature of significant differences (if any) that are identified? All research procedures are therefore designed to clearly distinguish results of early adopters from those of late adopters. To answer these questions, the research design is based on what Creswell (2003) describes as a ‘Mixed Methods Approach’, as it employs both quantitative and qualitative research methods. He explains that this approach is one in which knowledge claims tend to be based on pragmatic grounds. Pragmatism is further defined as not being committed to any one system of philosophy and reality. Creswell (2003, pp. 11- 12) claims that this approach places the emphasis on the research problem rather than the research methods and “ … opens the door to multiple methods, different world views, and different assumptions, as well as to different forms of data collection …”. The resulting research methodology is based on the assumption that collecting diverse types of data will best provide an understanding of the research problem. 37 This thesis exploits the benefits of this approach. It begins by using ‘Accounting Choice’ as a theoretical framework. This provides rich theoretical perspectives for the research problem and is supported by a well developed body of literature (reviewed in the previous chapter). Accounting choice theory is based on economic theories of the firm and costly contracting (Watts, 1992). As a result, much of the prior literature reviewed for this thesis is based on positivist ontological assumptions and hence epistemological assumptions which favour objectivism and quantitative research. Laughlin (1995) however comments on tension between the ‘economics’ and ‘behavioural’ wings of the accounting academic research community and notes a proliferation of alternative ‘behavioural’ approaches to empirical research in accounting during the 1990’s. In deference to Creswell’s (2003) arguments in favour of pragmatically incorporating the strengths of different approaches into one research methodology, qualitative methods traditionally associated with ‘behavioural’ or ‘naturalistic’ approaches are used, as well as the quantitative methods which are more commonly associated with accounting choice research.17 Overall, the study is however, characterised by an emphasis on objectivity, replicability, reliability and validity. It is guided by prior research and is concerned with cause-effect relationships (causes of early as opposed to late adoption). In summary, while pragmatism does not commit the researcher to any one system of ontological or epistemological assumptions, the characteristics of the research methodology for this study tend to be positivist in nature. Six major strategies for mixed methods research are explained by Creswell (2003), although he notes that these do not exhaust all possibilities.18 One of these, ‘Sequential Explanatory Strategy’, best describes the approach taken for this thesis. The strategy uses qualitative results to assist in explaining and interpreting findings of a primarily quantitative study. This strategy is particularly useful when unexpected results arise from the quantitative study. Also a broader perspective will be gained as a result of using both methods, than would be the case using only one of them. When compared to the other major strategies, a particular strength of ‘Sequential Explanatory Strategy’ is 17 Guba & Lincoln (1989) identify ‘Scientific’ and ‘Naturalistic’ approaches as the two broad research methodologies most commonly used. There are similarities between these concepts and the ‘economics’ and ‘behavioural’ wings referred to by Laughlin (1995). 18 The alternative strategies include: ‘sequential exploratory’, ‘sequential transformative’, ‘concurrent triangulation’, ‘concurrent nested’ and ‘concurrent transformative’. Differences in these strategies relate inter alia to the order in which quantitative and qualitative procedures are conducted, whether or not there is a predominant method, the extent to which integration of results takes place during interpretation, and theoretical perspectives used. Further consideration of these alternatives is considered to be beyond the scope of this study. 38 simplicity of design, but a particular weakness is the length of time involved in data collection. Data collection for this thesis was a lengthy process, but this was due in part to the extended period over which IFRS adoption took place, with most early adopter data becoming available during 2006 and 2007, while late adopter data became available mainly during 2008. Figure 1, which was used in Chapter 1 to provide a diagrammatic overview of this thesis, is adapted from Creswell (2003). It illustrates how the ‘Sequential Explanatory Strategy’ is applied in this thesis. An overview of the research design follows, with reference to Figure 1 (inserted again over page for ease of reference). This overview includes explanations of the research methods used to address the research questions which were introduced at the beginning of this chapter. Figure 1 demonstrates that the research is conducted using Accounting Choice theory as a framework. A bar highlights the research problem, namely to investigate the motivations for adoption timing of early as opposed to late adopters of IFRS. The numbering of procedures and the order in which quantitative and qualitative methods appear, indicate the order in which these processes take place. The study begins with quantitative methods to collect data relating to the impact of IFRS on financial information. This should provide context for subsequent qualitative and quantitative methods. To address the first research question, the impact of IFRS adoption on financial statements and some common financial ratios is investigated (section 4.2). To begin, the impact on financial statements is analysed by financial statement component and by the specific accounting standards concerned. In addition, five key ratios that may influence analysts are selected for analysis. Descriptive statistics are used to analyse the nature, frequency, extent and significance of IFRS differences (quantitative analysis). Further detail is then extracted regarding the most frequent IFRS adjustments to assist in identifying potential incentives for adoption timing decisions (qualitative analysis). 39 Figure 1 Sequential explanatory strategy used for mixed methods research approach       Theoretical framework of ‘Accounting Choice’ (Ch. 3)  Quantitative: • Capture OLD GAAP & IFRS values (e.g. assets, expenses etc.)  • Capture values($) and source (standard) of IFRS adjustments  • Compute ratios using OLD GAAP & IFRS values (e.g. leverage)  Motivations for IFRS adoption timing:  Early Adopters v Late Adopters? 1. COLLECT DATA FROM FINANCIAL STATEMENTS Qualitative: Extract narrative descriptions for IFRS adjustments 2. ANALYSE DATA  Quantitative: • Descriptive statistics for IFRS differences  • Significance testing of OLD GAAP/IFRS and EA/LA differences  • Most frequent causes of IFRS differences (standards)  Qualitative: Summary of narratives for most frequent of IFRS adjustments  3. INTERPRET INTERPRET RESULTS 4. COLLECT DATA FROM ANNUAL REPORTS Quantitative: • Count number words used in IFRS related disclosures  • Measure space devoted to IFRS related disclosures    5. ANALYSE DATA Quantitative:  • Form oriented content analysis (count data)  • Descriptive statistics (word count, space usage, theme score)  • Significance of EA/LA differences  • Multivariate regression on importance of IFRS for EA  Entire analysis of financial  statements data:   Impact of IFRS on financial  information (EA v LA)?  6. INTERPRET Entire content analysis of annual  report discretionary narratives:  Attitude to IFRS adoption          (EA v LA)?  7. CONDUCT SURVEYS (QUESTIONNAIRES & INTERVIEWS)  Quantitative:  Obtain values relating to costs & benefits of IFRS (questionnaires)  8. ANALYSE DATA Quantitative:  • Descriptive statistics on IFRS costs and benefits  • Significance of EA/LA differences  Qualitative: Meaning oriented content analysis (themes)  Qualitative: Summary of adoption timing insights  Qualitative: Extract/prepare discretionary narrative disclosures (outside of  financial statements) related to adoption of IFRS   Qualitative:  Obtain adoption timing insights (semi‐structured interviews)  9. INTERPRET Entire analysis of survey data:  Costs/benefits of IFRS (EA v LA)?  Other adoption timing insights?  CONCLUSION (Ch. 6) Ch. 4.2 Ch. 5.2 Ch. 4.3 Ch. 5.3 Ch. 4.4 Ch. 5.4 40 Next, IFRS-related discretionary narrative disclosures in annual reports are examined to address the second research question (section 4.3). Content analysis is used to investigate the extent and nature of such disclosures in annual reports for the year in which IFRS is first adopted. Descriptive statistics are used to analyse word counts and space devoted to IFRS related disclosures (quantitative analysis). The disclosures are then analysed to identify the main themes communicated by management, because such themes may provide insights regarding influences for IFRS adoption timing decisions (qualitative analysis). The quantitative measures, as well as derived scores for the IFRS themes identified, are analysed for significant differences between EA and LA. A multivariate model is also developed to control for firm specific factors which may influence disclosures. The costs and benefits of IFRS adoption are then considered to address the third research question (section 4.4). This evidence is obtained by means of questionnaires and semi-structured interviews. Information gathered through the questionnaires includes measures of the cost and level of effort involved in implementing IFRS, as well as opinions of senior financial executives regarding expected benefits of adopting IFRS. Descriptive statistics are used to report these measures and the significance of differences between EA and LA for these measures (quantitative analysis). Comments in questionnaires and semi-structured interviews provide further insights regarding IFRS adoption timing choices (qualitative analysis). Sections 4.2 to 4.4 provide further detail on the research questions, the research methods used to investigate them and the data examined. 41 4.2 IMPACT OF IFRS ADOPTION ON FINANCIAL INFORMATION 4.2.1 RESEARCH QUESTION AND HYPOTHESES Research Question 1 and related sub-questions RQ1 What impact does IFRS adoption have on the financial information of early as opposed to late adopters? 1.1 Are there significant differences between early and late adopters? 1.2 What is the nature of significant differences (if any) that are identified? The main research method used to investigate this research question involves analysis of descriptive statistics relating to IFRS differences for the EA and LA sample firms. The analysis is then extended to address the sub-questions. RQ1.1 Are there significant differences between early and late adopters? This sub-question is investigated by testing the null hypothesis (EA = LA) for each area of financial information analysed, resulting in the following hypotheses: H01: There are no significant differences in financial information (FI) prepared under IFRS and financial information prepared under NZ GAAP, for Early Adopters or for Late Adopters i.e., H01a: µFI of EA (IFRS) = µFI of EA (NZ GAAP); H01b µFI of LA (IFRS) = µFI of LA (NZ GAAP) HO2: Any change in financial information for Early Adopters (∆FI for EA) as a result of the adoption of IFRS will not be significantly different from the change in financial information for Late Adopters (∆FI for LA) i.e., H02: µ∆FI of EA = µ∆FI of LA Where: FI represents the financial statement components (e.g., assets, revenue) and calculated key ratios (e.g., return on equity, leverage) as specified in 4.2.2 below. Hypotheses 1a and 1b are tested using the Wilcoxon test for matched pairs (OLD GAAP figures are compared to equivalent IFRS figures for each sample firm); hypothesis 2 is 42 tested using the Mann-Whitney U test (IFRS adoption changes for early adopters are compared to those of late adopters). RQ1.2 What is the nature of significant differences (if any) that are identified? This sub-question is addressed by analysing IFRS adjustments in further detail. This requires consideration of descriptive evidence regarding a further level of sub- questions: 1.2.1 Which accounting standards are the most frequent sources of IFRS adjustments (e.g., NZ IAS 12 Income Taxes)? 1.2.2 To what extent do these accounting standards affect the components of financial statements (e.g., absolute minimum, maximum and percentile changes and percentage of firms affected)? 1.2.3 What is the nature of these adjustments? 1.2.4 How different are the results for EA from those for LA for sub-questions 1.2.1 and 1.2.2 above? 1.2.5 What are the implications for adoption timing choice (if any) of the findings for these sub-questions? 4.2.2 METHOD Following Hung & Subramanyam (2007) and Stent et al. (2010), two sets of financial statements are gathered for all sample firms: the first full-year NZ IFRS financial statements and the year prior to adoption of NZ IFRS. Prior year comparative figures, as restated under IFRS, for Total Assets, Total Liabilities, Equity, Revenue and Net Profit or Loss are extracted directly from the financial statements for the ‘IFRS Year’. The equivalent figures reported under old GAAP are extracted from the ‘Pre-IFRS Year’ financial statements. The impact of IFRS on financial statement elements is then measured as the difference between figures reported under IFRS and those reported under old GAAP (i.e., ‘IFRS differences’). Similar to prior studies (e.g., Goodwin et al., 2008; Hung & Subramanyam, 2007; Kabir et al., 2010; Stent et al., 2010), the reasons for the above IFRS differences are then investigated by analysing the reconciliations required by IFRS 1 to determine the amounts attributable to specific accounting standards. This information is extracted from the notes to the first full-year NZ IFRS financial statements. It is often the case 43 that the IFRS difference for one financial statement element (e.g., Total Assets) is due to a number of IFRS adjustments (e.g., requirements of standards relating to income taxes, financial instruments and business combinations). The analysis therefore incorporates processes which ensure that totals of adjustments per financial statement element (as extracted from IFRS reconciliations) agree to IFRS differences for financial statement elements (as extracted from the ‘IFRS Year’ and ‘Pre-IFRS Year’ financial statements). Finally, the effect of NZ IFRS on key ratios is analysed. While others (e.g., Goodwin et al., 2008; Kabir et al., 2010) also consider some ratios, the focus for this thesis is ratios which serve as proxies for variables that may influence or be of interest to analysts. Five key ratios are chosen: return on equity (net profit to equity), return on assets (net profit to total assets), leverage (total liabilities to equity), asset turnover (revenue to total assets) and return on sales (net profit to revenue).19 These ratios reflect the main ratios in the Du Pont analysis.20 The research design has the following features: first, it does not exclude firms which are unaffected by IFRS, hence providing more representative findings on the overall impact of IFRS; second, it incorporates procedures which assist in overcoming missing data in reconciliations, hence further reducing exclusions and promoting more representative findings (see 4.2.3 below); and third, it distinguishes between early and late adopters, hence the findings promote the understanding of adoption timing choices. 4.2.3 DATA NZ IFRS 1 First-Time Adoption of New Zealand Equivalents to International Financial Reporting Standards requires comparatives to be restated and reconciled in the first year of adopting NZ IFRS. Specifically, paragraphs 38-40 of NZ IFRS 1 requirements include reconciliations of equity and profit or loss, as well as sufficient detail to explain the material adjustments to the balance sheet and income statement arising from adoption of IFRS. 19 It is normal to use EBIT (earnings before interest and taxation) in estimating return on assets. However, the OLD GAAP/NZ IFRS reconciliations only provide details for net profit. Hence it is not possible to estimate EBIT. Similarly, it is not possible to estimate an Altman-type Z score model. 20 In the Du Pont analysis (Palepu et al., 2010), return on sales and asset turnover are components of return on assets. Return on assets and leverage are components of return on equity. 44 Reconciliations of sample firms are found to vary considerably in format and level of detail supplied. The most informative disclosures include presentation of full balance sheets and income statements, with reconciliations for each line item, as well as explanatory notes (see Appendix 1 example). Less informative disclosures simply note the amount of change in equity and profit or loss if any, with little or no additional explanation (see Appendix 2 example). In spite of these variations, it is important to identify which financial statement elements are impacted by NZ IFRS, the amounts involved and the accounting standards to which these impacts are attributed. The method described in 4.2.2 above explains how this is achieved and clarifies how this assists in reducing the number of firms excluded from samples due to missing data, as is necessary in many other studies (e.g., Goodwin et al., 2008; Kabir et al., 2010). 45 4.3 IFRS-RELATED DISCRETIONARY NARRATIVE DISCLOSURES 4.3.1 RESEARCH QUESTION AND HYPOTHESES Research Question 2 and related sub-questions RQ2 What do discretionary narrative disclosures in annual reports reveal about early and late adopter firms’ attitudes to IFRS adoption? 2.1 Are there significant differences between early and late adopters? 2.2 What is the nature of significant differences (if any) that are identified? The main research method used to investigate this research question involves content analysis of discretionary narrative disclosures relating to IFRS adoption for the EA and LA sample firms. This analysis is extended as follows to address the sub-questions. RQ2.1 Are there significant differences between early and late adopters? This sub-question is investigated by testing the null hypothesis (EA = LA) for both the nature and the extent of these discretionary narrative disclosures, resulting in the following hypothesis and sub-hypotheses: H03: There are no significant differences in the extent (ED) and nature (ND) of discretionary narrative disclosures about IFRS in annual reports for Early Adopters (EA) and Late Adopters (LA) i.e., H03a µED of EA = µED of LA’s H03b: µND of EA = µND of LA; Sub-hypotheses 3a and 3b are tested using the Mann-Whitney U test. As univariate tests do not control for other firm specific factors that might impact disclosure, a multivariate model is also developed to examine the impact of early adoption on the nature and extent of discretionary narrative disclosures: DISji = β0 + β1 EAi+ β2 SIZEi+ β3 AUDi + β4 OSHIPi + β5 ORIGINi + β6-10 INDki (1) 46 Development of this model and descriptions of the component variables are described in more detail in the section below (4.3.2). RQ2.2 What is the nature of significant differences (if any) that are identified? Significant differences between results for EA and LA are described with reference to the original extracts from firms’ annual reports. As suggested by Creswell (2003), quotations are used to convey information in the voice of the sample firms and to add to the richness of descriptions. 4.3.2 METHOD AND DATA Much of the literature on the motivations for, and effects of, adopting IFRS is restricted to quantitative analysis of financial statements and capital markets data. This part of the thesis extends the analysis beyond the financial statements to the discretionary narrative sections of the annual report. Narrative disclosures offer a broader perspective, as well as the opportunity to triangulate the findings with prior literature. In addition, this thesis extends the accounting choice literature by considering differences in narrative disclosures between early (voluntary) and late adopters of IFRS. Narrative disclosures are an important part of ‘quality’ corporate reporting as they offer insights and explanations about quantitative measures (Beattie et al., 2004; Beretta & Bozzolan, 2004; Clatworthy & Jones, 2003; Smith & Taffler, 2000). In particular, the ‘front half’ of the corporate annual report provides a tool for management to signal its response to IFRS (Stanton & Stanton, 2002). This thesis examines governing bodies’ perceptions of the importance of IFRS adoption, as reflected in the narrative disclosures in the annual report. The IFRS adoption literature closest to this thesis considers the perceived usefulness of IFRS from the preparers’ perspective based on survey results (e.g., Jermakowicz et al., 2007; Jones & Higgins, 2006). Jermakowicz et al. (2007) survey company executives of DAX-30 firms in Germany and find that most agree that IFRS should improve comparability of financial statements, but that complexity, cost, lack of guidance and increased volatility of earnings are important ‘challenges’. Jones & Higgins (2006) report, on the basis of a telephone survey, that account preparers in Australia are generally very sceptical about the claimed benefits of IFRS, as enunciated in the 47 Corporate Law Economic Reform Program. Content analysis is applied to the nature and extent of disclosures in annual reports of first time adopters of IFRS in New Zealand from 2005 through 2008. The focus is on discretionary narrative information in the annual report (i.e., information other than the financial statements). Creswell (2003) notes that an advantage of analysing documents (as opposed to performing interviews or other methods of data collection) is that documents represent data that participants have taken care and thought to compile. Thus, the annual report reflects views of the governing body with regard to the type, amount and importance of information disclosed (Niskala & Pretes, 1995). Another distinguishing feature is that content analysis is unobtrusive, as the documents can be evaluated without the communicators’ knowledge (Jones & Shoemaker, 1994). In a review of 68 studies that use content analysis in accounting research, Jones & Shoemaker (1994) conclude that the primary strength of thematic analysis is its ability to identify the motivations and concerns of accounting communicators. Analysis of IFRS disclosures in annual reports is an appropriate means of extending existing research regarding the importance and impact of IFRS. This thesis draws on the framework developed by Smith & Taffler (2000), to classify content characteristics for measurement of apparent and connotative meaning. The framework is used as a basis: (1) to determine the attitudes of governing bodies towards the adoption of IFRS (including extent of related disclosures); (2) to identify and classify the main themes communicated by governing bodies regarding the adoption of IFRS; and (3) to observe differences between early adopters and late adopters in the attitudes and themes communicated. Table 1 reports the framework as adapted for this thesis (Panel A) and also describes the original Smith & Taffler (2000) framework (Panel B). 48 Table 1 Framework for content analysis Panel A: Theme Categories adapted for IFRS Analysis (This thesis) Category Classification Importance High (prominence/high-emphasis/achievement) Low (contextual/low emphasis) Potency High (specific or significant impact ) Low (vague or minimal impact) Evaluative Beneficial (positive consequences) Adverse (negative consequences) Activity N/A N/A Manageability Expected Unexpected Panel B: Theme Categories (Smith and Taffler, 2000) Category Classification Evaluative Beneficial (positive achievement) Adverse (negative occurrence) Potency Tangible (degree of certainty) Intangible (vagueness) Activity Dynamic (measure of performance) Static (reluctant action) Manageability Expected (status quo) Unexpected (external factors) N/A = activity (i.e., early or late adoption) is used as a firm characteristic rather than as a reporting theme. Based on Smith & Taffler (2000) five broad reporting themes are developed. The first theme relates to whether the discretionary narrative sections of the annual report indicate the adoption of IFRS as being of high or low ‘importance’. This theme was not used in Smith & Taffler (2000), as it is specific to this thesis. The second theme indicates whether descriptions of the impact of IFRS have high ‘potency’ (i.e., descriptions make specific references to affected financial statement elements or to the extent of significant changes) or low ‘potency’ (i.e., descriptions are vague or indicate insignificant changes). The third, or ‘evaluative’, theme classifies the consequences of IFRS adoption as beneficial or adverse. The ‘activity’ theme in Smith & Taffler (2000) corresponds to whether the firm was active or static. In terms of IFRS adoption, this best corresponds to the distinction between early or late adopting firms. Hence, in this thesis, the activity dimension is a firm characteristic rather than a content analysis theme. Early or late adoption of IFRS is also built into the research design, as the content themes are compared along this dimension. The fifth or ‘manageability’ theme corresponds to whether the discretionary narrative sections of the annual report describe the impact of IFRS as expected or unexpected by management. 49 Based on the studies discussed in Chapter 3, the following expectations are developed for this thesis: early adopters of IFRS (relative to late adopters) are expected to consider adoption to be more important (theme 1), to be more beneficial (theme 3) and the impact to be more expected and hence manageable (theme 5). With regard to potency (theme 2) some firms may choose to early adopt because the impact is immaterial, rather than important. For firms where the impact of IFRS is material early adopters are expected to be more specific about the impact of the adoption (i.e., the narrative will have high potency). Data is hand-collected for the first year that full-IFRS financial statements are issued. Discretionary narrative sections (i.e., the non-financial statement sections) of the annual report are searched for the terms ‘IFRS’ or ‘International Financial Reporting Standards’.21 Where either of these terms is included in a sentence, the entire sentence is extracted to ensure that the reference is analysed in its correct context (Holsti, 1969; Steenkamp & Northcott, 2007). The sentences before and after an extract are also read to assess for relevance to IFRS. The extracted sentences are recorded and coded using NVivo to increase the reliability and validity of the analysis (see Boesso & Kumar, 2007; Krippendorff, 2004 for more detail on these issues).22 ‘Form oriented’ analysis or word counts are often used to provide a measure of the importance of, or attention devoted to, a concept or attribute (e.g., Holsti, 1969; Krippendorff, 2004; Steenkamp & Northcott, 2007). Hence, the initial analysis uses four count measures to assess the extent of IFRS adoption disclosures found in the narrative extracts: the number of IFRS mentions per firm; the number of words; the space devoted relative to one page of the annual report; and the space devoted relative to all pages in the discretionary narrative sections of the annual report. Mann-Whitney U tests are used to assess the statistical differences between EA firms and LA firms in each case. The ‘form oriented’ dimension of content analysis is generally accepted as being less informative than the ‘meaning oriented’ dimension (Krippendorff, 2004; Steenkamp & Northcott, 2007). Hence, the IFRS extracts are also analysed in accordance with the themes identified in Table 1. 21 Reviews of annual reports revealed no further potential search terms. 22 Coding decision rules were developed by two independent researchers and then clearly defined using the tools available within NVivo. All coding was then performed by one researcher in accordance with these decision rules. Coding by this researcher was test checked by the second researcher to confirm accuracy and consistency of coding. 50 In identifying the themes, annual reports of a set of early adopters are reviewed to develop an initial list of common themes. This initial list confirmed that little adaptation was necessary to the Smith & Taffler (2000) framework. Descriptions of the themes and illustrative examples are provided in Appendix 3. The process used allows for the possibilities that multiple themes may be present within one sentence and that a number of sentences may be associated with a single theme (Holsti, 1969; Krippendorff, 2004; Steenkamp & Northcott, 2007). That is, the coding approach does not assume that context units can be classified into mutually exclusive themes. Three main categories for themes emerge, namely ‘Importance’, ‘Potency’ and ‘Evaluative’. To test whether there are statistically significant differences in the nature of voluntary narrative disclosures of EA and LA firms, a score is derived for each of these theme categories for each sample firm through simple weighting schemes: • An ‘importance’ score is constructed for each firm using the ordinal nature of the sub-themes within the importance themes.23 Each firm is allocated a score as follows: ‘context’ mentions = 1; ‘highlight’ mentions = 2; ‘informing of importance’ mentions = 3. Hence there is a possible maximum score of 6 for each firm; • The potency theme is divided into two categories. ’High’ potency’ relates to discretionary narrative which is specific as to the nature and extent of IFRS impacts. There are sub-themes in this category for the financial statement elements that are described as being materially affected (e.g., assets, liabilities, equity, revenue, expenses or disclosure). The ‘Low’ potency category has two sub-themes: (1) where the discretionary narrative comments on a limited or immaterial impact, and (2) where the impact description is general or vague. A ‘high potency’ as well as a ‘low potency’ index is derived for each firm. ‘High potency’ sub-themes score 1 point for each specific reference to a financial statement element (i.e., a maximum possible score of 6 per firm). ‘Low potency’ themes are scored: ‘limited or immaterial impact’ = 2 and ‘general’ = 1 (i.e., a maximum possible score of 3 per firm). Similar to the high potency score, the low potency score emphasizes the specific nature of comments. The low potency score reflects the view that a ‘general’ comment is less specific than a ‘limited or immaterial’ comment. One weakness of the high potency score is 23 Low frequencies for many sub-themes mean that Chi-square tests cannot be used for testing at this level. The sub-themes are therefore aggregated and a disclosure index created to statistically test the differences between early and late adopters. 51 that a firm, where the effect of IFRS is widespread, (i.e., where the impact affects assets, liabilities, revenues and expenses) can have a higher ‘high potency’ score relative to a firm where the impact of IFRS is significant but isolated just to one area (e.g., assets). A combined potency score is therefore also developed by taking the ‘low potency’ score and adding 1 if the ‘high potency’ score is non-zero. • Evaluative themes are classified into positive (beneficial), negative (adverse), and neutral sub-themes. The ‘evaluative’ index for each firm is constructed by allocating 1 for ‘positive’ or ‘neutral’ comments, zero for no comment and -1 for ‘negative’ comments. The derived scale variables are used for Mann-Whitney U Tests of significant differences between EA and LA. The above univariate tests do not control for other firm specific factors that might impact disclosure. A multivariate model is therefore also developed to examine the impact of early adoption on the nature and extent of discretionary narrative disclosures: DISji = β0 + β1 EAi+ β2 SIZEi+ β3 AUDi + β4 OSHIPi + β5 ORIGINi + β6-10 INDki (1) Where: DISji is the disclosure score for disclosure index j for firm i. Where disclosure index j is related to themes (e.g., importance; potency); EAi is 1 if the firm is an early adopter and 0 if it is a late adopter; SIZEi represents firm size and is measured as the natural log of the total assets for firm i; AUDi is 1 if the firm has a Big 4 auditor and 0 otherwise; OSHIPi is the largest shareholding percentage ownership, ORIGINi is if the firm has Australian and European shareholding, and INDki is 1 if the firm is in industry k and 0 otherwise. The concern addressed is whether early adoption (EA) impacts narrative disclosures. Therefore the coefficient on this variable is expected to be positive and significant. The remaining variables in model 1 are to control for other factors that might influence a firm’s IFRS disclosures. A parsimonious model is chosen because the sample size is relatively low (i.e., n=80). In line with prior literature, larger firms are expected to provide more extensive IFRS disclosures (e.g., Palmer, 2008). Firm size is measured 52 using total assets and logs are taken to normalise the distribution. IFRS disclosures have been found to be related to whether the firm employs a Big 4 auditor (Kent & Stewart, 2008; Palmer, 2008). Hickey et al. (2003) suggest that many of the early adopters may be influenced by parent companies or subsidiary holdings in Australia or Europe. An ownership variable (i.e., the percentage of the largest shareholding) and an origin variable (i.e., =1 if the firm was listed in Australia or Europe) is therefore included in the multivariate model. Industry indicator variables are used to control for the differential impact of IFRS on industry sectors. Firms are allocated to one of six industries: energy, goods, investment, services, property and primary. Five industry variables are used, leaving the largest industry (services) as the base sector. 53 4.4 COSTS AND BENEFITS OF IFRS ADOPTION 4.4.1 RESEARCH QUESTIONS AND HYPOTHESES The collection of survey data allows extension of the earlier research questions and hypotheses, as follows: Research Question 3 and related sub-questions RQ 3 What are the costs and benefits of IFRS adoption for early as opposed to late adopters? 3.1 Are there significant differences between early and late adopters? 3.2 What is the nature of significant differences (if any) that are identified? RQ 3.1 Are there significant differences between early and late adopters? This question is addressed by testing the null hypothesis (EA = LA) for each area of cost and benefit information surveyed, resulting in the following hypothesis and sub- hypotheses: H04: There are no significant differences in costs or benefits arising from adoption of IFRS, for EA and LA i.e., H04a: µCosts of IFRS EA = µCosts of IFRS LA; H04b: µBenefits of IFRS EA = µBenefits of IFRS LA; Sub-hypotheses 4a and b are tested using the Mann-Whitney U test. RQ 3.2 What is the nature of significant differences (if any) that are identified? Significant differences between results for EA and LA are described with reference to any comments added to questionnaires and to information obtained during the follow-up interviews. This qualitative information is also analysed to identify any other adoption timing insights which may emerge. 54 4.4.2 METHOD AND DATA The previous two parts of this chapter concern the examination of publicly available data to assist in determining the influences affecting IFRS adoption timing choices. This part concerns information obtained by means of questionnaires and semi-structured interviews, which is not found in publicly available documents. Chapter 2 includes a review of the motivations for adopting IFRS in New Zealand. This review reveals a persistent underlying conviction by regulators and standard setters that the benefits of adopting IFRS should exceed the costs of doing so. Similar sentiments are likely to have influenced firms’ choices of adoption timing. The costs and benefits reviewed in Chapter 2 are therefore used as a basis for constructing a questionnaire and for providing structure for the semi-structured interviews. The survey questionnaire was sent to senior financial executives (e.g., Chief Financial Officer, Financial Controller, Finance Manager, Group Accountant) of each sample firm.24 The survey includes questions designed to collect evidence relating to cost and level of effort involved in implementing IFRS, as well as respondent opinions regarding expected benefits of adopting IFRS. Introductory questions related to the respondent’s position and responsibilities and the firm’s main shareholder are also included, as well as a question regarding the impact of IFRS on the firm’s contracts. The questionnaire was sent to early adopters over the period July 2007 to August 2008 and to late adopters in March 2011. Questionnaires were followed up by telephone or face-to-face interviews to clarify responses where necessary and to obtain more detailed answers to open-ended questions.25 Care was taken to identify and approach senior financial executives within the sample firms who had key roles in overseeing the adoption of IFRS. These individuals were then contacted by telephone to explain the research project and to request their assistance. A questionnaire, together with a covering letter bearing the researcher’s original signature, was then sent to those who agreed to assist. These survey instruments were usually sent by both physical mail and email to allow respondents more flexibility and convenience and hence to assist in improving the response rate. Where 24 The questionnaire and a sample covering letter are attached as Appendices 4 and 5. 25 Face-to-face interviews were preferred, but telephone interviews were used where respondents were in remote locations or where this improved the likelihood of securing an interview. 55 questionnaires had not been returned by respondents, they were politely reminded by telephone call or email approximately every two weeks (up to three reminders). Two potential issues in data collection are identified, namely response equivalence and timing of data collection (see Cavana, Delahaye, & Sekaran, 2001 for more detail in this regard). Steps taken to promote response equivalence included conducting all survey procedures personally (i.e., research assistants were not used to collect survey data); use of standardised procedures and explanations for introducing the research project to respondents; and use of a standardised questionnaire as well as a standardised survey instrument for semi-structured interviews. Timing of data collection was necessarily spread over a relatively long period for early adopters (approximately one year). Data collection for late adopters took place approximately two and a half years later than for early adopters. It is therefore possible that results could be affected by time-related changes (e.g., changes to IFRS standards have been ongoing since 2005 when the first early adopters began implementing IFRS; changes in personnel responsible for IFRS implementation prior to data collection). Timing of data collection is therefore recognised as a potential limitation of this study. The questionnaire and covering letter were checked and pretested as follows: (1) Design was checked and amended to minimise bias and achieve conformity with accepted principles of wording, measurement and general appearance (e.g., Cavana et al., 2001; Creswell, 2003). Closed questions were used throughout the questionnaire to make completion easier and quicker for respondents and to facilitate coding for subsequent analysis. However, each question was also supplemented with space for comments to provide an open-ended option for respondents, in case they wished to elaborate or qualify their answers in any way. Likert scales were used in two questions where respondents were asked to indicate their opinions; (2) Pre-testing for face and content validity was conducted by five academics and by four senior financial executives (i.e., similar to the intended respondents). Issues raised were discussed with the pre-testers and amendments made where necessary; and (3) The questionnaire was included in project material reviewed and approved by the Massey University Human Ethics Committee. 56 An adapted version of the questionnaire was used for the follow-up interviews.26 This adapted version shows potential interview prompts (optional depending on completion of the original questionnaire) and an additional final open-ended question relating to the respondent’s motivations for their adoption timing choice. All interviews were recorded with the permission of respondents and transcribed. Both the recorded and transcribed versions have been retained as evidence. The questionnaire sought to measure information regarding ‘Costs’ in terms of staff, time and financial commitments as well as in terms of the impact of IFRS on the firms’ contracts (e.g., costs of and need to renegotiate contracts): • Two measures of staff commitments are used: (1) the number of staff initially appointed to manage the implementation of IFRS; (2) the maximum number of staff responsible for managing the implementation of IFRS at any one time; • Two measures of time commitments are used: (1) the number of months prior to date of transition to IFRS that formal planning began for implementation of IFRS; (2) the firm’s number of person hours spent implementing IFRS, deflated by the firm’s average annual costs of preparing financial accounts;27 • Financial commitment is measured as the firm’s NZ dollar cost of implementing IFRS, deflated by the firm’s average annual cost of preparing financial accounts;28 • The impact of IFRS on firm contracts is measured by constructing an ‘impact’ score for each firm using respondents’ opinions. Respondents were asked to indicate how much of an impact IFRS had on four categories of contracts. The four categories of contracts were debt, revenue, expenses and other; impact rankings were indicated using a five point Likert scale (1 = ‘no impact’ through to 5 = ‘significant impact’). The ‘impact’ score was calculated as the sum of rankings across all contract categories. In addition to costs, the questionnaire sought to measure ‘Benefits’ by obtaining opinions of respondents in respect of nine benefits that are expected to arise as a result of the adoption of IFRS. Opinions were indicated using a 5 point Likert scale (1 = ‘strongly agree’ through to 5 = ‘strongly disagree’). Opinions are considered 26 The adapted questionnaire which was used as a survey instrument for the semi-structured interviews is attached as Appendix 6. 27 Sensitivity testing is also conducted using firstly Total Assets and secondly Audit Fees as deflation factors. 28 Sensitivity testing is conducted as for time commitments. 57 individually, as well as on an overall basis by constructing an ‘overall benefits’ score for each firm. The ‘overall benefits’ score is calculated as the sum of the rankings across all nine expected benefits. The result is divided by nine to arrive at an average score which can be interpreted using the Likert scale described above. Finally, interviews were concluded with the open-ended question reproduced below in order to probe for ‘other adoption timing insights’:29 “Adoption of IFRS was possible from 1 January 2005 (on a voluntary basis); mandatory from 1 January 2007. Your organisation elected to adopt early/ wait until it became mandatory. Could you give me your insights/background as to why early/ late?” The method for identifying themes and the framework used to analyse these themes is broadly consistent with that used for the content analysis of annual report disclosures (see Chapter 4.3.2). However, direct personal interaction with the respondents and the specific nature of the above question result in some adaptations and developments, so a brief summary of the adapted framework is presented below: • In addition to the three main themes which emerged in the earlier content analysis (i.e., ‘importance, ‘potency’ and ‘evaluative’), three additional themes are identified. The first two of these additional themes correspond to those identified by Smith & Taffler (2000), namely ‘activity’ and ‘ manageability’. The final theme which emerges is identified as ‘accounting choice’ as it highlights managements’ intentions with regard to adoption timing decisions, namely whether “…to affect the output of the accounting system or whether the impetus derives from other motives” (Fields et al., 2001, p. 256); • Use of the ‘evaluative’ theme remains unchanged. There are three sub-themes: ‘neutral’, ‘negative’ and ‘positive’; • Use of the ‘importance’ and ‘potency’ themes are simplified – in each case, only ‘high’ or ‘low’ emerge as sub-themes (various other sub-themes emerged in the earlier analysis, particularly for the ‘potency’ theme). • Earlier analysis in this thesis has differentiated the ‘activity’ dimension on the basis of action (i.e., early adopters were classified as ‘dynamic’ and late 29 First IFRS financial statements of interviewee firms were analysed before interviews. If interviewees were not forthcoming about whether material IFRS differences had influenced adoption timing, they were prompted in this regard (note that analysis of overall results was not complete prior to the conduct of interviews). 58 adopters as ‘static’). Interview responses to the above question however provide deeper insights into the thought processes behind the decisions to act. Analysis of ‘activity’ can therefore be expanded to more accurately assess whether firms were dynamic or static in relation to implementation of IFRS (e.g., where early adoption by a firm is simply due to the requirements of an Australian parent, this firm should be reclassified as ‘static’ rather than ‘dynamic’). • Use of the ‘manageability’ theme remains unchanged. There are two sub- themes: ‘expected’ and unexpected’. • ‘accounting choice’ is a new theme which emerges from interview responses. The sub-theme ‘yes’ indicates that managements’ intentions with regard to adoption timing decisions were to affect the output of the accounting system (e.g., early adoption due to income increasing effects). The sub-theme ‘no’ indicates that the impetus for the adoption timing decision derives from motives unrelated to accounting information (e.g., late adoption due to time constraints). 4.5 SAMPLE SELECTION Table 2 reports the outcome of the sample selection procedures. 161 companies are listed on the New Zealand Stock Exchange (NZX) on 1 March 2007. These are then stratified into early and late adopters of NZ IFRS. Early adopters (EA) are the reporting entities that chose to adopt NZ IFRS for periods beginning on or after 1 January 2005, but before it became mandatory (periods beginning on or after 1 January 2007). Selection as an EA is only made once the “… explicit and unreserved statement of compliance …” with NZ IFRS has been sighted in a first full-year set of financial statements.30 These procedures identify an initial set of 48 EA. The remaining total of 113 ‘non-EA’ forms the initial population of Late Adopters (LA). These are reporting entities listed on the NZX, which chose to wait until periods beginning on or after 1 January 2007, when it became mandatory to adopt NZ IFRS. 12 observations are lost where companies delisted from the NZX after 1 March 2007, resulting in a reduced population of 101 LA. Status as LA are confirmed by 30 This statement is required in terms of NZ IFRS 1 First Time Adoption of New Zealand Equivalents to International Financial Reporting Standards. 59 ensuring that the “… explicit and unreserved statement of compliance …” with NZ IFRS appears for the first time in a first full-year set of financial statements beginning on or after 1 January 2007. For the 48 EA (Panel B), observations are discarded because the entity uses GAAP other than NZ IFRS (4 observations); uses a functional currency other than NZ dollars (2); has no prior year financial statements available as the first year of listing on the NZX is also first year of application of NZ IFRS (1); or has no reconciliation to OLD GAAP, because NZ IFRS was adopted from its first year of operation (1). A total sample of 40 EA remains. Table 2 Effect of sample selection criteria Panel A: Late adopters    Total number of companies listed on NZX at 1/3/2007 161 Early adopters (see Panel B) -48 Total late adopters 113 Companies that delisted after 1/3/2007 -12 Population of late adopters 101 Panel B: Early adopters    Total early adopters 48 GAAP other than NZ IFRS -4 Currency other than NZ dollar -2 No prior-year financial statements -1 No NZ IFRS reconciliation -1 Population of early adopters 40 A random sample of 40 LA is then selected from the diminished population of 101 late adopters as a control group. The total sample therefore comprises 80 (50%) of the original 161 companies.31 The methodologies described in the preceding sections were applied to all sample companies, with one exception: three Christchurch companies in our LA sample were excluded from the survey procedures out of sensitivity to their situation and priorities after the first earthquake in 2010. 31 Boesso & Kumar (2007) suggest that a relatively small sample size is fairly common in studies using content analysis and refer to a number of studies with samples of 33 and fewer subjects. They note that the richness of data collected makes it a labour-intensive and time consuming methodology. 60 4.6 CHAPTER SUMMARY This chapter explains the research design used to investigate the nature and extent of significant differences between early adopters and late adopters of IFRS. Three research questions are stated involving three different research methods to examine three distinct data sets. These research questions are developed and descriptions provided of the methods and data sets relevant to each of them. The chapter ends with a description of the sample selection process. First, the impact of IFRS adoption on financial statements and some common financial ratios is quantified. Second, content analysis is applied to investigate the extent and nature of IFRS related discretionary narrative disclosures in annual reports. Third, senior financial executives are surveyed to assess the costs and benefits arising from IFRS adoption. In each case, results for early adopters are compared to those of late adopters. Mann-Whitney U tests are used to establish whether differences are statistically significant. A multivariate test is also used to control for firm specific factors that might impact disclosures. Qualitative research methods are then used to investigate significant differences in more depth. The research described above is applied to a total sample of 80 companies listed on the New Zealand Stock Exchange (50% of the original population of 161 companies). Sample selection criteria result in the selection of all early adopters, except for 8 which must be discarded for specific reasons, leaving a sub-sample of 40 early adopters. A random selection of 40 late adopters is then made as a control group. The research findings are presented in the chapter which follows, beginning with the results of the investigation into the impact of IFRS adoption on financial information. 61 CHAPTER 5: RESULTS 5.1 INTRODUCTION This chapter presents the research findings for the investigation of the three research questions developed in the previous chapter. These findings are presented in three parts (5.2 – 5.4). Part 5.2 relates to the first research question. Descriptive statistics are presented to quantify the impact of IFRS adoption on the financial statements and key ratios of early and late adopters. In particular, this includes: • the impact of IFRS adoption on the financial statement elements, such as assets and liabilities (5.2.2); • the impact of IFRS adoption analysed by the accounting standards concerned, such as NZ IAS 12 Income Taxes (5.2.3); • the effect of IFRS adoption on key ratios, such as return on equity (5.2.4). Part 5.3 relates to the second research question. Results are presented for the content analysis of discretionary narrative disclosures concerning IFRS adoption. These disclosures are extracted from annual reports issued in the first year of IFRS adoption. Findings are presented for form oriented analysis (count data) as well as for meaning oriented analysis (themes). These findings include the results of procedures which control for firm specific factors associated with the nature and extent of IFRS disclosures (e.g., industry, firm size and auditor influence). Part 5.4 relates to the third research question. Results of a survey are presented to provide evidence relating to the costs and benefits of IFRS adoption, as well as further insights relating to adoption timing motives. 62 5.2 IMPACT OF IFRS ADOPTION ON FINANCIAL INFORMATION 5.2.1 INTRODUCTION The results for this part of the chapter (5.2) relate to research question 1: RQ1 What impact does IFRS adoption have on the financial information of early as opposed to late adopters? These results are discussed in three sub-sections. The first sub-section (5.2.2) is concerned with differences in dollar values of financial statement elements as a result of IFRS adoption and the difference in these differences for EA relative to LA. The second sub-section (5.2.3) is concerned with the source and nature of the differences identified in 5.2.2. The third sub-section (5.2.4) is concerned with differences in key financial ratios which flow from IFRS adoption and the difference in these differences for EA relative to LA. 5.2.2 IMPACT OF IFRS ON FINANCIAL STATEMENT ELEMENTS The results for this section (5.2.2) relate to research question 1.1: RQ1.1 Are there significant differences between early and late adopters (with respect to the impact of IFRS on financial information)? Descriptive statistics of the impact of IFRS on the financial statement elements are presented in Tables 3 to 5. Table 3 reports descriptive statistics of the sample firms. In this table pre-IFRS GAAP figures (i.e., ‘OLD GAAP’) are reported as this is the base from which the differences due to the adoption of IFRS are measured.32 The sample data are not normally distributed and for the most part are leptokurtic and positively skewed. This results in the median being a better indicator of central tendency than the mean. Non-parametric tests are therefore relied upon to analyse statistical differences in the sample data. 32 Table 3 figures reported under ‘OLD GAAP’ are extracted from ‘Pre-IFRS Year’ financial statements. Figures in subsequent tables are extracted mainly from the OLD GAAP/IFRS reconciliations which are required under IFRS 1 in the financial statements for the year in which IFRS is first adopted. 63 Table 3 Descriptive statistics of sample under OLD GAAP ($000) Total Assets Total Liabilities Total Equity Total Revenue Net Profit Panel A: Early adopters (n = 40) Mean 1,700,000 1,350,000 355,143 515,250 66,442 Std. Deviation 6,222,000 5,807,000 703,101 1,145,000 161,647 Minimum 6 152 -146 0 -1,250 25 Percentile 28,458 5,235 21,859 20,165 1,288 Median 137,538 64,858 94,546 134,156 9,668 75 Percentile 601,369 251,611 292,764 476,559 40,072 Maximum 38,798,900 36,527,700 3,034,884 5,759,000 916,000 Panel B: Late adopters (n = 40) Mean 609,733 326,292 283,959 331,630 24,508 Std. Deviation 1,370,000 839,737 574,981 775,538 45,469 Minimum 164 12 -95 6 -9,543 25 Percentile 9,322 671 6,185 4,002 -178 Median 38,743 18,908 23,954 34,688 2,736 75 Percentile 307,671 124,216 170,672 262,778 35,845 Maximum 5,728,860 3,825,819 1,934,471 4,297,000 214,000 The median total assets figure for EA is $137.5 million (LA: $38.7 million), median total equity for EA is $94.5 million (LA: $23.9 million) and median net profit for EA is $9.7 million (LA: $2.7 million). These initial observations are consistent with prior literature (e.g. Iatridis & Joseph, 2006) which suggests that EA tend to be larger than LA. Minimums are, however often lower for EA than for LA, indicating that size is not the only important characteristic. Panels A and B of Table 4 report the changes in financial statement elements due to the adoption of IFRS for early and late adopters respectively. The magnitude of change in a particular financial statement element is measured as: IFRS/ OLD GAAP - 1 For example, the mean (median) change in total assets due to the adoption of IFRS for EA is 0.051 (0.010) and for LA is 0.057 (0.002). 64 Table 4 Impact of IFRS on financial statement elements Total Assets Total Liabilities Total Equity Total Revenue Net Profit Panel A: Early adopters (n = 40) Magnitude of change a Mean 0.051 0.089 0.096 0.008 -0.922 Std. Deviation 0.441 0.281 1.222 0.190 4.384 Minimum -0.640 -0.190 -1.120 -0.600 -26.740 25 Percentile 0.000 0.000 -0.061 -0.009 -0.017 Median 0.010 0.010 -0.001 0.000 0.043 75 Percentile 0.019 0.073 0.018 0.007 0.140 Maximum 2.610 1.590 7.430 0.600 1.170 Direction of change Increase 67.5% 75.0% 45.0% 30.0% 70.0% Decrease 25.0% 10.0% 50.0% 45.0% 25.0% No change 7.5% 15.0% 5.0% 25.0% 5.0% Statistical tests b Z statistic 2.693 4.112 0.877 0.648 2.647 p-value (two tailed) 0.007 0.000 0.380 0.517 0.008 Panel B: Late adopters (n = 40) Magnitude of change a Mean 0.057 0.238 -0.040 0.170 0.169 Std. Deviation 0.218 0.810 0.662 0.840 0.513 Minimum -0.060 0.000 -3.440 -0.990 -0.930 25 Percentile 0.000 0.000 -0.044 -0.025 -0.009 Median 0.002 0.035 -0.004 0.000 0.002 75 Percentile 0.023 0.180 0.000 0.016 0.182 Maximum 1.230 5.130 2.130 3.950 1.930 Direction of change Increase 52.5% 72.5% 27.5% 42.5% 52.5% Decrease 25.0% 2.5% 57.5% 40.0% 32.5% No change 22.5% 25.0% 15.0% 17.5% 15.0% Statistical tests b Z statistic 2.489 4.762 1.838 0.179 1.915 p-value (two tailed) 0.013 0.000 0.066 0.858 0.056 a = The change is estimated as (IFRS/OLD GAAP)-1 b = The reported test statistic is a Wilcoxon test for equality of matched pairs 65 Table 4 (cont.) Impact of IFRS on financial statement elements Total Assets Total Liabilities Total Equity Total Revenue Net Profit Panel C: Mann-Whitney U tests c Mean Rank Early 41.71 36.42 41.25 39.78 39.92 Late 39.29 44.58 39.75 41.23 41.08 Z statistic 0.467 1.575 0.289 0.280 0.221 p-value (two tailed) 0.640 0.115 0.773 0.779 0.825 c = The Mann-Whitney U test reports significance of differences in impact between EA and LA The direction of change reports the percentage of firms experiencing decreases, increases and no change. For EA, the adoption of NZ IFRS results in an increase in assets for 67.5 percent (27/40) of the observations, a decrease for 25 percent and 7.5 percent remain unchanged (for LA, an increase for 52.5 percent, a decrease for 25 percent and 22.5 percent remain unchanged). Statistical tests report Wilcoxon tests for difference in the distribution of a variable for matched pairs. That is, each firm is compared with itself for differences between OLD GAAP and IFRS. The general increase in total assets due to the adoption of IFRS is statistically significant for EA at the 0.01 level (for LA at the 0.05 level). Finally, Panel C reports the results of Mann- Whitney U tests to establish whether the difference in IFRS adoption changes for EA and LA are statistically significant. Table 4 reveals that the extent of change is similar for early and late adopters of IFRS. The largest impact of IFRS is for liabilities, where 75 percent of EA observations (LA: 72.5 percent) report an increase in liabilities and only 10 percent (LA: 2.5 percent) report a decrease. The impact of IFRS is widespread as 95 percent of EA firms are affected (LA: 85 percent). That is, only 5 percent of EA firms have no changes to equity or net profit (LA: 15 percent). Overall the impact of IFRS significantly increases assets, liabilities and net profit for both EA and LA firms. The impact on equity is not significant at conventional levels for EA firms, but LA firms show a decrease in equity which is weakly significant (at the 0.1 level). The impact on revenue is not significant at conventional levels for either EA or LA firms. The inter-quartile range for most elements is small indicating that for most firms the impact of IFRS is small. However, the maximum and minimum values indicate that the effect of IFRS can be quite 66 substantial for some firms. For example, the inter-quartile range for net profit for EA firms is 0.157 (0.140 – (-0.017)) and the range is 27.910 (1.170 – (-26.740)). The inter- quartile range for net profit for LA firms is 0.191 (0.182 – (-0.009)) and the range is 2.860 (1.930 – (-0.930)). Some differences between EA and LA firms are apparent from panels A and B. First, EA firms are more widely affected than LA firms (see above). Second, the increase in liabilities is less pronounced for EA than for LA firms (see all minimum, maximum, percentile and Z statistic figures). Third, the income-increasing effect of IFRS has a larger and more strongly significant effect on net profit for EA than for LA firms (medians of 0.043 and 0.002 respectively; p-values of 0.008 and 0.056 respectively). These findings are consistent with the debt and bonus hypotheses, but Panel C reveals that differences between EA and LA firms regarding the impact of IFRS are not significant at conventional levels for any of the financial statement elements.33 To conclude, the hypotheses for research question 1.1 are re-stated and tested against the above results: H01: There are no significant differences in financial information (FI) prepared under IFRS and financial information prepared under NZ GAAP, for Early Adopters or for Late Adopters i.e., H01a: µFI of EA (IFRS) = µFI of EA (NZ GAAP); H01b µFI of LA (IFRS) = µFI of LA (NZ GAAP) HO2: Any change in financial information for Early Adopters (∆FI for EA) as a result of the adoption of IFRS will not be significantly different from the change in financial information for Late Adopters (∆FI for LA) i.e., H02: µ∆FI of EA = µ∆FI of LA Hypotheses 1a and b are rejected, because significant differences are found in most financial statement elements for both EA and LA firms as a result of NZ IFRS adoption (exceptions are revenue for EA and LA firms and equity for EA firms). However, Hypothesis 2 is accepted as the change in financial statement elements for EA is not found to be significantly different to the change in financial statement elements for LA. 33 Stent et al. (2010), find contrasting results, consistent with early adopting firms being those firms on which IFRS have a lower impact. This is likely due to use of a smaller sample of only 16 early adopters whereas the results for this thesis are based on a sample of 40 early adopters. 67 These findings are consistent with those of Kabir et al. (2010). 5.2.3 IMPACT OF IFRS ANALYSED BY ACCOUNTING STANDARDS The results for this section (5.2.3) relate to research question 1.2 and its related sub- questions: RQ1.2 What is the nature of significant differences (if any) that are identified? 1.2.1 Which accounting standards are the most frequent sources of IFRS adjustments (e.g., NZ IAS 12 Income Taxes)? 1.2.2 To what extent do these accounting standards affect the components of financial statements (e.g., absolute minimum, maximum and percentile changes and percentage of firms affected)? 1.2.3 What is the nature of these adjustments? 1.2.4 How different are the results for EA from those for LA for sub-questions 1.2.1. and 1.2.2 above? 1.2.5 What are the implications for adoption timing choice (if any) of the findings for these sub-questions? Accounting standards most frequently causing IFRS adjustments and their effects on financial statements Table 5 reports the impact of the adoption of IFRS analysed by the accounting standard that caused the accounting change. Panels A and B report results for early and late adopters respectively.34 These results have been sorted to show which standards most frequently affect firms for each financial statement element (i.e., from lowest to highest ‘no change’ percentages). 34 The statistical significance of IFRS adoption changes for each line of Table 5 is not reported due to limitations in data (e.g., figures reported under OLD GAAP compared to those reported under IFRS may be due to more than one standard; and insufficient data in many reconciliations). 68 Table 5 Impact of NZ IFRS by financial statement element and accounting standard Minimum 25th Percentile Median 75th Percentile Maximum % Increase % Decrease % no change $000 $000 $000 $000 $000 Panel A: Early adopters (n = 40) Assets Business combinations (NZ IFRS 3) -6,898 0 0 856 81,000 45.00 10.00 45.00 Financial instruments (NZ IAS 32/39) -6,000 0 0 114 121,217 37.50 12.50 50.00 Income taxes (NZ IAS 12) -68,000 0 0 18 130,504 27.50 22.50 50.00 Intangible assets (NZ IAS 38) -145,849 0 0 0 51,000 17.50 10.00 72.50 Liabilities Income taxes (NZ IAS 12) -216 0 0 5,038 664,841 42.50 5.00 52.50 Employee benefits (NZ IAS 19) -161 0 0 79 16,000 30.00 2.50 67.50 Financial instruments (NZ IAS 32/39) -3,000 0 0 0 545,593 22.50 10.00 67.50 Equity Income taxes (NZ IAS 12) -663,457 -3,438 -5 2 29,919 25.00 50.00 25.00 Business combinations (NZ IFRS 3) -6,898 0 0 856 81,000 45.00 10.00 45.00 Financial instruments (NZ IAS 32/39) -543,736 0 0 81 3,971 32.50 22.50 45.00 Employee benefits (NZ IAS 19) -74,000 -79 0 0 161 2.50 30.00 67.50 Intangible assets (NZ IAS 38) -145,849 0 0 0 51,000 17.50 10.00 72.50 Provisions (NZ IAS 37) -2,416 0 0 0 47 5.00 15.00 80.00 Revenue recognition (NZ IAS 18) -30,000 0 0 0 5,735 2.50 17.50 80.00 Incomes Financial instruments (NZ IAS 32/39) -5,000 0 0 0 2,126 5.00 5.00 90.00 Expenses Business combinations (NZ IFRS 3) -69,000 -1,334 -36 0 4,434 10.00 50.00 40.00 Income taxes (NZ IAS 12) -6,000 -24 0 0 28,260 22.50 32.50 45.00 Employee benefits (NZ IAS 19) -124 0 0 0 3,000 17.50 5.00 77.50 Financial instruments (NZ IAS 32/39) 0 0 0 0 35,394 20.00 0.00 80.00 Share-based payment (NZ IFRS 2) 0 0 0 0 963 10.00 0.00 90.00 69 Table 5 (cont.) Minimum 25th Percentile Median 75th Percentile Maximum % Increase % Decrease % no change $000 $000 $000 $000 $000 Panel B: Late adopters (n = 40) Assets Financial instruments (NZ IAS 32/39) -68,000 0 0 318 101,300 40.00 15.00 45.00 Income taxes (NZ IAS 12) -1,599 0 0 8 13,700 25.00 12.50 62.50 Business combinations (NZ IFRS 3) -1,595 0 0 0 94,384 15.00 2.50 82.50 Intangible assets (NZ IAS 38) 0 0 0 0 19,035 12.50 0.00 87.50 Liabilities Income taxes (NZ IAS 12) -243,000 0 0 1,930 297,000 40.00 2.50 57.50 Employee benefits (NZ IAS 19) -1,000 0 0 98 18,002 37.50 2.50 60.00 Financial instruments (NZ IAS 32/39) -17,445 0 0 99 145,000 27.50 10.00 62.50 Equity Financial instruments (NZ IAS 32/39) -213,000 -2 0 182 89,700 37.50 25.00 37.50 Income taxes (NZ IAS 12) -283,600 -1,930 0 0 243,000 17.50 45.00 37.50 Employee benefits (NZ IAS 19) -18,002 -98 0 0 4,000 2.50 37.50 60.00 Business combinations (NZ IFRS 3) -2,882 0 0 0 94,384 15.00 2.50 82.50 Provisions (NZ IAS 37) -181,000 0 0 0 1,026 7.50 10.00 82.50 Intangible assets (NZ IAS 38) 0 0 0 0 19,035 12.50 0.00 87.50 Revenue recognition (NZ IAS 18) -4,770 0 0 0 13,464 2.50 7.50 90.00 Incomes Financial instruments (NZ IAS 32/39) -92,040 0 0 0 133,000 17.50 5.00 77.50 Expenses Income taxes (NZ IAS 12) -41,706 -3 0 6 18,000 25.00 25.00 50.00 Financial instruments (NZ IAS 32/39) -1,081 0 0 0 118,000 20.00 15.00 65.00 Share-based payment (NZ IFRS 2) -1,134 0 0 0 1,000 20.00 5.00 75.00 Employee benefits (NZ IAS 19) -158 0 0 0 1,036 15.00 10.00 75.00 Business combinations (NZ IFRS 3) -94,382 0 0 0 5,810 10.00 15.00 75.00 70 Table 6 reports on the significance of differences between the IFRS adoption changes recorded for early and late adopters. Table 6 Significance of difference in changes arising from individual accounting standards: Early adopter change compared to late adopter change a Total Assets Total Liabilities Total Equity Incomes Expenses Income Taxes 0.032 1.531 1.082 N/A 0.164 0.974 0.126 0.959 N/A 0.869 Financial instruments 0.427 0.094 0.974 1.000 2.658 0.670 0.925 0.330 0.317 0.008 Business combinations 0.185 1.732 1.136 1.342 0.917 0.853 0.083 0.256 0.180 0.359 Employee benefits 0.053 0.582 0.826 N/A 0.322 0.958 0.561 0.409 N/A 0.747 Intangible Assets 0.242 0.000 0.655 1.000 0.189 0.809 1.000 0.513 0.317 0.850 Revenue Recognition 0.473 1.658 0.878 0.868 0.913 0.636 0.097 0.380 0.386 0.361 Provisions 1.083 0.689 0.548 0.000 1.414 0.279 0.491 0.584 1.000 0.157 Share-based payment N/A N/A N/A N/A 1.556 N/A N/A N/A N/A 0.120 a = Reported Mann-Whitney U test statistics are Z scores with two-tailed p-values italicised. N/A = Accounting standard affects fewer than 20 percent of EA or LA firms for financial statement element concerned. The data for this analysis is extracted from the IFRS/OLD GAAP reconciliations required by NZ IFRS 1.35 As IFRS comprises in excess of 40 standards, the table presented in this paper is restricted to separate consideration of only those standards which affect at least 20 percent of the EA, LA or combined sample for one or more of the elements of the balance sheet and of the income statement.36 Eight such standards are identified in respect of balance sheet 35 See Chapter 4 (4.1) for more detail regarding variation in format and detail of reconciliations for sample firms, as well as methodology for data extraction. 36 ‘No change’ figures greater than 80 percent appear in the table only where this is necessary for consistency and comparison across panels (i.e., where the equivalent figure in the second panel is less than 80 percent). 71 elements, all of which are evident for equity.37 One additional standard (NZ IFRS 2) is identified for income statement elements. The most striking feature of Table 5 is that the median observation is zero across all financial statements elements and accounting standards, with two exceptions for EA (income tax impact on equity of -$5,000 and business combinations impact on expenses of - $36,000).38 The majority of the 25th and 75th percentiles are also zero for both EA and LA. The last three columns in Table 5 report the percentage of firms experiencing increases, decreases or no change as a result of each standard. Increases are more frequent than decreases (i.e., IFRS tends to increase values of financial statement elements). The range of ‘% no change’ is from 25 percent to 90 percent after excluding most standards from Table 5 as they affect fewer than 20 percent of sample firms. This indicates that for most EA and LA firms a specific standard of IFRS has no impact. However, the minimums and maximums indicate that a specific IFRS can be very material for a small number of firms (e.g., NZ IAS 38 Intangible Assets, where the equity of 72.50 percent of EA (LA: 87.50 percent) are unaffected by this standard but minimum and maximum effects of -$145.8 million (LA: nil) and $51 million (LA: $19 million) respectively are reported). Table 6 results are also striking in that the IFRS adoption changes for EA and LA, arising from individual accounting standards are significantly different in only three instances. These significant differences are highlighted in Table 6 and discussed in further detail below. Kabir et al. (2010), rank the 11 standards that they identify as most frequently giving rise to IFRS adjustments and report, inter alia, the median percentage changes for these standards. The nine standards identified in Table 5 are consistent with those identified by Kabir et al. (2010), although frequency rankings differ slightly, probably due to sampling differences. Nature of IFRS adjustments, EA/LA differences and adoption timing implications Table 5 shows that the most frequent adjustments (i.e., lowest ‘% no change’) relate to the effects of NZ IAS 12 Income Taxes on equity. Only 25 percent of EA are unaffected by this standard (LA: 37.5 percent). NZ IAS 12 decreases equity in 50 percent of the EA 37 NZ IAS 32 and 39 are considered together as they both relate to financial instruments and hence appear as one line item in Table 5. 38 Stent et al. (2010) make similar observations. 72 observations (LA: 45 percent) and in general NZ IAS 12 adjustments to other financial statement elements are more widespread for EA than for LA. Proportionately more EA than LA experience decreases in expenses as a result of this standard (EA: 22.50 percent increases and 32.50 percent decreases; LA: increases as well as decreases of 25 percent). No clear incentives for adoption timing decisions are apparent for this accounting standard. The income tax adjustments for sample firms are due mainly to deferred tax differences which arise because NZ IAS 12 adopts a ‘balance sheet approach’, which is significantly different to the ‘income statement approach’ formerly used under OLD GAAP. While the inter-quartile range for the impact of NZ IAS 12 on EA equity is $3.44 million (LA: $1.93 million), the range is $693 million for EA (LA: $526.6 million). These results provide an insight as to the variation in impact of the deferred tax adjustments across the sample firms, as well as a reminder that these adjustments may be both book-value increasing and decreasing. Overall, the significant increases in liabilities under IFRS noted in Table 4 can be attributed mainly to this standard. The absolute size of adjustments noted above for EA and LA are similar. While this finding is in contrast with the proportions evident in Table 3 (e.g., EA median liabilities are $64.9 million while LA median liabilities are $18.9 million), it is consistent with Table 4 results which show larger increases in liabilities for LA (e.g., median increase of 0.035) than for EA (e.g., median increase of 0.010), suggesting possible incentives for delaying IFRS adoption. Adjustments under NZ IAS 32 and 39 (financial instruments) are the second most frequent across the EA and LA samples. Equity is affected for 55 percent of EA (LA: 62.5 percent). In contrast to IFRS adjustments for income taxes, adjustments for financial instruments affect more LA than EA. Overall effects are more beneficial (for equity and earnings) and more widespread for LA than EA, indicating potential incentives for late adopters to delay adoption of IFRS. Table 6, indicates no significant differences in changes experienced by EA and LA for balance sheet elements, but the difference for expenses is strongly significant (p = 0.008). IFRS adjustments for business combinations increase EA equity and assets by 45 percent (LA: 15 percent) and decrease EA expenses by 50 percent (LA: 15 percent). These adjustments arise mainly as a result of the requirements in NZ IFRS 3 that goodwill should not be amortised, but should be subject to impairment testing. Hence it follows that, because EA firms are generally larger than LA firms (as reported in Table 4), group reporting and the incidence of goodwill will be more frequent for EA. These are the largest frequency differences between EA and LA and indicate potential bonus and debt hypothesis incentives 73 for early adoption, but surprisingly Table 6 reports that differences between EA and LA changes are not significant for any financial statement elements.39 Employee benefits (under NZ IAS 19) also contribute to the significant increases in liabilities affecting 30 percent of EA (LA: 37.5 percent). New requirements under NZ IAS 19 include requirements for employers to recognise an asset or liability in respect of any employee defined benefit plans and to accrue for both vested and non-vested employee benefits (e.g., long service leave and sick leave). These requirements are therefore potential incentives for adoption timing e.g., a preference to delay the negative effects for leverage ratios/debt contracting, particularly for the more widely affected late adopters. IFRS adjustments for intangible assets (NZ IAS 38) result in increases in assets and equity for 17.5 percent of EA (LA: 12.5 percent). The most common explanation provided for this adjustment relates to software and other intangibles that had to be recognised and reclassified to separate them from other related non-current assets (e.g., computer hardware and investments). Overall, effects are more beneficial and more widespread for EA than LA, indicating potential debt hypothesis incentives for early adoption. Remaining equity adjustments reported in Table 5 are less widespread, affecting only EA equity by more than 20 percent (i.e., effects of less than 20 percent on both assets and liabilities). IFRS adjustments relating to revenue recognition (NZ IAS 18) were due mainly to requirements to defer recognition of revenue for land development and real estate sales, product installations and various services, pending performance of required recognition criteria. Examples of IFRS adjustments relating to provisions (NZ IAS 37) include requirements to adjust provisions to take account of: sales returns, make good clauses in leases, executory contracts on entertainment programme rights and impairments of receivables. In both cases (NZ IAS 18 and 37), the effects are more negative and more widespread for EA than for LA, indicating potential political incentives for EA adoption timing decisions. Table 6 provides additional support for this possibility, although only one weakly significant difference between EA and LA relating to these changes exists, namely the revenue recognition changes to liabilities (p = 0.097). Share-based payments are the last category of adjustments which have non-zero effects 39 Liabilities are reported as being weakly significant but this is due to an isolated adjustment relating to restatement of business acquisitions involving onerous leases for one LA. There were 60 adjustments to equity and assets. This finding is therefore considered to be of little relevance. 74 exceeding 20 percent and this is so only for LA expenses. These adjustments increase LA expenses for 20 percent of observations while for EA only 10 percent of observations record increased expenses. While the minimum and maximum values indicate that these adjustments are relatively small in comparison to the other adjustments, the effects are more negative and wide-spread for late adopters. This is, therefore, another potential incentive for late adoption decisions, particularly in view of likely increased ownership concentration in the management structures of smaller firms (Ayres, 1986). In summary, this section investigates the nature of significant differences arising from adoption of IFRS, distinguishing between results for EA and LA, with the following results: • The accounting standards which are found to be the most frequent sources of IFRS adjustments are those relating to income taxes, financial instruments, business combinations, employee benefits, intangible assets, revenue recognition, provisions and share-based payment. These accounting standards affect at least 20 percent of EA and/or LA sample firms for one or more of the financial statement components. These findings are consistent with many of the expectations discussed within the accounting profession prior to IFRS adoption (e.g., Ernst & Young, 2004). In general, they are also consistent with those of Hung & Subramanyam (2007), except that differences between IFRS and local GAAP are more pronounced for their results for German firms; • For most EA and LA firms a specific standard of IFRS has no impact. The majority of the 25th and 75th percentiles are zero for both EA and LA, as well as all but two of the median observations. However, the minimums and maximums indicate that a specific standard can be very material for a small number of firms; • Percentage increases are more frequent than percentage decreases, which is consistent with the findings in 5.2.2 that IFRS tends to increase values of financial statement elements; • Descriptions of the nature of the most frequent adjustments are provided. These include the effect that adjustments have on financial statement elements (e.g., increases in assets and equity) and the most common explanations provided for the adjustments (e.g., dispensing with the amortisation of goodwill); • Effects on assets and equity are generally more widespread for EA than for LA. Also, EA generally report smaller ‘no impact’ percentages than LA. A further difference between EA and LA is that absolute amounts of adjustments for early 75 adopters tend to be larger than those of late adopters.40 Although these and various other differences are reported, only two significant differences (after disregarding one as explained in the footnote on pg. 73) between EA and LA are found regarding IFRS adoption adjustments: adjustments to expenses relating to financial instruments and adjustments to liabilities relating to revenue recognition; • While no apparent trend emerges regarding the implications of the above findings, proportions of EA and LA firms affected by some IFRS adjustments and the nature of these effects suggest potential incentives for adoption timing decisions. These include potential incentives related to the bonus and debt hypotheses for both EA and LA, as well as the political hypothesis for EA. 5.2.4 EFFECT OF IFRS ON KEY RATIOS The results for this section (5.2.4) relate to the third component of research question 1. Having considered differences in financial statement elements as a result of IFRS adoption in the previous sections, the focus in this section is on ratios. Table 7 presents results of the analyses into the effect of IFRS on key ratios, which serve as proxies for variables which may influence or be of interest to analysts. Five key ratios are chosen: return on equity (net profit to equity), return on assets (net profit to total assets), leverage (total liabilities to equity), asset turnover (revenue to total assets) and return on sales (net profit to revenue).41 These ratios reflect the main ratios in the Du Pont analysis.42 Panel A (EA) and Panel B (LA) report that under IFRS the median ROE for EA (LA) increases from 0.120 (0.074) to 0.130 (0.120). Panel C (EA) and Panel D (LA) report that ROE increases for 71.79 percent (60 percent) of observations and decreases for 23.08 percent (27.50 percent) of observations. The change in this ratio is more significant for LA (p = 0.006) than for EA (p = 0.034), although Panel E reports that this difference between EA and LA is not significant (p = 0.806). 40 For example, the largest amounts recorded for minimums and maximums for EA were -$663.4 million and $664.8 million respectively, whereas for LA the equivalent figures were -$283.6 million and $297 million. 41 It is normal to use EBIT (earnings before interest and taxation) in estimating return on assets. However, the OLD GAAP/ IFRS reconciliations only provide details for net profit. Hence it is not possible to estimate EBIT. Similarly, it is not possible to estimate an Altman-type Z score model. 42 In the Du Pont analysis return on sales and asset turnover are components of return on assets. Return on assets and leverage are components of return on equity. 76 Table 7 Descriptive statistics on effect of IFRS on key ratios ROE OLD GAAP ROE NZ IFRS ROA OLD GAAP ROA NZ IFRS LEV OLD GAAP LEV NZ IFRS ATO OLD GAAP ATO NZ IFRS ROS OLD GAAP ROS NZ IFRS Panel A: Ratio comparisons for early adopters Mean 0.084 0.071 -1.473 -1.476 1.445 1.194 0.979 1.023 0.118 0.098 Std. Deviation 0.784 0.858 9.547 9.546 2.766 3.705 0.813 0.911 0.299 0.151 Minimum -4.050 -4.500 -60.330 -60.330 -1.040 -12.830 0.000 0.000 -0.480 -0.480 25 Percentile 0.040 0.059 0.011 0.013 0.335 0.298 0.356 0.344 0.026 0.033 Median 0.120 0.130 0.050 0.049 0.625 0.642 0.641 0.593 0.070 0.088 75 Percentile 0.221 0.240 0.099 0.101 1.339 1.434 1.696 1.799 0.155 0.167 Maximum 2.480 2.480 0.250 0.210 16.080 16.100 3.010 3.280 1.630 0.410 Panel B: Ratio comparisons for late adopters Mean 0.076 0.316 -0.156 -0.142 0.611 0.608 1.023 0.777 -8.444 -2.400 Std. Deviation 1.852 1.638 0.661 0.711 1.299 1.378 1.271 0.771 43.227 7.780 Minimum -5.580 -2.050 -3.340 -3.850 -6.270 -6.270 0.000 0.010 -272.660 -35.540 25 Percentile -0.072 -0.020 -0.091 -0.068 0.220 0.222 0.202 0.133 -0.245 -0.485 Median 0.074 0.120 0.039 0.049 0.621 0.657 0.825 0.580 0.048 0.061 75 Percentile 0.270 0.276 0.119 0.114 1.128 1.179 1.292 1.179 0.131 0.144 Maximum 9.730 9.730 0.320 0.320 2.760 2.670 7.030 2.590 1.680 3.030 Ratios under OLD GAAP and NZ IFRS are reported, rather than the change to these ratios, for ease of reference to base figures. Key ratios are defined as follows: ROE is return on equity which equals net profit divided by book value of equity; ROA is return on assets, which equals net profit divided by total assets; LEV is leverage, which equals total liabilities divided by book value of equity; ATO is asset turnover, which is revenue to total assets; ROS is return on sales, which is net profit to revenue. 77 Table 7 (cont.) ROE Change ROA Change LEV Change ATO Change ROS Change Panel C: Change statistics for early adopters a Increases (%) 71.79 67.50 58.97 27.50 73.68 Decreases (%) 23.08 27.50 35.90 65.00 26.32 No change (%) 5.13 5.00 5.13 7.50 0.00 Z statistic 2.120 2.226 1.848 1.697 2.951 p-value (two-tailed) 0.034 0.026 0.065 0.090 0.003 Panel D: Change statistics for late adopters a Increases (%) 60.00 52.50 62.50 32.50 65.00 Decreases (%) 27.50 35.00 22.50 55.00 25.00 No change (%) 12.50 12.50 15.00 12.50 10.00 Z statistic 2.768 1.097 2.197 2.113 2.435 p-value (two-tailed) 0.006 0.272 0.028 0.035 0.015 Panel E: Change comparison for early and late adopters b Z statistic 0.245 1.232 0.029 0.313 0.380 p-value (two-tailed) 0.806 0.218 0.977 0.754 0.704 a = The reported test statistic is a Wilcoxon test for equality of matched pairs b = The reported test statistic is a Mann-Whitney U test for significance of differences in ratio change between EA and LA 78 The median values for the remaining ratios all increase except for ROA, where only EA decrease (from 0.050 to 0.049) and ATO, where both EA and LA show decreases (EA: from 0.641 to 0.593 and LA from 0.825 to 0.580). In summary, Table 7 results indicate that: • The ‘no-change’ effect of IFRS on financial ratios is small but that more LA are unaffected (10 to 15 percent) than EA (0 to 7.5 percent); • The impact of NZ IFRS does not simply result in a uniform increase in financial statement ratios but has a firm-specific effect. For some firms a particular ratio may increase, while for other firms that ratio may decrease; • The percentages of ratio increases are typically double those of decreases, except for ATO. Hence the tendency for IFRS to increase values of financial statement elements (noted in 5.2.2 and 5.2.3) results in tendencies towards improved profitability ratios but deteriorations in leverage and efficiency ratios; • Changes in all ratios are statistically significant at conventional levels, except for one LA observation (return on assets); • The difference between EA and LA changes are not found to be significant for any ratios. The results above are generally consistent with those of Stent et al. (2010). However, the increase in the sample size of early adopters and use of Mann-Whitney tests for this thesis has clarified that, while there are some differences between results for early and late adopters, these are not significant. Kabir et al. (2010) report similar findings for return on assets and leverage (i.e., increases for both), although they do not differentiate between early and late adopters or consider the other ratios above. To conclude, the analysis of the effect of IFRS on ratios confirms the results of the analysis of the effect of IFRS on financial statement elements. Hypotheses 1a and b are again rejected, because significant differences are found in all but one of the ratios for both EA and LA firms as a result of NZ IFRS adoption (the one exception was return on assets for LA firms). However, Hypothesis 2 is accepted as the change in ratios for EA is not found to be significantly different to the change in ratios for LA. 79 5.2.5 SUMMARY The results for this part of the chapter, on the impact of IFRS adoption on financial information, provide evidence of significant changes to most financial statement elements and ratios for both early and late adopters. These changes may provide incentives for adoption timing decisions. Contrary to some of the earlier literature (e.g. Brown, 1985; Brown & Brandi, 1986; Langer & Lev, 1993) however, there is no evidence of statistically significant differences when the changes for early adopters are compared to those for late adopters . These earlier studies concentrate on investigation of the bonus, debt and political cost hypotheses. Subsequent studies (e.g., Hung & Subramanyam, 2007; Iatridis & Joseph, 2006; Li, 2010) emphasise the interaction of the incentives implicit in these hypotheses with firm characteristics. The findings summarised above for this thesis align with the wider view of the later studies. That is, to take account of the interaction between incentives and firm characteristics when investigating accounting choices such as adoption timing. 80 5.3 IFRS-RELATED DISCRETIONARY NARRATIVE DISCLOSURES 5.3.1 INTRODUCTION The results for this part of the chapter (5.3) relate to research question 2: RQ2 What do discretionary narrative disclosures in annual reports reveal about early and late adopter firms’ attitudes to IFRS adoption? These results are discussed in two sub-sections, with reference to the five theme categories identified for IFRS analysis of annual reports: importance, potency, evaluative, activity and manageability. The first sub-section (5.3.2), form oriented analysis, is concerned with the extent of disclosures. The second sub-section (5.3.3), meaning oriented analysis, is concerned with the nature of disclosures.43 5.3.2 FORM ORIENTED ANALYSIS The results for this section (5.3.2) relate to research question 2.1: RQ2.1 Are there significant differences between early and late adopters (with respect to the extent of discretionary narrative disclosures about IFRS in annual reports)? In Table 8 four ‘count’ measures are provided to indicate the extent to which the adoption of IFRS is reported in the narrative extracts. Count measures are a reasonably objective way of measuring the importance of an issue (i.e., the first theme category from Table 1). The measures reported are: the number of IFRS mentions per firm, the number of words, the space devoted relative to one page of the annual report, and the space devoted relative to all pages in the narrative section of the annual report. The early adopters have greater means and medians on all measures. The means and medians in Table 8 also indicate that the data are not normally distributed. A non- parametric Mann-Whitney U test is therefore reported to assess the statistical differences between EA firms and LA firms. In all cases the differences are statistically significant at conventional levels. The results in Table 8 support the expectation that IFRS adoption is more important to EA relative to LA firms. 43 See Chapter 4.3.2 for further detail regarding content analysis procedures and the themes and sub- themes identified in Table 1 and Appendix 3. 81 Table 8 Form oriented analysis (count data per firm) Mentions b Words c Space devoted to IFRS a    % of 1 page of annual report % of narrative pages Early Adopters    Mean 7 208 12% 0.4% Median 5 143 9% 0.3% Minimum 0 0 0% 0.0% Maximum 27 1,044 58% 1.5% Standard deviation 15 224 1% 0.1% Late Adopters Mean 4 125 6% 0.3% Median 3 69 3% 0.1% Minimum 0 0 0% 0.0% Maximum 17 834 42% 2.1% Standard Deviation 8 191 3% 0.0% Mann-Whitney U tests   Mean rank   Early 45.43 46.68 47.65 46.48 Late 35.58 34.33 33.35 34.53 Z statistic 1.919 2.399 2.777 2.321 (p-value) d 0.0275 0.0080 0.0025 0.0100 a = Calculated in cm2 (1 PDF page is 29.5cm x 21 cm = 619.5cm2). b = Total number of mentions of terms "IFRS" or "International Financial Reporting Standards". c = Total word count associated with IFRS discretionary disclosures. d = One-tailed tests in terms of expectation that EA firms will consider IFRS adoption more important and hence make more IFRS related discretionary disclosures. The minimum value observed in Table 8, across all measures for both EA and LA is zero. This observation raises the possibility that these results (i.e., little or no disclosure by firms) may be due to IFRS having minimal impact on the financial statements. Further investigation is therefore carried out on firms for which no IFRS narrative disclosures are found with the following results (untabulated): There are 21 ‘no IFRS mentions’ firms: seven EA and 14 LA. Two of the seven EA and three of the 14 LA make statements asserting that IFRS has had no impact on 82 the financial statements.44 The reconciliations between NZ GAAP and IFRS (required by NZ IFRS 1, para 40) are also examined. These show that the net adjustment to equity or income is more than 10 percent for the remaining five of the seven EA firms and for six out of the remaining 11 LA firms. Furthermore, for the five LA firms where the overall net effect of IFRS is less than 10 percent, the impact on individual line items is found to be material.45 Hence low levels of IFRS disclosure in annual report narratives are not explained by the adoption of IFRS having an immaterial impact on financial statements. To conclude, the hypotheses for research question 2.1 are re-stated and the relevant sub- hypothesis (see bold text below) tested against the above results: H03: There are no significant differences in the extent (ED) and nature (ND) of discretionary narrative disclosures about IFRS in annual reports for Early Adopters (EA) and Late Adopters (LA) i.e., H03a µED of EA = µED of LA’s H03b: µND of EA = µND of LA; Hypothesis 3a is rejected, because significant differences are found between EA and LA for all four measures of the extent of discretionary narratives disclosures related to IFRS adoption. 5.3.3 MEANING ORIENTED ANALYSIS The results for this section (5.3.3) relate to research questions 2.1 and 2.2: RQ2.1 Are there significant differences between early and late adopters (with respect to the nature of discretionary narrative disclosures about IFRS in annual reports)? RQ2.2 What is the nature of significant differences (if any) that are identified? Table 9 reports on the themes identified in the IFRS extracts as follows: the number of firms that report a particular theme; and the word count for that theme. IFRS mentions are found in the narratives of annual reports for a total of 33 (82.5 percent) EA and 26 (65 percent) LA. The total word counts associated with all IFRS mentions are 8,141 words for EA and 4,553 for LA. 44 These statements appear in the notes to the financial statements as opposed to the discretionary narrative sections of the annual report. 45 NZ IFRS 1, para 40 requires “ … sufficient detail to understand the material adjustments …”. 83 Table 9 Meaning oriented analysis (themes) Firms Word counts Early Adopters a Late Adopters a Early Late No. %  No. %  Adopters Adopters Firms 33 26 8,141 4,553 Importance Contextual 24 60% 20 50% 1,823 1,896 Highlight 12 30% 1 3% 500 53 Informing of importance 25 63% 19 48% 4,173 2,998 High potency Assets 12 30% 6 15% 2,251 1,643 Liabilities 7 18% 2 5% 1,773 758 Equity or income 15 38% 8 20% 2,229 2,031 Revenue 9 23% 2 5% 1,137 490 Expenses 14 35% 5 13% 2,747 967 Disclosure 7 18% 3 8% 1,417 1,059 Low potency Limited or immaterial impact 4 10% 2 5% 371 251 General or vague impact 9 23% 1 3% 936 40 Evaluative Neutral 4 10% 1 3% 433 28 Negative 14 35% 10 25% 1,585 2,258 Positive 3 8% 0 0% 254 0 a = For both Early and Late Adopters, n = 40 Importance The ‘importance’ theme comprises three sub-themes: ‘contextual’, ‘highlight’ and ‘informing of importance’ (as described more fully in Appendix 3). A ‘contextual’ mention simply indicates that accompanying financial information is prepared under IFRS rather than previous GAAP. Such IFRS contextual mentions are evident in annual report narratives of 24 (60 percent) of EA and 20 (50 percent) of LA. Total word counts associated with this sub-theme are 1,823 for EA and 1,896 for LA. Twelve EA ‘highlight’ the move to IFRS as an achievement (500 words), whereas only one LA (53 words) uses a ‘highlight’ sub-theme. The ‘informing of importance’ sub-theme is used by 25 EA (63 percent) with 4,173 words, compared to 19 (48 percent) of LA with 2,998 words. ‘Informing of importance’ is the most dominant theme in Table 9 in terms of both the number of firms making such IFRS mentions and the number of words 84 associated with this theme. To test whether there is statistical support for the observed differences between early and late adopters, an ‘importance’ score is constructed for each firm, using the ordinal nature of the sub-themes within the importance themes.46 Each firm is allocated a score as follows: ‘context’ mentions = 1; ‘highlight’ mentions = 2; ‘informing of importance’ mentions = 3. Given that the coding did not partition IFRS extracts into mutually exclusive themes, there is a possible maximum score of 6 for each firm.47 As reported in Table 10 (over page), a Mann-Whitney U test is used and findings indicate a statistically significant difference between early and late adopters for the ‘importance’ theme at the 1 percent level. Potency The potency theme is divided into two categories. ‘High potency’ relates to narrative which is specific as to the nature and extent of IFRS impacts. There are sub-themes in this category for the financial statement elements that are described as being materially affected (e.g., assets, liabilities, equity, revenue, expenses or disclosure). The ‘low potency’ category has two sub-themes: (1) where the narrative comments on limited or immaterial impacts, and (2) where the impact description is general or vague. Table 9 shows that less than 50 percent of firms provide high potency IFRS narratives for any of the sub-themes (i.e., financial statement elements). The highest observations for EA are in relation to equity (15 firms and 2,229 words) and expenses (14 firms and 2,747 words). For LA the highest observations are for equity (8 firms and 2,031 words). EA have a higher number of firms and words than LA across all high potency categories. Minimal low potency information is provided by both EA and LA firms. EA provide more low potency comments than LA, but this is consistent with the higher levels of disclosure overall by EA firms. 46 Low frequencies for many sub-themes mean that Chi-square tests cannot be used for testing at this level. The sub-themes are therefore aggregated to create a disclosure index which is used to statistically test the differences between early and late adopters. 47 As an alternative index, weightings for the sub-themes are discarded and each simply given a mention score of 1. Hence the maximum score is 3. The results are similar to those reported in Table 10. 85 Table 10 Mann-Whitney test statistics relating to themes Early Adopters Late Adopters 'Importance' index Mean score 3.075 1.95 Mean rank 46.575 34.425 Z statistic 2.414 p-value a 0.008 'High potency' index Mean score 1.6 0.65 Mean Rank 46.3 34.7 Z statistic 2.488 p-value a 0.007 'Low potency' index Mean score 0.425 0.125 Mean rank 46.575 34.428 Z statistic 2.414 p-value a 0.008 ‘Combined potency' index Mean score 0.975 0.425 Mean rank 47.563 33.438 Z statistic 2.978 p-value a 0.001 ‘Evaluative’ index Mean score -0.2 -0.25 Mean rank 40.98 40.03 Z statistic 0.235 p-value a 0.407 a = One-tailed tests in terms of expectations that EA firms will consider IFRS adoption more important and more beneficial than LA firms and also that EA firms will be more specific than LA firms about the impact of adoption (i.e. EA narratives will have higher potency than LA firms). 86 To test for statistically significant differences between EA and LA, a high potency and a low potency index is derived for each firm. The high potency index scores 1 point for a specific reference to a financial statement element (i.e., a maximum possible score of 6 per firm). Low potency themes are scored: ‘limited or immaterial impact’ = 2 and ‘general’ = 1 (i.e., a maximum possible score of 3 per firm). Similar to the high potency score, the low potency score emphasizes the specific nature of comments. The low potency score reflects the view that a ‘general’ comment is less specific than a ‘limited or immaterial’ comment. The differences between potency indices for early and late adopters are statistically significant for both ‘high potency’ themes (p = .007) and ‘low potency’ themes (p = .008). One weakness of the high potency index is that where the effect of IFRS is widespread (i.e., where the impact affects assets, liabilities, revenues and expenses) a firm can have a higher ‘high potency’ score relative to a firm where the impact of IFRS is significant but isolated just to one area (e.g., assets). A combined potency score is therefore developed by taking the ‘low potency’ score and adding 1 if the ‘high potency’ score is non-zero. The combined potency index is higher for early adopters and the difference between early and late adopters is statistically significant (p = .001). Evaluative Evaluative themes are classified into positive (beneficial), negative (adverse), and neutral sub-themes. Commentary on the adverse consequences of adopting IFRS is notably higher than for the other two evaluative sub-themes. For EA, negative comments came from 14 firms (35 percent) and comprise 1,585 words. For LA the corresponding amounts are 10 firms (25 percent) and 2,258 words. To test for statistically significant differences between EA firms and LA firms the ‘evaluative’ index for each firm is allocated 1 for ‘positive’ or ‘neutral’ comments, zero for no comment and -1 for ‘negative’ comments. Findings indicate that there is no statistically significant difference between EA and LA firms (p = .407).48 That is, both early and late adopters include evaluative narrative disclosures that are predominantly negative in nature. 48 As for the other themes, a sensitivity test is conducted using an alternative index: ‘Neutral’ mentions = 1; ‘Negative’ mentions = -2; ‘Positive’ mentions = +2 (i.e. a range of possible scores between -2 and +3 per firm with signed weightings to indicate whether the tendency for each firm’s ‘Evaluative’ narrative is positive or negative overall). The result (p = 0.824) confirms the finding for the original test. 87 Manageability The manageability theme relates to whether the impact of IFRS adoption was expected or unexpected. No such theme emerged during the content analysis. Word searches were also undertaken of the IFRS extracts with regard to all truncation variations of: ‘*expect*’ OR ‘*surpris*’ OR ‘*foresee*’ OR ‘*anticipat*’ OR ‘*predict*’ OR ‘*startl*’ OR ‘*astonish*’ OR ‘*usual’ OR ‘*likel*’ OR ‘*manageabl*’. These word searches confirm that there is no evidence of this theme. Accordingly, there is no analysis to report for this theme. Magnitude of change in earnings or equity were considered as proxies for this theme, but rejected for two reasons: (a) it conflicts with the methodology, which is restricted to discretionary narrative and, (b) while these may be good proxies for investigating situations where information asymmetry is likely between managers and parties external to the firms (e.g., Wang & Welker, 2011), the setting for this study concerns only managers. Further, it is noted that manageability is likely reflected in the decision to early or late adopt IFRS, analysis of which is already incorporated in our research design. The results in Table 9 and 10 therefore provide further support for the expectation that IFRS adoption is more important to EA relative to LA firms. They also provide evidence supporting the expectation that early adopters who are materially affected will be more specific about the impact of IFRS adoption. However, in contrast to the expectation that EA (relative to LA) would consider IFRS adoption to be more beneficial, predominantly negative perceptions are found and in this respect, there is no significant difference between EA and LA. The univariate analysis above indicates that EA make more disclosures than LA on all themes, except the ‘evaluative’ theme. However, univariate results do not control for other firm specific factors that might impact discretionary disclosure. The impact of early adoption is therefore also examined using the multivariate model described in Chapter 4:49 DISji = β0 + β1 EAi+ β2 SIZEi+ β3 AUDi + β4 OSHIPi + β5 ORIGINi + β6-10 INDki (1) As each of the disclosure indices can be viewed as an ordinal ranking, a multivariate ordered logistic regression is applied, the results of which are reported in Table 11.50 49 See 4.3.2 for descriptions of variables and development of model. 50 Ordinary least squares regression yield similar results. 88 Only the result of the ‘importance’ and the ‘combined potency’ disclosure indices are reported. The evaluative disclosure index is omitted as the univariate result is not significant.51 The results in Table 11 are as expected. The coefficient on early adoption (EA) is positive and significant. The coefficients on the control variables are as expected, and are weaker than EA. Table 11 Ordered logistic regression on disclosures score Importance Index Potency Index Expected sign Coefficient p-value Coefficient p-value EA + 0.862 0.044 1.093 0.022 SIZE + 0.243 0.048 -0.041 0.395 AUD + 0.972 0.114 1.448 0.045 OWNERSHIP + 0.01 0.154 0.022 0.027 ORIGIN + -0.412 0.28 -0.41 0.285 IND - energy ? -0.151 0.42 -0.751 0.358 IND - goods ? 0.315 0.362 -0.535 0.601 IND - investment ? -0.13 0.428 -0.971 0.206 IND - primary ? 0.534 0.246 0.084 0.914 IND - property ? -0.082 0.346 0.003 0.989 Model Chi-square 21.98 0.015 20.43 0.026 Pseudo R2 0.083 0.119 None of the industry indicator variables is significant. The model Chi-square shows the model is significant, although the explanatory power is low for both models with a pseudo R2 of 8.3 percent and 11.9 percent. These findings demonstrate that there is an early adoption impact with regard to the importance and potency of IFRS adoption that is reflected in the narrative disclosures and which exists after controlling for factors such as firm size, auditor and industry. 51 The regression results for the evaluative model did not converge because there is little variation in the data (i.e., too many disclosure scores of zero). 89 To conclude, the hypothesis for research question 2.1 is re-stated and the relevant sub- hypothesis (see bold text below) tested against the above results: H03: There are no significant differences in the extent (ED) and nature (ND) of discretionary narrative disclosures about IFRS in annual reports for Early Adopters (EA) and Late Adopters (LA) ie. H03a µED of EA = µED of LA’s H03b: µND of EA = µND of LA; Hypothesis 3b is rejected, because significant differences are found between EA and LA for two out of three themes tested. These results indicate that EA disclosures are more potent (i.e., specific) and place more importance on IFRS adoption than LA disclosures. The results hold when controlling for other firm specific factors that might impact discretionary disclosure (e.g., firm size, auditor, industry). However, with regard to the third theme relating to evaluative disclosures, no statistically significant difference is found as both EA and LA disclosures are predominantly negative in nature. The nature of significant differences is covered in the above discussion, hence also addressing research question 2.2. Appendix 3 provides detailed descriptions of each theme and sub-theme, including quotations extracted from sample firms’ annual reports. These are used to convey information in the voice of the sample firms and to add to the richness of descriptions, hence providing further insight into the nature of significant differences between EA and LA disclosures. 5.3.4 SUMMARY The results for this part of the chapter provide evidence of significant differences between EA and LA with respect to discretionary narrative disclosures about IFRS adoption in annual reports. The study investigates what narrative disclosures reveal about managements’ perceptions as to the importance and effects of IFRS. Both form oriented and meaning oriented content analysis is used to examine the extent and nature of narrative disclosures. Although the sample size is small (n=80), all EA (40) have been considered as well as an equal number of randomly selected LA. The form oriented analysis concerns measurement of the extent of these disclosures (e.g., number of firms making such disclosures, word counts and page area devoted to the disclosures). The meaning oriented analysis is concerned with the nature of disclosures 90 and identifies three major themes: ‘importance’, ‘potency’ and ‘evaluative’. These themes are consistent with the content analysis framework used by Smith & Taffler (2000). Findings indicate that there is significantly more narrative disclosure relating to IFRS in annual reports of EA relative to LA for the year in which IFRS is adopted. 21 firms make no mention of IFRS at all, but of these, only seven are EA while 14 are LA. Furthermore, the EA median figures for three of the four measures relating to the extent of IFRS adoption disclosures are more than double those of LA. Differences between EA and LA are significant at the 1 percent level for two of these measures and at the 5 percent level for the remaining two measures. Meaning oriented analysis provides deeper insights. ‘Informing of importance’ of IFRS adoption is the most dominant theme. It applies to 25 EA (i.e., 63 percent) and 19 LA (i.e., 48 percent) but accounts for more than half of the total word count for both EA and LA. ‘High potency’ descriptions of IFRS impacts are much more prevalent than ‘low potency’ descriptions for both EA and LA. That is, the narratives reflect specific detail regarding the impact of IFRS rather than references that are vague, general or report immaterial impact. Differences between EA and LA as regards ‘Importance’ and ‘Potency’ themes are statistically significant. The multivariate analysis of these themes provides evidence that early adoption is a more strongly significant explanatory variable than many other incentives for voluntary disclosure, such as firm size, auditor and industry. While relatively few firms make ‘evaluative’ narrative disclosures, these are found to be predominantly negative with respect to IFRS adoption and there is no statistically significant difference in this regard between EA and LA. 91 5.4 COSTS AND BENEFITS ARISING FROM IFRS ADOPTION AS WELL AS OTHER SURVEY INSIGHTS 5.4.1 INTRODUCTION The results for this part of the chapter (5.4) relate to research question 3: RQ3 What are the costs and benefits of IFRS adoption for early as opposed to late adopters? These results are discussed in three sub-sections. The first sub-section (5.4.2), investigates the costs of IFRS adoption in terms of staff, time and financial commitments. The second sub-section (5.4.3) assesses perceived benefits of IFRS adoption. The third sub-section (5.4.4) considers additional insights as to potential adoption timing motivations which arise from qualitative survey data, namely comments in questionnaires and interview responses. Questionnaires were distributed to all EA and LA sample firms, with the exception of three Christchurch companies in our LA sample, which were excluded from the survey procedures out of sensitivity to their situation and priorities after the first earthquake in 2010. Responses were received from 17 EA (43 percent) and 17 LA (46 percent).52 These response rates compare favourably with the 35 percent response rate of Jones & Higgins (2006). A comparison of early surveys with later surveys for both EA and LA was carried out (late respondents are proxies for non-respondents).53 Mann-Whitney tests were carried out to compare responses of early surveys to those of late surveys for a total of 20 survey questions/sub-questions. No statistically significant differences are found except for 2 of the EA (i.e., 10 percent) and 3 of the LA (i.e., 15 percent) responses, indicating that non-response error is unlikely to be a serious issue. 52 Interview follow-ups were achieved with 13 EA (76%) and 14 LA(82%). 53 Early surveys are differentiated from later surveys by reference to records relating to questionnaire mailing dates, follow-up reminder dates and email correspondence. 92 5.4.2 COSTS OF IFRS ADOPTION The results for this section (5.4.2) relate to research question 3.1 and 3.2: RQ3.1 Are there significant differences between early and late adopters (with respect to the costs of IFRS adoption)? RQ3.2 What is the nature of significant differences (if any) that are identified? Descriptive statistics are presented in Table 12. Panels A and B of Table 12 report three sets of benchmark statistics as well as the descriptive statistics relating to the costs of IFRS adoption for EA and LA. Panel C reports the results of statistical tests to determine the significance of differences between EA and LA, with respect to costs of implementing IFRS.54 The benchmark statistics in Panels A and B were derived as follows to provide context against which to consider the costs of IFRS adoption. First, ‘Total Assets CY IFRS ($ 000's)’ are extracted from the total current year assets figures reported under IFRS in the first full-year IFRS financial statements of respondent firms. These figures suggest that characteristics of the reduced samples of EA and LA, for which questionnaire responses are received, are consistent with those of the full samples (e.g., the median for EA is $294.5 million and for LA is $116.1 million). These observations are consistent with earlier observations that early adopters tend to be larger than late adopters (see 5.2.2 and Table 3). Second, the ‘Full accounting team’ information was obtained from responses to Question 7 of the survey instrument: “What is the size of your entire financial accounting team (i.e., persons involved in preparation of financial accounts as opposed to routine data capture/administration tasks)?” These figures provide context against which to view the level of staff and time committed to IFRS adoption. The median figure is three persons for both EA and LA, but most other figures indicate that accounting teams for EA are larger than those of LA (e.g., mean, 75th percentile and maximum). Third, the ‘Pre-IFRS accounting costs (annual av. $)’ information was obtained from responses to Question 8 of the survey instrument: “Before IFRS, what were the direct labour costs of preparing financial accounts for your organisation? (Please estimate by calculating the annual average cost for the 3 years immediately prior to when your organisation formally began incurring IFRS related expenditures).” 54 Costs are measured in terms of staff, time and financial commitments, as well as in terms of the impact of IFRS on the firms’ contracts (e.g., costs of and need to renegotiate contracts). 93 Table 12 Descriptive statistics of sample relating to cost of IFRS implementation Benchmark Staff Commitment Benchmark Time Commitment Benchmark $ Commitment Total Assets c ($ 000's) IFRS team: Initial IFRS team: Maximum Full accounting team a Months planning b Hours implementing Pre-IFRS acctg costs (annual av.) $ Total cost of IFRS adoption $ Panel A: Early Adopters N 17 17 17 17 16 14 10 14 Mean 3,656,435 3.3 4.2 9.5 3.3 1269 143,900 284,286 Std. Deviation 10,680,000 6.9 6.8 14.5 10.6 936 229,614 329,436 Minimum 16,060 1 1 1 -15 300 5,000 5,000 25 Percentile 39,338 1 1 1.5 -6 488 14,500 50,500 Median 294,465 1 2 3 6 950 25,000 180,000 75 Percentile 1,772,332 2.5 4 11.5 12.8 2120 205,000 431,000 Maximum 44,568,000 30 30 60 18 2800 700,000 1,206,000 Panel B: Late Adopters N 17 17 17 16 16 16 12 15 Mean 719,965 2.9 3.4 7.6 9 1242 242,167 131,700 Std. Deviation 1,290,443 2.5 2.5 12.3 14.519 1420 298,794 110,143 Minimum 1,252 1 1 1 -22 70 4,000 5,000 25 Percentile 16,893 1 2 2 6 126 26,250 24,500 Median 116,104 2 3 3 12 850 125,000 100,000 75 Percentile 566,961 3 4 6.8 17 1675 315,000 205,000 Maximum 4,399,400 10 10 50 36 4900 1,000,000 335,000 a = Number of persons involved in preparation of financial accounts as opposed to routine data capture/administration tasks b = Number of months prior to date of transition to NZ IFRS (i.e., start of first comparative period under NZ IFRS) that formal planning for implementation of IFRS began. Negative figures indicate that planning began after date of transition. c = Total current year assets reported under IFRS in first full-year IFRS financial statements 94 Table 12 (cont.) Descriptive statistics of sample relating to cost of IFRS implementation Team Size Initial Team Size Maximum Months Planning Hrs implmnt/ Audit fees Total cost d/ Audit fees Contracts Impact e Panel C: Mann-Whitney U tests f Mean Rank Early 15.06 16.5 14.22 12.07 12.75 13.18 Late 19.94 18.5 18.78 18.5 17.1 12.77 Z statistic 1.51 0.6 1.39 1.995 1.375 0.143 p-value (two tailed) 0.131 0.549 0.165 0.046 0.169 0.887 d = Total cost of IFRS adoption e = Based on an index created to measure the impact of IFRS on contracts of the firms f = The Mann-Whitney U test reports significance of differences in costs/impact between EA and LA 95 These figures facilitate consideration of the total cost of IFRS adoption in relation to the annual average costs of preparing financial accounts prior to IFRS. Almost all of these figures indicate that accounting costs for EA are less than those for LA (e.g., medians are $25,000 for EA and $125,000 for LA). This raises concerns as to the reliability of this data, because as observed earlier, EA tend to be larger than LA and hence it is logical that their accounting costs should follow the same tendency. Questions related to the size of the project teams appointed to manage the implementation of IFRS, reveal that the staff commitment (in number of persons) for EA and LA is similar. Mean, median and percentile figures indicate that, for most EA and LA, between 1 and 3 persons were appointed for the IFRS implementation process. For many EA and LA, this meant involvement of the full accounting team, although this level of commitment reduces for larger firms (i.e., above the 75th percentile). Panel C of Table 12 reports that there are no significant differences between EA and LA as regards the initial or maximum staffing commitments to manage the implementation of IFRS (p = 0.131 and 0.549 respectively). ‘Months planning’ reflects responses to the question: “How many months prior to date of transition to NZ IFRS (i.e., start of first comparative period under NZ IFRS), did your organisation formally begin planning for the implementation of IFRS?” In general, planning by LA was further in advance of the date of transition to IFRS than planning by EA (e.g., EA median is six months while for LA, it is 12 months). The minimum figures indicate that for some firms, planning only commenced well after date of transition (EA: -15 months and LA: -22 months) and for two LA firms, meant that planning began only two months before the period end for their first IFRS full-year financial statements. In spite of the apparent differences between EA and LA noted above, Panel C of Table 12 reports no statistically significant difference between EA and LA as regards the number of months planning for IFRS implementation (p = 0.165). ‘Hours implementing’ reflects responses to the question: “How many person hours (to nearest hundred) do you estimate were spent on the implementation of IFRS?” In the questionnaire, the following guidance was provided to assist respondents: “1800 hours = one person working full time for one year (assumes 40 hour working week less 4 weeks leave, all public holidays, 3 days sick leave)”. Most of the descriptive statistics indicate that EA spent more time implementing IFRS than LA (e.g., EA median of 950 hours in relation to the LA median of 850 hours). Panel C of Table 12 confirms a difference between EA and LA, that is statistically significant, as regards the number of 96 hours spent on the implementation of IFRS (p = 0.046).55 ‘Total cost of IFRS adoption ($NZ)’ reveals that absolute costs for EA were approximately double those of LA (e.g., means, medians and percentiles). The median for EA is $180,000 in relation to $100,000 for LA. This is consistent with earlier observations that EA tend to be larger than LA and that LA should benefit from the experience of those involved with earlier adoption (hence absolute costs are likely to be larger for EA than for LA). The questionnaire sought to gather additional detail regarding the breakdown of these costs. Respondents were asked to indicate both budgeted and actual costs as well as how these were split between system changes (hardware and software); staff training; external consultants; internal personnel costs and other costs. Fewer than 50 percent of respondents budgeted for IFRS implementation costs or incurred costs other than in respect of external consultants and internal personnel, so no further detail in this regard is reported. Untabulated median costs of EA (LA) for external consultants are $75,000 ($40,000) and for internal personnel $45,000 ($50,000). Panel C of Table 12 reports that, after deflating total costs by audit fees, the differences between EA and LA as regards the dollar costs of implementing IFRS are not statistically significant (p = 0.169).56 Respondents were also asked to indicate how much of an impact IFRS had on four categories of contracts (e.g., costs of and need to renegotiate contracts) and to rank these categories in terms of monetary value. The four categories of contracts were debt, revenue, expenses and other; impact rankings were indicated using a five point Likert scale (1 = no impact; 5 = significant impact). Results for this question are not tabulated as most firms indicate that IFRS had little or no impact on their contracts. For EA, no mean values exceed 2 and all median values equal 1; for LA results are similar except that two median values equal 2. In view of the lack of impact on contracts for most firms, fewer than 50 percent of respondents indicated rankings for the monetary values of the contract categories, so no further detail in this regard is reported. 55 ‘Audit fees’ are used as a deflation factor, rather than ‘pre-IFRS accounting costs’ as originally intended, in view of concerns as to the reliability of the ‘pre-IFRS accounting costs’ raised earlier (see discussion of benchmark statistics above). ‘Audit fees’ on the other hand, are verified figures from financial statements and have time, IFRS and size relationships, which support their suitability as a deflation factor. Sensitivity tests are conducted on ‘hours implementation’ and ‘total cost of IFRS adoption’ using ‘pre-IFRS accounting costs’ and ‘total assets’. In both cases, results using ‘audit fees’ as a deflation factor are the only results confirmed by one of the other two sensitivity tests. 56 ‘Audit fees’ are used as a deflation factor for ‘total costs’ (dollars), for the same reasons as are provided in relation to ‘hours implementation’ earlier. 97 Panel C of Table 12 again reports no statistically significant difference between EA and LA, this time as regards the impact on firms’ contracts. To conclude, the hypotheses for research question 3.1 are re-stated and the relevant sub- hypothesis (see bold text below) tested against the above results: H04: There are no significant differences in costs or benefits arising from adoption of IFRS, for EA and LA i.e.: H04a: µCosts of IFRS EA = µCosts of IFRS LA; H04b: µBenefits of IFRS EA = µBenefits of IFRS LA; Overall, hypothesis 4a is accepted, because for five of the six measures of costs of IFRS, no significant differences are found between EA and LA. However, with respect to the sixth measure, concerning the number of hours spent on the implementation of IFRS, a significant difference is found. It is therefore appropriate to address research question 3.2 with regard to the one significant difference noted above: RQ3.2 What is the nature of significant differences (if any) that are identified? As discussed earlier, the descriptive statistics in Table 12, Panels A and B indicate that EA spent more time implementing IFRS than LA. This is consistent with qualitative data gathered during interviews, indicating that for some LA, the motivation for delaying adoption timing was to benefit from the experience of those involved with earlier adoption. Examples of such benefits included specific mentions that systems and knowledge of auditors were improved as a result of their experience with earlier adopters. 98 5.4.3 BENEFITS OF IFRS ADOPTION The results for this section (5.4.3) relate to research questions 3.1 and 3.2: RQ3.1 Are there significant differences between early and late adopters (with respect to the benefits of IFRS adoption)? RQ3.2 What is the nature of significant differences (if any) that are identified? Table 13 reports relevant descriptive statistics, using panels A, B and C to differentiate between EA, LA and significant differences respectively. Respondents were asked to indicate their opinions in respect of nine benefits that are expected to arise as a result of the adoption of IFRS.57 Opinions were indicated using a 5 point Likert scale (1 = ‘strongly agree’ through to 5 = ‘strongly disagree’). Opinions are considered individually, as well as on an overall basis by constructing an ‘overall benefits’ score for each firm.58 Mean and mode values in both panels A and B are predominantly between ‘3’ and ‘4’, indicating that respondents are generally ‘unsure’ or ‘disagree’ as to whether the expected benefits of IFRS will arise. EA are slightly more positive, tending towards 3 (i.e., unsure), while LA tend towards 4 (i.e., disagree). The only benefit which is popularly agreed with is ‘comparability’ for which both EA and LA record modes of 2. However, the mean is 2.29 for EA (i.e., indicating mainly ‘agree’) and 2.88 for LA (i.e., indicating mainly ‘unsure’) and the difference between EA and LA opinions in this regard is significant, as reported in panel C (p = 0.056). Significant differences between EA and LA opinions are also found with regard to expected long-term benefits of IFRS for the individual firms (p = 0.073) and for New Zealand (p = 0.033). LA ‘disagree’ that IFRS will result in long-term benefits for their firms (mean and mode values of ‘4’), while the tendency for EA is to indicate that they are ‘unsure’ with a mean of 3.41 (panel A shows the lowest of multiple mode values to be 3; the other mode value was 4 and each mode accounted for 35.3 percent of respondent opinions). These results are similar but slightly more positive with regard to 57 See Appendix 4, question 9 for details of questions asked regarding the nine expected benefits. 58 The ‘overall benefits’ score is calculated as the sum of the rankings across all nine expected benefits. The result is divided by nine to arrive at an average score which can be interpreted using the Likert scale described above. 99 the long-term benefits of IFRS for New Zealand. A significant difference is also found for the derived ‘overall benefits’ score (p = 0.046). Table 13 Descriptive statistics of sample relating to benefits of IFRS implementation Panel A: Early Adopters Panel B: Late Adopters Panel C: Significance of Differences c Benefit a Mean Std dev. Mode Mean Std dev. Mode Z- score p-value (two tailed) Quality 3.35 0.996 4 3.88 0.993 4 1.524 0.127 Comparability 2.29 1.047 2 2.88 0.928 2 1.914 0.056 Analysis costs 4.12 0.697 4 3.88 0.781 4 1.055 0.291 Capital barriers 3.65 0.996 3 4.07 0.594 4 1.315 0.189 Reporting costs 3.82 1.074 4 3.93 0.917 4 0.147 0.883 User preferences 3.24 1.147 2 3.20 0.941 3d 0.000 1.000 User confidence 3.41 0.870 4 3.88 0.781 4 1.416 0.157 Long-term: Firm 3.41 0.939 3d 4.00 0.791 4 1.796 0.073 Long-term: NZ 2.82 1.185 2 3.65 0.996 3 2.134 0.033 Overall benefits b 3.35 0.566 3.11d 3.67 0.501 3.44d 1.996 0.046 a = Opinions on all benefits were rated on a 5 point Likert scale: 1 = ‘strongly agree’; 2 = 'agree'; 3 = 'unsure'; 4 = 'disagree'; 5 = ‘strongly disagree’. b = An 'overall benefits’ score is calculated for each firm (sum of rankings across all nine expected benefits divided by nine). c = A Mann-Whitney U test reports significance of differences between EA and LA with respect to their opinions on the expected benefits of IFRS. d = Multiple modes exist. The smallest value is shown. To conclude, the hypotheses for research question 3.1 are re-stated and the relevant sub- hypothesis (see bold text below) tested against the above results: H04: There are no significant differences in costs or benefits arising from adoption of IFRS, for EA and LA i.e., H04a: µCosts of IFRS EA = µCosts of IFRS LA; H04b: µBenefits of IFRS EA = µBenefits of IFRS LA; 100 Overall, hypothesis 4b is rejected, because for the constructed ‘overall benefits’ score and for three of the nine measures of benefits of IFRS, significant differences are found between EA and LA. Research question 3.2 is therefore addressed below, with regard to these significant differences, namely ‘comparability’; ‘long-term benefits: firm’; ‘long-term benefits: NZ’ and for the ‘overall benefits score: RQ3.2 What is the nature of significant differences (if any) that are identified? Analysis of comments written on questionnaires as well as the interview data relating to the expected benefits of IFRS revealed that sentiments expressed tended to be general rather than related to specifically identifiable expected benefits. The insights from both EA and LA indicate that there is little evidence of IFRS benefits to date or that this is expected to change in the long term. These findings are consistent with the findings for Table 13 discussed above, as well as the findings relating to discretionary narrative disclosures (see 5.3.4 where ‘evaluative’ narrative disclosures are found to be predominantly negative with respect to IFRS adoption for both EA and LA). LA questionnaire and interview comments however, tend to be more negative than those of EA. EA comments were dominated by the sentiment that IFRS had resulted in a significant amount of extra effort and cost with little or no noticeable benefits. A perception shared by many EA was that much of the extra disclosure and accounting for fair value movements required by IFRS was not read, understood or used by most users of financial statements. A quote from one of the EA illustrates this sentiment: “I think IFRS was the ‘Y2K’ of this decade”.59 Two other EA drew similar comparisons between IFRS and Y2K, indicating that, in their opinion, both projects had resulted in much cost and effort for very little benefit. Other EA described IFRS as “a non-event in relation to the main business of the company” and “a damp squib” in relation to the reaction of shareholders and the market to significant changes in reported profits and financial position as a result of IFRS. 59 ‘Y2K’ is a reference to concerns prior to the year 2000, that there would be widespread computer malfunctions due to time-dependent computer coding which had not taken into account the potential effects of zeros in dates after the end of 1999. These concerns resulted in significant cost and effort to test and upgrade computer systems prior to 2000 to ensure that they continued working, rather than necessarily resulting in improved performance. 101 LA comments reflect sentiments similar to those expressed by EA, but whereas EA are relatively benign or neutral as to the effects of IFRS on users of financial information, LA voice strong concerns in this regard. Comments by LA consistently reflect concern that IFRS is detrimental to the understandability, relevance, reliability and even comparability of financial reports. A number of LA make references to an emerging trend to supplement IFRS information with information regarding ‘underlying profits’ or ‘normalised profit’. This supplementary information typically reverses most IFRS adjustments and fair value movements. A common explanation for the emergence of this trend is that this supplementary information is being requested by boards, analysts and users in preference to IFRS information, in order to better understand the results of operational activities as opposed to valuation adjustments. Of greater concern, is the view expressed by a number of LA that certain IFRS requirements result in misleading information, less accurate indications of financial performance and financial position and even situations where financial statements no longer reflect a true and fair view of financial affairs. Particular reference was made to IAS 12 requirements which resulted in recognition of material deferred tax liabilities due to a change in certain depreciation allowances in 2010 in New Zealand.60 The following quote from one LA interview is representative of the sentiments in this regard: “I think that got highlighted as an issue more recently with the income tax depreciation and the government removing depreciation on long lived buildings. And therefore following IAS 12 in its exact form required quite significant increases in deferred tax liabilities, running through the income statement, which if a user today asked me: ‘When are we going to pay that $100 odd million of deferred tax?’ the likely answer to that is: ‘Probably never’, and so the answer to that is: ‘Why are we recognising a liability if we are never going to pay it?’” Quotes from different respondents illustrating concerns about comparability include: “ … there is much more volatility in the income statements these days some of which is non-cash and may not be important to some users of the financial statements but also makes comparability between years a lot more difficult …”; and: 60 See news release retrieved from http://www.scoop.co.nz/stories/BU1008/S00686/reports-skewed-by- deferred-tax-liabilities.htm on 5/4/2011, for details of companies affected and estimate that this requirement resulted in an accumulated distortion detrimental to the profits of New Zealand listed companies in excess of one billion dollars. 102 “Comparability - you can’t even say that retail is comparable, basically because everyone has a different interpretation.” Two quotes extracted from interviews which illustrate concern expressed that the extent of disclosure and technical terminology is compromising the quality and understandability of financial reporting since IFRS, include: “Look I buy the concept of trying to get everybody onto the same playing field. I think we’ve got to a position where we are bogged down with so much detail, half the time I don’t understand what the standard is trying to tell us to do and report. I don’t know how people in the street have got a bloody clue;” and: “I actually kind of feel that when I meet with our shareholders once a year … most of them say they don’t even it read it now. It’s just become such a big volume book that they don’t even understand half of what’s there and some of the notes like the derivatives and things like that take a long time to do the work on and actually the shareholders don’t even understand what it’s about.” To conclude this section, written comments regarding expected benefits have been extracted from the final page of five different LA questionnaires as they provide a good summary of the sentiments expressed in LA interviews, as well as highlighting the stronger concerns of LA relative to EA: • “We have begun reporting underlying profit to both the board and shareholders which essentially removes most of the IFRS adjustments. The Group’s debt covenants have also been restructured to remove IFRS adjustments associated with financial instruments.” • “Comments from the users of our annual report indicate that there is far too much complicated information provided in the annual report as a result of IFRS requirements, in particular, Financial Instruments which does not add any value to the investors and readers. It results in the readers being overloaded with information which they cannot understand unless they are accountants. Further, it has not only increased the size of the annual report at a cost to the investors, but they no longer reflect a true and fair view of the financial affairs of the business.” • “NZIFRS has confused our investors, Chairman of the Audit Committee & Canadian investors. It has significantly increased costs & is perceived to be of no benefit.” • “The only benefit has been comparability of financial statements between organisations. Some of the requirements add no value whatsoever particularly the disclosures required 103 by IFRS 7. Analysts ignore these and they just confuse 99% of readers of financial statements.” • “Overall, IFRS has been a waste of time. In particular IFRS 41 is rendering the financial statements of a large number of New Zealand companies irrelevant. In my experience the cost of explaining financial statements to shareholders and analysts has increased. In many cases, statements of cash flows have increased in relevance at the expense of the P&L and balance sheets, which are now considered to be less accurate indications of a business’s performance and financial position. Increasingly, we are relying on non- financial information to communicate with investors.” 5.4.4 ADDITIONAL INSIGHTS AS TO POTENTIAL ADOPTION TIMING MOTIVATIONS Table 14 presents a thematic analysis of the responses to the open-ended question used to conclude interviews with early and late adopters.61 The information reported in Table 14 for the ‘importance’, ‘potency’ and ‘evaluative’ themes adds little to the results discussed earlier for content analysis of discretionary narrative disclosures in annual reports (see 5.3.3 discussion of Tables 9 and 10). This discussion therefore focuses on the three subsequent themes. Responses relating to the ‘activity’ theme indicate very little difference between EA and LA (‘dynamic’: 11 EA and 10 LA; ‘static’ 6 EA and 5 LA). ‘Dynamic’ EA responses typically demonstrated a positive, proactive approach to change and hence a desire to adopt IFRS early. ‘Dynamic’ LA responses demonstrated similarly proactive approaches, but management had typically concluded that deferring adoption was in the best interests of the firm (e.g., in order to learn from and leverage off the experience of early adopters or to allow sufficient time to “…ensure that a rigorous process was put in place, necessary systems changes were made and properly informed decisions made …”). ‘Static’ EA responses typically indicate that early adoption was driven by a need to align with group reporting requirements for related parties who were Australian or were New Zealand public sector entities. ‘Static’ LA responses tended to indicate a 61 “Adoption of IFRS was possible from 1 January 2005 (on a voluntary basis); mandatory from 1 January 2007. Your organisation elected to adopt early/ wait until it became mandatory. Could you give me your insights/background as to why early/ late?” See final section of Chapter 4.4.3 for further details on research method used to conduct content analysis of responses to this question. 104 preference to follow others and delay adoption until it became mandatory. Table 14 Thematic analysis of interview responses regarding adoption timing motivations No. of Responses Early Adopters Late Adopters (n = 13) (n = 13) Importance High 3 1 Low 1 2 Potency High 6 0 Low 0 2 Evaluative Neutral 1 2 Negative 0 5 Positive 0 0 Activity Dynamic 11 10 Static 6 5 Manageability Expected 4 0 Unexpected 1 5 Accounting Choice Yes 9 6 No 21 14 The ‘manageability’ theme highlights striking differences between EA and LA. For EA, four out of five responses indicate perceptions that IFRS adoption would proceed in line with expectations (i.e. ‘expected’). These responses included perceptions that requirements for late adopters would become more onerous and less flexible and that promises of a stable platform of IFRS standards were encouraging. In contrast, all five LA responses indicate that ‘unexpected’ factors influenced their decision to delay IFRS 105 adoption. Four of these responses expressed concern at the extent of continuing change to IFRS standards (e.g. “Our sense was it was a bit of a moving feast and so we wanted to wait until, not only we had a bit more of a sense for what other people were doing, but also our advisors (Big 4 audit firm) had a sense of what the practice was becoming”). Finally, the ‘accounting choice’ theme suggests that, while accounting choice theory does have substantial explanatory power, it explains less than one third of responses regarding motivations for adoption timing decisions. For EA, nine out of 30 (LA six out of 20) responses indicate that the early adoption decision was influenced by the effect of IFRS on accounting information. Examples of such responses for EA include four responses which relate to the income increasing effect of discontinuing amortisation of goodwill. Examples for LA include three responses indicating that they deferred adoption, because their accounting information would be influenced by the way in which IFRS was adopted by competitors and other earlier adopters. A further two LA indicated that a motivation for late adoption was that it deferred perceived adverse effects of IFRS on the quality of their financial statements. Responses which indicate that adoption timing had been motivated by factors other than the impact of IFRS on accounting information may be summarised as follows: 62 • For EA, five of the 21 responses indicate that early adoption was motivated by a positive, proactive management approach (i.e., associated with the ‘dynamic’ activity theme). A further six responses concern relationships with New Zealand public sector entities or Australian companies (i.e., associated with the ‘static’ activity theme for EA). Two responses relate to a desire to be seen as market leaders (i.e., associated with the ‘importance’ theme) and a further two concern a perception that early adoption would afford better access to consultants, while few were demanding their services. • For LA, three of the 14 responses relate to a need to defer adoption in order to ensure sufficient time for preparation and planning prior to implementation of IFRS (i.e., associated with the ‘dynamic’ activity theme). A further three responses cited deferral as being preferable in view of perceived adverse consequences of adopting IFRS, while two more responses simply perceived no benefit in early adopting. A further two responses indicated that adopting IFRS was a low priority for them (i.e., associated with ‘low’ importance). 62 Non-recurring responses are not included in the summary. 106 5.5 CHAPTER SUMMARY This chapter reports the results of research which aims to investigate what motivates entities to adopt IFRS early or late. In particular, three research questions are investigated: 1. What impact does IFRS adoption have on the financial information of early as opposed to late adopters? 2. What do discretionary narrative disclosures in annual reports reveal about early and late adopter firms’ attitudes to IFRS adoption? 3. What are the costs and benefits of IFRS adoption for early as opposed to late adopters? In addressing each of these research questions, two main sub-questions are investigated: a) Are there significant differences between early and late adopters? b) What is the nature of significant differences (if any) that are identified? These research questions and sub-questions are investigated by using what Creswell (2003) describes as a ‘Mixed Methods Approach’, as it employs both quantitative and qualitative research methods. The results relating to the impact of IFRS adoption on financial information provide evidence of significant changes to most financial statement elements and ratios for both early and late adopters. However, no evidence is found of statistically significant differences when the changes for early adopters are compared to those for late adopters. In addition, the nature of significant changes to financial statement elements arising from the adoption of IFRS was investigated, distinguishing between results for early and late adopters. This analysis begins by identifying the accounting standards found to be the most frequent sources of IFRS adjustments and providing common explanations of the adjustments. Further analysis identifies that for most EA and LA firms a specific standard of IFRS has no impact, but that a specific standard can be very material for a small number of firms. Positive effects are more frequent than negative effects (i.e., IFRS tends to increase values of financial statement elements). Although effects on assets and equity are generally more widespread for EA than for LA, only two (out of a possible 40) significant differences between EA and LA are found regarding IFRS adoption adjustments: adjustments to expenses relating to financial instruments and 107 adjustments to liabilities relating to revenue recognition. The examination of discretionary narrative disclosures about IFRS adoption in annual reports, however, provides much stronger evidence of significant differences between early and late adopters. Form oriented analysis reveals significant differences for all four measures of the extent of discretionary narrative disclosures related to IFRS adoption (e.g., the number of words and the amount of space devoted to these disclosures). The extent of IFRS adoption disclosures for early adopters is typically more than double that of late adopters. Meaning oriented analysis identifies three major themes: ‘importance’, ‘potency’ and ‘evaluative’. ‘Informing of importance’ of IFRS adoption is the most dominant theme. Differences between early and late adopters as regards ‘Importance’ and ‘Potency’ themes are statistically significant. Multivariate analysis provides evidence that early adoption persists as a significant explanatory variable, after controlling for other incentives for voluntary disclosure, such as firm size, auditor and industry. ‘Evaluative’ narrative disclosures are made by relatively few firms and are found to be predominantly negative regarding IFRS adoption. There is no statistically significant difference in this regard between early and late adopters. Qualitative information such as theme descriptions and examples of disclosures are provided to afford deeper insight into the nature of the significant differences between early and late adopters. The costs and benefits of IFRS adoption are then examined, as well as motivations for adoption timing. For five of the six measures of costs of IFRS considered, no significant differences are found between early and late adopters. However, with respect to the sixth measure, concerning the number of hours spent on the implementation of IFRS, a significant difference is found: early adopters spent more time implementing IFRS than late adopters. This is consistent with qualitative data gathered during interviews, indicating that for some late adopters, the motivation for delaying adoption timing was to benefit from the experience of those involved with earlier adoption. As regards the benefits of IFRS, significant differences are found between early and late adopters for a constructed ‘overall benefits’ score and for three of nine individual benefits assessed. Qualitative data gathered during interviews, suggests that these significant differences are due to relatively benign or neutral attitudes by early adopters as to the effects of IFRS on users of financial information; whereas late adopters voice strong concerns in this regard. Concerns that IFRS may be detrimental to the understandability, relevance, reliability and even comparability of financial reports are 108 prevalent amongst late adopters but not early adopters. Finally, a thematic analysis of interview responses is presented which provides additional insights regarding motivations for adoption timing. These insights relate mainly to findings for three new themes which emerge, namely ‘activity’, ‘manageability’ and ‘accounting choice’. Little difference is found between early and late adopters regarding the ‘activity’ theme with approximately twice as many responses indicating ‘dynamic’ as opposed to ‘static’ approaches to IFRS adoption by firms. Results are similar for ‘accounting choice’, where responses indicate that ‘other factors’ are twice as likely to influence adoption timing, than the impact of IFRS on accounting information. However, responses relating to the ‘manageability’ theme suggest that, while most early adopters were confident about what to expect from IFRS, late adopters had concerns regarding the ‘unexpected.’ 109 CHAPTER 6: CONCLUSIONS 6.1 SUMMARY AND IMPLICATIONS OF FINDINGS IFRS is now permitted or required in over 120 countries.63 It is a recent phenomenon and is regarded as one of the most significant regulatory changes in accounting history. This study exploits the New Zealand setting to investigate accounting choices made by firms with respect to the timing of IFRS adoption. Companies listed on the New Zealand Stock Exchange were required to adopt IFRS for periods beginning on or after 1 January 2007, but unlike many other countries (e.g., Australia and European Union countries) they were permitted the choice of voluntarily adopting IFRS up to two years earlier. To investigate what motivates entities to adopt IFRS early or late, mixed methods research (quantitative and qualitative methods) is applied to three different sources of information. Although the sample size is small (n=80), all early adopters are considered (40) as well as an equal number of randomly selected late adopters. First, the impact of IFRS on the financial information of early as opposed to late adopters is investigated. This research provides evidence of significant differences in most financial statement elements and ratios for both early and late adopters. The magnitude of these differences and the extent to which individual firms are affected varies considerably. There is however, little evidence of statistically significant differences when the IFRS differences for early adopters are compared to those for late adopters. These findings have implications for regulators, firms, analysts and others in countries which have signalled an intention to adopt IFRS in the future, but have yet to do so and may be useful in informing their decisions. The findings will be particularly pertinent for countries with characteristics similar to those of New Zealand (e.g., GAAP similar to IFRS; strong institutional arrangements, enforcement and reporting incentives). Further implications which are specific to the literature relating to IFRS and accounting choice are considered in the section below (6.2). 63 Extracted with reference to http://www.iasplus.com/country/useias.htm#totals on 9/9/2011. 110 Second, discretionary narrative disclosures in annual reports are investigated to assess what they reveal about early and late adopter firms’ attitudes to IFRS adoption. Form oriented as well as meaning oriented content analysis is employed.64 Significantly more disclosure is found for early adopters relative to late adopters for all four measures of extent of disclosure. There is, however, relatively little narrative disclosure overall (e.g., none at all for fourteen late adopters and seven early adopters and generally less than a quarter of a page of the annual report for remaining firms). To the extent that such form oriented analysis reflects the importance of an event or issue, the adoption of IFRS appears to have been a relatively minor event in relation to overall business operations of sample firms, particularly for late adopters. However, the ‘form oriented’ dimension of content analysis is generally accepted as being less informative than the ‘meaning oriented’ dimension. Meaning oriented analysis of narrative disclosures reveals three main themes, namely ‘informing of importance’, ‘potency’ and ‘evaluative’. ‘Informing of importance’ of IFRS adoption, is found to be the most dominant theme. Furthermore, descriptions of IFRS impacts are found to be predominantly ‘high potency’.65 ‘Evaluative’ narrative disclosures are less frequent and found to be predominantly negative concerning IFRS adoption for both early and late adopters. Hence, IFRS related disclosure indicates that firms attach importance to the adoption of IFRS, but tend to have negative perceptions of the consequences of adoption. This has implications for the continuing acceptance of IFRS in its current form and emphasises the importance of ongoing improvement efforts by the International Accounting Standards Board (IASB), such as reducing disclosure requirements. Further implications which are specific to the literature relating to discretionary narrative disclosures and accounting choice are considered in the section below (6.2). Third, early and late adopters are surveyed to collect data relating to the costs and benefits of IFRS adoption, as well as additional insights regarding the motivations for their adoption timing decisions. This study provides results for six different measures of the cost of implementing IFRS, as well as the perceptions of senior financial executives relating to nine expected benefits of IFRS (see 6.2 below for further detail). Further work is required to properly attempt to assess whether the benefits of IFRS exceed the 64 Form oriented analysis is concerned with extent of disclosures, while meaning oriented analysis is concerned with the nature of disclosures. 65 ‘High potency’ narratives reflect specific detail regarding the impact of IFRS; ‘low potency’ narratives are vague, general or report immaterial impacts. 111 costs, however there are some indications that, in the short term, this may not be the case. The findings have planning and budgeting implications for firms who are to adopt IFRS in the future and may be useful in informing their decisions. A further implication is the need for regulators to carefully manage expectations of firms to take a long-term view of the cost-benefit relationship in regard to IFRS adoption. The results also provide additional support for earlier comments regarding implications for the continuing acceptance of IFRS in its current form and the importance of ongoing improvement efforts by the IASB. Further implications which are specific to the literature relating to accounting choice are considered in the section below (6.2). 6.2 REFLECTIONS ON RESULTS AS LINKED TO THE LITERATURE This study draws on accounting choice theory to provide a framework for investigating adoption timing choices. Literature relevant to this theory suggests that the primary purpose of an accounting choice decision is to influence accounting outputs in a particular way. Such literature suggests that accounting choices can be used in contrasting ways, either to reduce or to exploit information asymmetries. Much of this prior literature is restricted to quantitative analysis of financial statements and capital markets data. The use of mixed methods research as explained above (6.1) contributes towards a broader, deeper and richer understanding of adoption timing in the context of the accounting choice literature. The findings related to the impact of IFRS on financial information (IFRS differences) and the nature of the most common IFRS adjustments are broadly consistent with those of Hung & Subramanyam (2007), Kabir et al., (2010) and Stent et al., (2010). However, differences between IFRS and local GAAP are more pronounced for German firms than for New Zealand firms. The significant IFRS differences found for most financial statement elements and ratios, for both early and late adopters, may provide incentives for adoption timing decisions. To elaborate, IFRS differences which affect information such as earnings or leverage, may influence adoption timing decisions (because in general, prior literature suggests that efficient contracting incentives such as bonuses do work – see Chapter 3). There is however, little evidence of statistically significant differences when the changes for early adopters are compared to those for late adopters. The findings for this study therefore do not support earlier studies which focus on 112 investigation of the incentives inherent in the bonus, debt and political cost hypotheses (e.g., Brown, 1985; Brown & Brandi, 1986; Langer & Lev, 1993). Rather, they suggest the need to take the wider view of the interaction between incentives and other factors (e.g., firm characteristics) which is evident in later studies, when investigating accounting choices such as adoption timing (e.g., Hung & Subramanyam, 2007; Iatridis & Joseph, 2006; Li, 2010). Prior literature related to discretionary narratives in annual reports suggests that such disclosures communicate important information (Beattie et al., 2004; Beretta & Bozzolan, 2004; Clatworthy & Jones, 2003; Smith & Taffler, 2000) and provide a tool for management to signal its response to important events such as IFRS adoption (Stanton & Stanton, 2002). Documentary disclosures represent data that firms have taken care and thought to compile (Creswell, 2003) and the annual report reflects firms’ views regarding the type, amount and importance of data (Niskala & Pretes, 1995). Findings for this study indicate significantly more disclosure for early adopters than for late adopters, as well as significant differences regarding themes, notably the dominant ‘Informing of importance’ theme. Multivariate analysis of these themes provides further evidence that early adoption persists as a significant explanatory variable, after controlling for other incentives for voluntary disclosure, such as firm size, auditor and industry. These findings triangulate to indicate that early adopters attach a higher level of importance to the adoption of IFRS than late adopters and that firm attitudes in this regard influence the priority given to adoption timing. Prior literature related to costs and benefits of IFRS adoption as well as further adoption timing insights suggests a concern by the IASB and regulators in New Zealand that the benefits of reporting should exceed the costs (Bradbury & van Zijl, 2006; Dunstan, 2002; IASB, 2004). Many recent studies contribute to a growing body of evidence relating to expected benefits of IFRS (e.g., effects on liquidity, interest costs and accounting quality were reviewed in Chapter 3). However, evidence relating to costs and perceptions regarding such benefits are rare. Griffin et al. (2009) find an increase in audit fees associated with IFRS, Li (2010) finds average compliance costs for IFRS adoption of £360,000–£625,000 and a PriceWaterhouseCoopers (2006) survey finds that only 14% of senior executives surveyed had positive perceptions regarding benefits arising from IFRS adoption. Contributions to the literature from the current study include descriptive statistics such as median figures which provide an indication of costs for EA (LA): IFRS initial team size of 1 (2); IFRS maximum team size of 2 (3); months planning prior to date of transition to IFRS of 6 (12); hours spent implementing 113 IFRS of 950 (850) and total dollar costs of IFRS adoption of $180,000 ($100,000). Cost is also measured in terms of the impact that IFRS had on firm contracts (e.g., renegotiation), but most firms indicate little or no such impacts. Results for the survey of benefits indicates that only one expected benefit rates popular agreement, namely ‘comparability’. For all other expected benefits surveyed, firms tend to indicate that they are unsure or disagree. A significant difference between early and late adopters is found for only one of six measures of costs, but for three of nine expected benefits as well as for an ‘overall benefits’ score. Qualitative data in questionnaires and from interviews suggests that late adopters incur lower costs, because they have the advantage of being able to leverage off the experience of early adopters. Late adopters’ opinions regarding expected benefits of IFRS are generally more negative than those of early adopters. Qualitative data in this regard indicates that late adopters have concerns that IFRS is a detrimental change, while early adopters are more inclined to the view that there is disappointingly little change as a result of IFRS, in view of the level of cost and effort required. Concerns raised by late adopters indicate that many users disregard or are confused by the increased volume and complexity of disclosures and fair value movements which are required by IFRS. A more serious concern raised is that certain IFRS requirements have the potential to mislead users. These concerns are said to be resulting in an increasing tendency to rely on non-financial information, cash flow information and alternative financial information (e.g., supplementary reporting of ‘underlying’ or ‘normalised’ profits, which remove many IFRS adjustments). These survey findings therefore suggest that significant differences, which exist between early and late adopters as regards the costs and perceived benefits of IFRS adoption, may affect adoption timing decisions. The survey also reveals additional insights regarding motivations for adoption timing. Approximately twice as many responses relating to an ‘activity’ theme indicate ‘dynamic’ as opposed to ‘static’ approaches to IFRS adoption by both early and late adopters. Results for this theme provide important evidence contrary to prior presumptions made on the basis of actions (i.e., that early adopters are ‘dynamic’ and late adopters are ‘static’). Responses relating to a ‘manageability’ theme suggest that early adopters were relatively confident about what to expect regarding IFRS adoption (‘expected’ sub-theme); while late adopters had concerns in this regard (e.g., lack of a stable platform of IFRS standards) which influenced their decision to delay adoption (‘unexpected’ sub-theme). Finally, responses relating to an ‘accounting choice’ theme suggest that ‘other factors’ are twice as likely to influence adoption timing, as those 114 relating to the impact of IFRS on accounting information. These additional insights add depth and richness to the earlier findings, highlighting the benefits of using both quantitative and qualitative methods. In conclusion, the findings triangulate to suggest that ‘accounting choice’ has less explanatory power, with regard to adoption timing decisions, than ‘other factors’ which are unrelated to the impact of IFRS on accounting information. First, IFRS differences are not significantly different for EA and LA; second, discretionary narrative disclosures suggest that EA attach more importance to IFRS adoption than LA; and third, survey results suggest that adoption timing may be influenced by cost and various other factors, many of which are not ‘accounting choice’ related decisions. Further examples of such ‘other factors’ include evaluations of the consequences of IFRS (predominantly neutral or negative) and perceptions as to the manageability of IFRS adoption (unexpected factors influence deferral). A further important finding is that adoption timing decisions of both early and late adopters are found to be predominantly ‘dynamic’ (proactive) rather than ‘static’ (passive) activity. 6.3 LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH 6.3.1 Limitations While mixed methods research exploits some of the benefits of both quantitative and qualitative research, it also carries with it some of the limitations of each: • Accounting Choice Theory offers a rich and appropriate theoretical framework for this study. It does, however, also have the potential to introduce researcher bias towards this theory rather than being open to other theories which may be suggested or inferred during qualitative discovery processes. • Data collection for qualitative research is laborious and time consuming and typically results in relatively small sample sizes. The sample sizes used in this study are relatively large in relation to many other qualitative studies, but are relatively small in relation to many quantitative studies. As a result most data is not normally distributed and so there are implications for use of statistical methods and generalisation of results. 115 Similarly, while the New Zealand setting for this study offers certain unique benefits, it also carries with it certain limitations: • Listed companies in New Zealand tend to be smaller than companies in many other countries. • They may also have other unique characteristics that were not identified in this study and hence may not be representative of other early or late adopters around the world. • Timing of data collection was necessarily spread over a number of years. Collection of financial statement information and discretionary narratives began in early 2006 as the first full-year IFRS financial statements of early adopters began emerging and ended three years later in early 2008, when financial statements of the last late adopters became available. Collection of survey information followed a similar pattern but lagged so that annual report information could be considered in preparation for interviews. Results could therefore potentially be affected by time-related changes (e.g., changes to IFRS standards since 2005, changes in economic conditions affecting fair values and changes in firm personnel involved in IFRS implementations). Finally, this study considers motivations for IFRS adoption on the basis of short- term data gathered from annual reports for the year in which sample firms adopted IFRS and from surveys soon after adoption. Further insights are likely to emerge in the long run. 6.3.2 Suggestions for further research As a recent and global phenomenon, IFRS offers rich opportunities for research and such research is still at an early stage. The suggestions offered below for further research flow from the limitations of this thesis only and are a small sample of much wider possibilities. Mixed methods research is a relatively little used avenue in accounting research. It offers a pragmatic approach to scientific enquiry which allows considerable flexibility and benefits. Where collection of diverse types of data is likely to offer the best understanding of a research problem, future accounting researchers may find that developing this approach provides the potential to make significant 116 contributions to existing literature. The research results presented in this paper relate to companies listed on the New Zealand Stock Exchange. Similar research on entities in other countries which have been or are to be offered the choice of early adoption of IFRS would undoubtedly provide further useful contributions to the accounting choice literature. For example, in the UK, AIM quoted companies were not required to produce accounts under IFRS until accounting periods beginning on or after 1 January 2007 (UK Financial Reporting Review Panel, 2007, p1), hence offering the potential to extend findings for this study to the UK.66 In addition to extending the scope of this research beyond New Zealand, it could also be extended to consider factors which may become evident over a longer period, which provide further insights regarding adoption timing decisions. 66 AIM is the London Stock Exchange’s international market for smaller growing companies (extracted from http://www.londonstockexchange.com/companies-and-advisors/aim/aim/aim.htm on 19 September 2011) 117 REFERENCES Accounting Standards Review Board. (2009). Proposed Application of Accounting and Assurance Standards under the proposed new Statutory Framework for Financial Reporting (Discussion Document). Wellington: Accounting Standards Review Board. Alford, A., Jones, J., Leftwich, R., & Zmijewski, M. (1993). The Relative Informativeness of Accounting Disclosures in Different Countries. Journal of Accounting Research, 31(3), 183-223. Ali, A., & Kumar, K. R. (1994). The Magnitudes of Financial Statement Effects and Accounting Choice - the Case of the Adoption of SFAS 87. Journal of Accounting & Economics, 18(1), 89-114. Amir, E., & Ziv, A. (1997a). Economic Consequences of Alternative Adoption Rules for New Accounting Standards. Contemporary Accounting Research, 14(3), 543-568. Amir, E., & Ziv, A. (1997b). Recognition, Disclosure, or Delay: Timing the Adoption of SFAS No. 106. Journal of Accounting Research, 35(1), 61-81. Andersen, BDO, DeloitteToucheTohmatsu, Ernst&Young, GrantThornton, KPMG, et al. (2001). GAAP 2001: A Survey of National Accounting Rules Benchmarked against International Accounting Standards. Armstrong, C. S., Barth, M. E., Jagolinzer, A. D., & Riedl, E. J. (2010). Market Reaction to the Adoption of IFRS in Europe. Accounting Review, 85(1), 31-61. Ashbaugh, H., & Olsson, P. (2002). An Exploratory Study of the Valuation Properties of Cross- Listed Firms' IAS and U.S. GAAP Earnings and Book Values. Accounting Review, 77(1), 107. Auer, K. V. (1996). Capital market reactions to earnings announcements: empirical evidence on the difference in the information content of IAS-based earnings and EC-Directives- based earnings. European Accounting Review, 5(4), 587-623. Austin, L. (1997). Accounting for Intangible Assets. University of Auckland Business Review, 9(1), 63-72. Ayres, F. L. (1986). Characteristics of firms electing early adoption of SFAS 52. Journal of Accounting and Economics, 8(2), 143-158. Ball, R., Kothari, S. P., & Robin, A. (2000). The effect of international institutional factors on properties of accounting earnings. Journal of Accounting and Economics, 29(1), 1-51. Balshaw, V., & Lont, D. (2010). New Zealand operating expense disclosure. Pacific Accounting Review (Emerald Group Publishing Limited), 22(2), 108-123. Barth, M. E., Clinch, G., & Shibano, T. (1999). International accounting harmonization and global equity markets. Journal of Accounting & Economics, 26(1-3), 201-235. Barth, M. E., Landsman, W. R., & Lang, M. H. (2008). International Accounting Standards and Accounting Quality. Journal of Accounting Research, 46(3), 467-498. Beattie, V., McInnes, W., & Fearnley, S. (2004). A methodology for analysing and evaluating narratives in annual reports: a comprehensive descriptive profile and metrics for disclosure quality attributes. Accounting Forum, 28(3), 205-236. Benjamin, J. J., Grossman, S. D., & Wiggins, C. E. (1986). The Impact of Foreign Currency Translation on Reporting during the Phase-in SFAS No. 52. Journal of Accounting, Auditing & Finance, 1(3), 177-184. Beretta, S., & Bozzolan, S. (2004). A framework for the analysis of firm risk communication. The International Journal of Accounting, 39(3), 265-288. Bishop, H., Bradbury, M., & van Zijl, T. (2005). NZ IAS 32: An Evaluation of the Potential Impact on Financial Reporting by Issuers of Convertible Financial Instruments. Pacific Accounting Review (Pacific Accounting Review Trust), 17(2), 34-52. Boesso, G., & Kumar, K. (2007). Drivers of corporate voluntary disclosure. Accounting, Auditing & Accountability Journal, 20(2), 269-296. 118 Botosan, C. A. (1997). Disclosure level and the cost of equity capital. Accounting Review, 72(3), 323. Bradbury, M. (1998). Harmonising with Oversesas Accounting Standards: A New Zealand Perspective. Australian Accounting Review, 18-23. Bradbury, M., & van Zijl, T. (2006). Due Process and the Adoption of IFRS in New Zealand. Australian Accounting Review, 16(39), 86-94. Brown, B. (1985). The Relationship between Firm Attributes and Early Adoption of the Foreign Currency Translation Standard SFAS No. 52: An Empirical Investigation. International Journal of Accounting, 21(1), 1-19. Brown, B., & Brandi, J. T. (1986). Security Price Reactions to Changes in Foreign Currency Translation Standards. Journal of Accounting, Auditing & Finance, 1(3), 185-205. Byard, D., Li, Y., & Yu, Y. (2011). The Effect of Mandatory IFRS Adoption on Financial Analysts' Information Environment. Journal of Accounting Research, 49(1), 69-96. Cavana, R. Y., Delahaye, B. L., & Sekaran, U. (2001). Applied Business Research: Qualitative and Quantitative Methods. Milton, Queensland: John Wiley & Sons. Chan, K. C., & Seow, G. S. (1996). The association between stock returns and foreign GAAP earnings versus earnings adjusted to U.S. GAAP. Journal of Accounting & Economics, 21(1), 139-158. Cheong, C. S., Sujin, K., & Zurbruegg, R. (2010). The impact of IFRS on financial analysts' forecast accuracy in the Asia-Pacific region. Pacific Accounting Review (Emerald Group Publishing Limited), 22(2), 124-146. Choi, F. (1973). Financial disclosure in relation to a firm’s capital costs. Accounting and Business Research, 4(1), 159-175. Christensen, H. B., Lee, E., & Walker, M. (2009). Do IFRS Reconciliations Convey Information ? The Effect of Debt Contracting. Journal of Accounting Research, 47(5), 1167-1199. Christie, A. A. (1990). Aggregation of test statistics: An evaluation of the evidence on contracting and size hypotheses. Journal of Accounting and Economics, 12(1-3), 15-36. Clatworthy, M. A., & Jones, M. J. (2003). Financial reporting of good news and bad news: evidence from accounting narratives. Accounting and Business Research, 33(3), 171- 185. Covrig, V. M., DeFond, M. L., & Hung, M. Y. (2007). Home bias, foreign mutual fund holdings, and the voluntary adoption of international accounting standards. Journal of Accounting Research, 45(1), 41-70. Creswell, J. W. (2003). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches (Second ed.): Sage Publications. D'Souza, J. M. (1998). Rate-Regulated Enterprises and Mandated Accounting Changes: The Case of Electric Utilities and Post-Retirement Benefits Other than Pensions (SFAS No. 106). Accounting Review, 73(3), 387. Daske, H., Hail, L., Leuz, C., & Verdi, R. (2008). Mandatory IFRS Reporting around the World: Early Evidence on the Economic Consequences. Journal of Accounting Research, 46(5), 1085-1142. Deloitte. (2003). International Financial Reporting Standards Healthcheck 2003: Deloitte. Dunstan, K. L. (2002). The Case for the Use of International Financial Reporting Standards in New Zealand : A Briefing Paper prepared on behalf of the New Zealand Securities Commission. Wellington, N.Z.: Centre for Accounting Governance and Taxation Research School of Accounting and Commercial Law Victoria University of Wellington. Dye, R. A. (1985). Disclosure of Nonproprietary Information. Journal of Accounting Research, 23(1), 123-145. Ernst&Young. (2004). IFRS Impact Report: Ernst & Young. Fields, T. D., Lys, T. Z., & Vincent, L. (2001). Empirical research on accounting choice. Journal of Accounting and Economics, 31(1-3), 255-307. Goodwin, J., Ahmed, K., & Heaney, R. (2008). The Effects of International Financial Reporting Standards on the Accounts and Accounting Quality of Australian Firms: A 119 Retrospective Study Journal of Contemporary Accounting & Economics, 4(2), 89-119. Griffin, P. A., Lont, D. H., & Sun, Y. (2009). Governance Regulatory Changes, International Financial Reporting Standards Adoption, and New Zealand Audit and Non-Audit Fees: Empirical Evidence. Accounting and Finance, 49(4), 697 - 724. Guba, E. G., & Lincoln, Y. S. (1989). Fourth Generation Evaluation: Sage Publications. Harris, M. S., & Muller, K. A. (1999). The market valuation of IAS versus US-GAAP accounting measures using Form 20-F reconciliations. Journal of Accounting and Economics, 26(1-3), 285-312. Hickey, L., Spencer, J., van Zijl, T., & Perry, J. (2003). Adoption of IFRS - Background and Process. Chartered Accountants Journal of New Zealand, 82(6), 4-7. Holsti, O. R. (1969). Content Analysis for the Social Sciences. Reading, Massachussetts: Addison-Wesley. Holthausen, R. W. (1990). Accounting method choice: Opportunistic behavior, efficient contracting, and information perspectives. Journal of Accounting and Economics, 12(1- 3), 207-218. Holthausen, R. W., & Leftwich, R. W. (1983). The economic consequences of accounting choice implications of costly contracting and monitoring. Journal of Accounting and Economics, 5, 77-117. Hung, M., & Subramanyam, K. R. (2007). Financial statement effects of adopting international accounting standards: the case of Germany. Review of Accounting Studies, 12(4), 623- 657. IASB. (2004). International Financial Reporting Standards (IFRSs). London: International Accounting Standards Board Iatridis, G. E., & Joseph, N. L. (2006). Characteristics of UK firms related to timing of adoption of Statement of Standard Accounting Practice No. 20. Accounting and Finance, 46(3), 429-455. Jermakowicz, E. K., Prather-Kinsey, J., & Wulf, I. (2007). The Value Relevance of Accounting Income Reported by DAX-30 German Companies. Journal of International Financial Management & Accounting, 18(3), 151-191. Jones, M. J., & Shoemaker, P. A. (1994). Accounting Narratives: A review of Empirical Studies of Content and Readability. Journal of Accounting Literature, 13, 142-184. Jones, S., & Higgins, A. D. (2006). Australia's switch to international financial reporting standards: a perspective from account preparers. Accounting and Finance, 46(4), 629- 652. Kabir, M. H., Laswad, F., & Islam, M. A. (2010). Impact of IFRS in New Zealand on Accounts and Earnings Quality. Australian Accounting Review, 20(4), 343-357. Kasznik, R. (1999). On the Association between Voluntary Disclosure and Earnings Management. Journal of Accounting Research, 37(1), 57-81. Kent, P., & Stewart, J. (2008). Corporate Governance and Disclosures on the Transition to International Financial Reporting Standards. Accounting and Finance, 48(4), 649-671. Kim, M.-S., & Kross, W. (1998). The impact of the 1989 change in bank capital standards on loan loss provisions and loan write-offs. Journal of Accounting and Economics, 25(1), 69-99. KPMG. (2006). The Application of IFRS: Choices in Practice. Krippendorff, K. (2004). Content Analysis: An Introduction to its Methodology (2nd ed.). London: Sage Publications. la Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and finance. Journal of Political Economy, 106(6), 1113. Lambert, R. A. (2001). Contracting theory and accounting. Journal of Accounting and Economics, 32(1-3), 3-87. Langer, R., & Lev, B. (1993). The FASB's Policy of Extended Adoption for New Standards: An Examination of FAS No. 87. Accounting Review, 68(3), 515-533. Laughlin, R. (1995). Empirical research in accounting: alternative approaches and a case for "middle range" thinking. Accounting, Auditing & Accountability, 8(1), 63-87. Lee, E., Walker, M., & Christensen, H. B. (2008). Mandating IFRS: its Impact on the Cost of 120 Equity Capital in Europe (No. RR105): ACCA. Leuz, C. (2003). IAS Versus U.S. GAAP: Information Asymmetry-Based Evidence from Germany's New Market. Journal of Accounting Research, 41(3), 445-472. Leuz, C., & Verrecchia, R. (2000). The Economic Consequences of Increased Disclosure. Journal of Accounting Research, 38(3), 91-124. Li, S. (2010). Does Mandatory Adoption of International Financial Reporting Standards in the European Union Reduce the Cost of Equity Capital? Accounting Review, 85(2), 607- 636. Milgrom, P. R. (1981). Good News and Bad News: Representation Theorems and Applications. The Bell Journal of Economics, 12(2), 380-391. Ministry of Economic Development. (2009). The Statutory Framework for Financial Reporting. Niskala, M., & Pretes, M. (1995). Environmental Reporting in Finland - a Note on the Use of Annual-Reports. Accounting Organizations and Society, 20(6), 457-466. NZICA. (2006). Audit Guidance Statement No. 1012: Audit Implications of the Transition to New Zealand Equivalents to International Financial Reporting Standards: New Zealand Institute of Chartered Accountants. Palepu, K. G., Healy, P. M., Bernard, V. L., Wright, S., Bradbury, M., & Lee, P. (2010). Business Analysis & Valuation (First ed.). Melbourne: Cengage. Palmer, P. D. (2008). Disclosure of the impacts of adopting Australian equivalents of International Financial Reporting Standards. Accounting and Finance, 48(5), 847-870. PriceWaterhouseCoopers. (2006). New Zealand 2006 Corporate Issues Report. Renders, A., & Gaeremynck, A. (2007). The Impact of Legal and Voluntary Investor Protection on the Early Adoption of International Financial Reporting Standards (IFRS). De Economist (0013-063X), 155(1), 49-72. Sami, H., & Welsh, M. J. (1992). Characteristics of early and late adopters of pension accounting standard SFAS No. 87. Contemporary Accounting Research, 9(1), 212-236. Scott, T. W. (1991). Pension disclosures under SFAS No. 87: Theory and evidence. Contemporary Accounting Research, 8(1), 62-81. Scott, W. R. (2006). Financial Accounting Theory (Fourth ed.). Toronto: Pearson Prentice Hall. Smith, M., & Taffler, R. (2000). The chairman's statement - A content analysis of discretionary narrative disclosures. Accounting, Auditing & Accountability Journal, 13(5), 624-647. Stanton, P., & Stanton, J. (2002). Corporate annual reports: research perspectives used. Accounting, Auditing & Accountability Journal, 15(4), 478-500. Steenkamp, N., & Northcott, D. (2007). Content Analysis in Accounting Research: The Practical Challenges. Australian Accounting Review, 17 No.3(43), 12-25. Stent, W. J., Bradbury, M., & Hooks, J. (2010). IFRS in New Zealand: Effects on Financial Statements and Ratios. Pacific Accounting Review 22(2), 92-107. Sweeney, A. P. (1994). Debt-covenant violations and managers' accounting responses. Journal of Accounting and Economics, 17(3), 281-308. Texeira, A., & Pickens, D. (2004). Global FRS: Risks and Strategies. Chartered Accountants Journal of New Zealand, 83(11), 18-19. Trueman, B. (1990). Theories of Earnings-announcement Timing. Journal of Accounting & Economics, 13(3), 285-301. UK Financial Reporting Review Panel. (2007). Annual Activity Report for year to 31 March 2007 (No. FRRP PN104): Financial Reporting Council (UK). Verrecchia, R. (1983). Discretionary disclosure. Journal of Accounting and Economics, 5, 179- 194. Wang, S., & Welker, M. (2011). Timing Equity Issuance in Response to Information Asymmetry Arising from IFRS Adoption in Australia and Europe. Journal of Accounting Research, 49(1), 257-307. Watts, R. L. (1992). Accounting choice theory and market-based research in accounting,. The British Accounting Review, 24(3), 235-267. Watts, R. L., & Zimmerman, J. L. (1986). Positive Accounting Theory. Englewood Cliffs, New Jersey: Prentice-Hall International. Wong, J., & Wong, N. (2005). The Impact of Not Amortizing Intangible Assets on Valuation 121 Multiples. Pacific Accounting Review (Pacific Accounting Review Trust), 17(1), 79-93. Zeff, S. A., & Nobes, C. W. (2010). Commentary: Has Australia (or Any Other Jurisdiction) ‘Adopted’ IFRS? Accounting Horizons, 20(2), 178-179. 122 123 124 125 Appendix 3: Themes used for coding of IFRS related extracts from discretionary narrative disclosures Theme Description Example Source Importance IFRS mention indicates the level of importance attached to IFRS adoption Contextual The IFRS mention provides context to information being presented. "Note: 2004 and 2005 figures are reported under NZ IFRS" Millennium and Copthorne Hotels NZ Ltd - 31/12/2005 Highlight The IFRS mention highlights the company is a market leader by adopting early or achieving something unique with regard to adopting IFRS. "NZ IFRS New Zealand equivalents to International Financial Reporting Standards. ANZO adopted these accounting standards from 1 July 2006, the first of New Zealand’s listed property vehicles to do so." AMP NZ OfficeTrust - 30/6/2007 Informing of Importance The IFRS mention draws attention to the change in a way that emphasises its importance. "(Financials for 2004/05 have been restated in line with NZ IFRS as required by NZ IFRS 1 First-time Adopters of New Zealand Equivalents to International Financial Reporting Standards)" - Highlighted in green at top of the Chairman's Report Abano Healthcare Group Ltd - 31/5/2006 126 Theme Description Example Source Potency IFRS mention highlights significance of change to accounting information as a result of IFRS. High potency Assets The IFRS mention refers to a significant financial change in Assets as a result of moving from old GAAP to IFRS. "Note: The 2002-2004 financial statements have not been restated to comply with NZ IFRS. The most significant differences in accounting treatment include the treatment of goodwill amortisation, classification of biological assets, classification of fixed term licences and software to intangible assets, and impairment of property, plant and equipment to fair value if appropriate." Tourism Holdings Ltd - 30/6/2006 Liabilities The IFRS mention refers to a significant financial change in liabilities as a result of moving from old GAAP to IFRS. "The resultant impact on the deferred tax liability of the above changes." Southern Capital Ltd (Hirequip) - 30/6/2006 Equity (and income) The IFRS mention refers to a significant financial change in profit, loss or equity as a result of moving from old GAAP to IFRS. "NPAT of $25.22 million reported for the January 2006 year under NZ GAAP has been restated under NZ IFRS to $24.77 million." Briscoes - 28/1/2007 Revenue The IFRS mention refers to a significant financial change in revenue as a result of moving from old GAAP to IFRS. "Under NZ GAAP, property sales and other income for 2005 would have been $30,805,000. This represents an increase of 52.9% on 2004 (2004: $20,150,000)." CDL Investments Ltd - 31/12/2005 Expenses The IFRS mention refers to a significant financial change in expenses as a result of moving from old GAAP to IFRS. "With the adoption of NZIFRS, the calculation basis for income tax expenses has changed." Contact Energy Ltd - 30/6/2006 Disclosure The IFRS mention refers to a significant change in the nature and extent of disclosure as a result of moving from old GAAP to IFRS "In accordance with NZIFRS, Contact is required to report segmental information. Contact’s primary reporting format is business segments. As these segments are fully integrated within New Contact Energy Ltd - 30/6/2006 127 Theme Description Example Source Zealand, this disclosure does not attempt to present the segments as stand-alone entities." Low Potency Limited or immaterial impact The IFRS mention states there is minimal change to accounting information as a result of applying IFRS. "The transition to NZ IFRS has not had a material impact on the reported profit before tax or cash flow." Ebos Group Ltd - 30/6/2006 General Statements which do not refer to specific amounts or components of financial statements. "This is the first full year that ASB has reported under the new International Financial Reporting Standards, therefore some of the financial data cannot strictly be compared to previous year’s figures." ASB Capital Ltd - 30/6/2006 Evaluative IFRS mentions highlight impact other than accounting information (e.g., benefits, costs, operational changes, key ratios). Neutral Neutral statements regarding non- accounting impacts of IFRS. "These are changes in accounting definitions and do not change Contact's business strategy ... " Contact Energy Ltd - 30/6/2006 Negative Non-financial impacts which are adverse consequences as a result of IFRS. "Overall the requirements to report under IFRS have been onerous and we trust that the additional reading for shareholders provides greater value in the understanding of the Finzsoft business." Finzsoft Solutions Ltd - 31/3/2007 Positive Non-financial impacts which are improvements or reflect beneficial change. "Early adoption of New Zealand International Financial Reporting Standards (NZ IFRS), making EBOS a leader in financial transparency." Ebos Group Ltd - 30/6/2006 128 APPENDIX 4: Survey questionnaire LA32 IMPACT OF ADOPTING INTERNATIONAL FINANCIAL REPORTING STANDARDS (IFRS) THIS SURVEY AND ITS OBJECTIVES Thank you for taking the time to complete this questionnaire – your efforts are very much appreciated! The questionnaire is part of a research project to assess the impact of adoption of IFRS on New Zealand listed companies. Please refer to the Information Sheet which has been provided for further details relating to the research project and your rights as a potential participant. Please complete by ticking (9) the appropriate box(es) where applicable, or supplying the information requested PRELIMINARY INFORMATION 1 Please state your job title ………………………………………………………………………………………….. 2 Please state your responsibilities with regard to the implementation of IFRS within your organisation …………………………………………………………………………………………..……… …………………………………………………………………………………..……………… ………………………………………………………………………….. 3 Please provide the following information regarding your company: a. Percentage of Equity shares held by main shareholder (rounded up): 0 – 10% 11 - 19% 20 – 50% 51 – 100% ‰ ‰ ‰ ‰ b. Country of origin of main shareholder (if known) ………………………………………………………………………………….. 129 A) IMPLEMENTATION PROJECT 1. What was the size of the project team initially appointed to manage the implementation of IFRS for your organisation*? Number of Persons: …………………………………………………………….. Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… 2. What was the maximum size of the project team responsible for managing the implementation of IFRS for your organisation, at any one time during the implementation process *? Number of Persons: …………………………………………………………….. Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… 3. How many months prior to date of transition to NZ IFRS (ie. start of first comparative period under NZ IFRS), did your organisation formally begin planning for the implementation of IFRS? Approximate Number of Months: …………………………………………….. Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… * = It is accepted/understood that persons responsible for managing the implementation of IFRS are likely to have had other responsibilities and may not necessarily have been dedicated only to the IFRS implementation project. 130 4. How many person hours (to nearest hundred) do you estimate* were spent on the implementation of IFRS? Number of Person Hours*: ……..…………………………………… hours* * = Please consider the following in arriving at your estimate: Guideline: 1800 hours = one person working full time for one year (assumes 40 hour working week less 4 weeks leave, all public holidays, 3 days sick leave) Period: From when formal planning for IFRS began until the first full-year IFRS compliant financial statements were finalised Who: Any external consultants as well as any employees involved Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… 5. Please provide the following information relating to the costs of implementing IFRS in your organisation: Type of Cost Budgeted (NZ$ 000’s) Actual (NZ$ 000’s) a) System changes (hardware/software) b) Staff training c) External consultants d) Internal personnel/employee costs e) Other costs (please include a brief description of these under “Comments”) Total Cost of implementing IFRS Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… 131 6. Please indicate how much of an impact IFRS has had on your organisation’s contracts (eg. costs of/ need to renegotiate contracts)? Please also use the second column to rank the 4 categories of contract: The category with the highest monetary value should be ranked first (1); the category with the lowest should be ranked last (4); if no such contracts exist then simply indicate “N/A”. Category of Contract Applicability (N/A) or Value Rank(eg.1st = 1; 4th = 4) No impact (1) ........ Significant impact (5) Debt (eg. Overdrafts, Loans) 1 2 3 4 5 Revenue (eg Royalty Income) 1 2 3 4 5 Expenses (eg. Management remuneration) 1 2 3 4 5 Other (Specify below): ………………………… 1 2 3 4 5 To give some perspective of the effort required for the implementation of IFRS in your organisation, please provide the following overall information: 7. What is the size of your entire financial accounting team (ie. persons involved in preparation of financial accounts as opposed to routine data capture/administration tasks)? Number of Persons: …………………………………………………………. Comments: ……………………………………………………………………………………………… ……………………………………………………………………………… 8. Before IFRS, what were the direct labour costs of preparing financial accounts for your organisation? (Please estimate by calculating the annual average cost for the 3 years immediately prior to when your organisation formally began incurring IFRS related expenditures). Annual Average Cost ((Y1$ + Y2$ + Y3$) ÷ 3): NZ$..................,000 Comments: ……………………………………………………………………………………………… ……………………………………………………………………………… 132 B) IMPACT OF ADOPTING IFRS 9. Please consider the nine statements below relating to expected benefits of adopting IFRS and indicate your opinion on each. Please then rank these statements, using the far right hand column, from most important (Ranking Value = 1) to least important (Ranking Value = 9): Expected benefits of adopting IFRS Strongly Agree Agree Unsure Disagree Strongly Disagree Importance Ranking (1 = Most; 9 = Least) a. Improved quality of financial reporting for our organisation. ‰ ‰ ‰ ‰ ‰ b. Increased comparability of our financial reports with those of similar organisations internationally. ‰ ‰ ‰ ‰ ‰ c. Reduced costs of financial analysis which our organisation performs. ‰ ‰ ‰ ‰ ‰ d. Removal of barriers previously encountered regarding international capital flows to/from our organisation. ‰ ‰ ‰ ‰ ‰ e. Reduced costs regarding our extra* financial reporting requirements (e.g. listing or reporting requirements in foreign countries/ reconciling NZ to foreign GAAP). [* = Disregard ‘basic” cost of preparing NZ IFRS financial statements - consider only further financial reporting requirements you have, if any ] . ‰ ‰ ‰ ‰ ‰ 133 Expected benefits of adopting IFRS Strongly Agree Agree Unsure Disagree Strongly Disagree Importance Ranking (1 = Most; 9 = Least) f. Greater alignment of our financial reporting with preferences of our institutional investors and analysts. ‰ ‰ ‰ ‰ ‰ g. Users are indicating that they have greater confidence in the financial information produced by our organisation. ‰ ‰ ‰ ‰ ‰ h. Long-term net benefits will arise from adopting IFRS (ie. benefits should exceed costs) for my organisation ‰ ‰ ‰ ‰ ‰ i. Long-term net benefits will arise from adopting IFRS (ie. benefits should exceed costs) for New Zealand. ‰ ‰ ‰ ‰ ‰ Comments: (Please include here, any further benefits you believe should be considered and indicate an importance ranking for each, using the same scale as above e.g. if you give an importance ranking of 3 to a further benefit, we will assume that your rankings 3 – 9 provided in the boxes above will all move down by one and become 4 – 10). …………………………………………………………………………………………………………… …………………………………………………………………………………………………………… …………………………………………………………………………………………………………… …………………………………………………………………………………………………………… ………………………….. Thank you very much for your time! 134 The impact of International Financial Reporting Standards (IFRS) in New Zealand INFORMATION SHEET General Information This questionnaire is part of a doctoral research project designed to assess impacts of the adoption of IFRS that are not readily apparent from information which is disclosed in company financial reports. Participant Recruitment All New Zealand listed companies that elected to adopt IFRS early (ie. prior to periods beginning on or after 1 January 2007) were selected for research purposes. A further group of companies that adopted IFRS once it became mandatory was also selected as a control group. Participant’s Rights You are under no obligation to participate in this study. If you decide to participate, you have the right to: • decline to answer any particular question; • withdraw from the study at any time prior to finalization of the results of the study; • ask any questions about the study at any time during participation; • provide information on the understanding that your name will not be used unless you give permission to the researcher; • be given access to a summary of the project findings when it is concluded. Project Contacts If you have any questions about the project, please do not hesitate to contact me, or my supervisors. Our full contact details are: Researcher Supervisors Mr Warwick Stent School of Accountancy Private Bag 102-904 Massey University Auckland New Zealand Direct Dial +64 9 414-0800 ext 9542 Fax +64 9 441-8133 ¾ Professor Michael Bradbury ¾ Professor Jillian Hooks School of Accountancy Private Bag 102-904 Massey University Auckland New Zealand Direct Dial +64 9 414-0800 ext 9458 Fax +64 9 441-8133 Committee Approval Statement This project has been reviewed and approved by the Massey University Human Ethics Committee: Northern, Application 07/008. If you have any concerns about the conduct of this research, please contact Associate Professor Ann Dupuis, Acting Chair, Massey University Human Ethics Committee: Northern, telephone 09 414 0800 x9054, email humanethicsnorth@massey.ac.nz. 135 1 March 2012 Mr Stuart XXXXX Chief Financial Officer XXXXXX Industries Ltd P.O Box XX, City, XXXX Dear Stuart, Research regarding the impact of International Financial Reporting Standards I am currently completing a comprehensive research project regarding the impact of International Financial Reporting Standards in New Zealand”. As you are no doubt aware, the decision to replace New Zealand Generally Accepted Accounting Practice (NZ GAAP) with International Financial Reporting Standards (IFRS) is regarded as one of the most significant developments in the history of financial reporting in New Zealand. This change has had wide ranging impacts for entities which are currently required by law to produce financial statements in New Zealand. It is these impacts, as well as the motivations of companies regarding adoption timing (ie. choosing to adopt IFRS early or waiting until it became mandatory), that is the focus of my research project. In this regard, I would be most grateful for your assistance in completing the enclosed questionnaire as part of this research project. The questionnaire is designed to assess impacts of the adoption of IFRS that are not readily apparent from information which is disclosed in company financial reports. Benefits for respondents are that they may request summary information on the results of this study, which may be useful in: ¾ Benchmarking your own firm’s experience/ performance regarding the impact of IFRS against the summary information, ¾ Assisting with any future/ongoing representations to bodies such as the outgoing Financial Reporting Standards Board or incoming External Reporting Board and its sub-committees regarding the implementation and application of IFRS. The questionnaire should take no more than 20 minutes to complete and all responses will be treated as strictly confidential. Only aggregated information will be used for any conference presentation and/ or publication. For additional information on this research project or project contact persons, please refer to the information sheet provided at the back of the questionnaire. My sincere thanks for your time and assistance. Yours faithfully Warwick Stent (Lecturer and Researcher at Massey University) Tel: +64 9 414 0800 X9542 Email: W.J.Stent@massey.ac.nz APPENDIX 5: Covering letter for survey 136 APPENDIX 6: Survey instrument with interview prompts 1 IMPACT OF ADOPTING INTERNATIONAL FINANCIAL REPORTING STANDARDS (IFRS) THIS SURVEY AND ITS OBJECTIVES Thank you for taking the time to complete this questionnaire – your efforts are very much appreciated! The questionnaire is part of a research project to assess the impact of adoption of IFRS on New Zealand listed companies that have elected to adopt IFRS early (ie. prior to periods beginning on or after 1 January 2007). Please refer to the Information Sheet which has been provided for further details relating to the research project and your rights as a potential participant. Please complete by ticking (9) the appropriate box(es) where applicable, or supplying the information requested PRELIMINARY INFORMATION 1 Please state your job title ………………………………………………………………………………………….. 2 Please state your responsibilities with regard to the implementation of IFRS within your organisation …………………………………………………………………………………………..……… …………………………………………………………………………………..……………… ………………………………………………………………………….. 4 Please provide the following information regarding your company: a. Percentage of Equity shares held by main shareholder (rounded up): 0 – 10% 11 - 19% 20 – 50% 51 – 100% ‰ ‰ ‰ ‰ b. Country of origin of main shareholder (if known) ………………………………………………………………………………….. 137 C) IMPLEMENTATION PROJECT 10. What was the size of the project team initially appointed to manage the implementation of IFRS for your organisation*? Number of Persons: …………………………………………………………….. Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… 11. What was the maximum size of the project team responsible for managing the implementation of IFRS for your organisation, at any one time during the implementation process *? Number of Persons: …………………………………………………………….. Comments: Interview Prompt: Who made up the team? What expertise did they have? ……………………………………………………………………………………………… ……………………………………………………………………………… 12. How many months prior to date of transition to NZ IFRS (ie. start of first comparative period under NZ IFRS), did your organisation formally begin planning for the implementation of IFRS? Approximate Number of Months: …………………………………………….. Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… * = It is accepted/understood that persons responsible for managing the implementation of IFRS are likely to have had other responsibilities and may not necessarily have been dedicated only to the IFRS implementation project. 138 13. How many person hours (to nearest hundred) do you estimate* were spent on the implementation of IFRS? Number of Person Hours*: ……..…………………………………… hours* * = Please consider the following in arriving at your estimate: Guideline: 1800 hours = one person working full time for one year (assumes 40 hour working week less 4 weeks leave, all public holidays, 3 days sick leave) Period: From when formal planning for IFRS began until the first full-year IFRS compliant financial statements were finalised Who: Any external consultants as well as any employees involved Comments: ……………………………………………………………………………………………… ……………………………………………………………………………………………… ……………………………………………………………………… 14. Please provide the following information relating to the costs of implementing IFRS in your organisation: Type of Cost Budgeted (NZ$ 000’s) Actual (NZ$ 000’s) a) System changes (hardware/software) b) Staff training c) External consultants d) Internal personnel/employee costs e) Other costs (please include a brief description of these under “Comments”) Total Cost of implementing IFRS Comments: Interview Prompts: Check “other costs” and ask for any clarifications needed. Impressions re impact on audits (disclosures are being analysed) – time/costs? In summary, was effort/cost required more or less than expected? ……………………………………………………………………………………………… ……………………………………………………………………………… 139 15. Please indicate how much of an impact IFRS has had on your organisation’s contracts (eg. costs of/ need to renegotiate contracts)? Please also use the second column to rank the 4 categories of contract: The category with the highest monetary value should be ranked first (1); the category with the lowest should be ranked last (4); if no such contracts exist then simply indicate “N/A”. Category of Contract Applicability (N/A) or Value Rank(eg.1st = 1; 4th = 4) No impact (1) ........ Significant impact (5) Debt (eg. Overdrafts, Loans) 1 2 3 4 5 Revenue (eg Royalty Income) 1 2 3 4 5 Expenses (eg. Management remuneration) 1 2 3 4 5 Other (Specify below): Clarify if necessary…… 1 2 3 4 5 To give some perspective of the effort required for the implementation of IFRS in your organisation, please provide the following overall information: 16. What is the size of your entire financial accounting team (ie. persons involved in preparation of financial accounts as opposed to routine data capture/administration tasks)? Number of Persons: …………………………………………………………. Comments: Interview Prompts: Has it grown specifically as a result of IFRS? …………………………………………………………………………… 17. Before IFRS, what were the direct labour costs of preparing financial accounts for your organisation? (Please estimate by calculating the annual average cost for the 3 years immediately prior to when your organisation formally began incurring IFRS related expenditures). Annual Average Cost ((Y1$ + Y2$ + Y3$) ÷ 3): NZ$..................,000 Comments: ……………………………………………………………………………………………… ……………………………………………………………………………… 140 D) IMPACT OF ADOPTING IFRS 18. Please consider the nine statements below relating to expected benefits of adopting IFRS and indicate your opinion on each. Please then rank these statements, using the far right hand column, from most important (Ranking Value = 1) to least important (Ranking Value = 9): Expected benefits of adopting IFRS Strongly Agree Agree Unsure Disagree Strongly Disagree Importance Ranking (1 = Most; 9 = Least) j. Improved quality of financial reporting for our organisation. ‰ ‰ ‰ ‰ ‰ k. Increased comparability of our financial reports with those of similar organisations internationally. ‰ ‰ ‰ ‰ ‰ l. Reduced costs of financial analysis which our organisation performs. ‰ ‰ ‰ ‰ ‰ m. Removal of barriers previously encountered regarding international capital flows to/from our organisation. ‰ ‰ ‰ ‰ ‰ 141 Expected benefits of adopting IFRS Strongly Agree Agree Unsure Disagree Strongly Disagree Importance Ranking (1 = Most; 9 = Least) n. Reduced costs regarding our extra* financial reporting requirements (such as listing or reporting requirements in foreign countries/ reconciling NZ to foreign GAAP). [* = Disregard ‘basic” cost of preparing NZ IFRS financial statements - consider only further financial reporting requirements you have, if any ] . ‰ ‰ ‰ ‰ ‰ o. Greater alignment of our financial reporting with preferences of our institutional investors and analysts. ‰ ‰ ‰ ‰ ‰ p. Users are indicating that they have greater confidence in the financial information produced by our organisation. ‰ ‰ ‰ ‰ ‰ q. Long-term net benefits will arise from adopting IFRS (ie. benefits should exceed costs) for my organisation ‰ ‰ ‰ ‰ ‰ r. Long-term net benefits will arise from adopting IFRS (ie. benefits should exceed costs) for New Zealand. ‰ ‰ ‰ ‰ ‰ 142 Comments: (Please include here, any further benefits you believe should be considered and indicate an importance ranking for each, using the same scale as above eg. if you give an importance ranking of 3 to a further benefit, we will assume that your rankings 3 – 9 provided in the boxes above will all move down by one and become 4 – 10). Interview Prompt: Check if any further benefits identified that need clarification and ask if respondents would like to elaborate where there is a trend or strong opinion. …………………………………………………………………………………………………………… …………………………………………………………………………………………………………… …………………………………………………………………………………………………………… …………….. Thank you very much for your time! E) EARLY/ LATE ADOPTION OF IFRS 19. Adoption of IFRS was possible from 1 January 2005 (on a voluntary basis); mandatory from 1 January 2007. Your organisation elected to adopt early/ wait until it became mandatory. Could you give me your insights/background as to why early/ late? 143 The impact of International Financial Reporting Standards (IFRS): A study of Early Adopters in New Zealand INFORMATION SHEET General Information This questionnaire is part of a doctoral research project entitled “The impact of International Financial Reporting Standards: A study of Early Adopters in New Zealand”. The questionnaire is designed to assess impacts of the adoption of IFRS that are not readily apparent from information which is disclosed in company financial reports. Participant Recruitment All New Zealand listed companies that have elected to adopt IFRS early (ie. prior to periods beginning on or after 1 January 2007) have been selected for research purposes. These companies have been identified by reference to information disclosed in their issued financial reports. Current indications are that approximately 37 companies have elected to adopt IFRS early. Participant’s Rights You are under no obligation to participate in this study. If you decide to participate, you have the right to: • decline to answer any particular question; • withdraw from the study at any time prior to finalization of the results of the study; • ask any questions about the study at any time during participation; • provide information on the understanding that your name will not be used unless you give permission to the researcher; • be given access to a summary of the project findings when it is concluded. Project Contacts If you have any questions about the project, please do not hesitate to contact me, or my supervisors. Our full contact details are: Researcher Supervisors Mr Warwick Stent School of Accountancy Private Bag 102-904 Massey University Auckland New Zealand Direct Dial +64 9 414-0800 ext 9542 Fax +64 9 441-8133 ¾ Professor Michael Bradbury ¾ Associate Professor Jillian Hooks School of Accountancy Private Bag 102-904 Massey University Auckland New Zealand Direct Dial +64 9 414-0800 ext 9458 Fax +64 9 441-8133 Committee Approval Statement This project has been reviewed and approved by the Massey University Human Ethics Committee: Northern, Application 07/008. If you have any concerns about the conduct of this research, please contact Associate Professor Ann Dupuis, Acting Chair, Massey University Human Ethics Committee: Northern, telephone 09 414 0800 x9054, email humanethicsnorth@massey.ac.nz.