Check the numbers: Accounting information still matters to you, me and investors around the world

When: 9 August 2018

Where: University of Western Australia, Perth, Australia

Ann Tarca delivered the CPA Philip Brown Annual Lecture, which is organised annually by CPA Australia together with the University of Western Australia Business School and Curtin Business School.

Introduction

Good evening. It is with great pleasure that I deliver the Philip Brown CPA Lecture this evening. I will discuss three things: (1) the relationship between accounting information and share prices, that is, what we know about the impact of accounting information in capital markets, based on empirical accounting research, (2) whether current accounting practices and changes in the economy have diminished the role for accounting information, and (3) the ways the International Accounting Standards Board (Board) is addressing the issues raised by those who consider accounting has lost its relevance.

I first met Professor Philip Brown in 1996 when I commenced post graduate studies at the University of Western Australia. I was immediately impressed by Philip’s enquiring mind and the rigour with which he examined research questions. Subsequently Philip was one of my PhD supervisors and then my coauthor on several research papers. Like many of Philip’s PhD students, I have continually referred to principles, techniques and strategies I learned from Philip – applicable to accounting research and beyond.

Background—Ball & Brown 1968

Philip’s own career received a crucial boost from a paper he published with Ray Ball in 1968. In this paper Philip and Ray investigated whether (or, rather, to what extent) accounting information was useful in capital markets, that is, they investigated whether the release of accounting earnings was reflected in stock prices. Today this seems a question with an obvious answer—yes. But at the time the answer was not obvious. In the 1960s accounting research was largely normative—investigating what ought to be rather than what was happening.

Like some of the very best research, Philip and Ray’s underlying idea was simple: could they observe any relationship between stock market returns and the release of companies’ earnings announcements? The innovation of their work was to apply a scientific method and provide empirical evidence to answer their research question. As many of you know, their study led to the development of an entire research school which continues to be influential in the lives and work of many academics and practitioners.

The findings of Ball and Brown showed that cumulative abnormal returns increased for firms with unexpected good news prior to the earnings announcement with some effects following the announcement. An opposite and stronger reaction was observed for bad news firms. The evidence was taken to support semi strong market efficiency, that is, market prices reflect all publicly available information. Since the publication of Ball and Brown many scholars have investigated various relationships of accounting information, share prices and market returns.

However, since the 1980s some researchers have argued that financial accounting numbers are becoming less useful, that is, less relevant to investors. This conclusion is based on examining the statistical association between accounting earnings and share prices (or returns) over time.

Loss of value relevance—Lev and other research

A recent strong voice in this area has been Professor Baruch Lev from New York University. In his December 2017 presentation at the ICAEW Information for Better Markets conference in London he stated that there was ‘widespread and growing dissatisfaction with the relevance of financial reporting information particularly among investors and corporate executives’. He pointed to research that shows a growing gap between capital market indicators and financial information, particularly earnings.

He linked the deterioration of the usefulness of financial information to: (1) the abandonment by accounting standard-setters of the traditional income statement (matching) model in favour of a balance sheet (asset valuation) model, and (2) standard-setters’ failure to adjust asset recognition rules to the fundamental shift in corporate value-creating resources from tangible to intangible assets.¹ I will return to these criticisms later.

The research referred to by Professor Lev investigates the association of share prices (or returns) and a range of accounting measures in addition to earnings.

Measures include book value of equity; costs of goods sold; selling, general and administrative expenses; research and development expense; advertising expense, capital expenditure and revenue growth. Explanations proposed for the decline in relevance of accounting information include (1) the rise of the ‘new economy’ featuring companies with future earnings that depend largely on intangible assets and (2) the presence of more loss companies (for which earnings is less relevant to determining share price).

Research that provides another view

So, is there any hope for the usefulness of accounting information, the preparation and consumption of which is the focus of the working lives of many of us here this evening? The good news is yes, according to some recent new research.

In a paper published in 2014 Dechow, Sloan and Zha (two of Philip Brown’s outstanding honours students from the class of 1986 and one of Patty Dechow’s PhD students) replicated the B&B study for 1971–2012 and concluded that the relationships demonstrated in B&B were still present. That is, share prices reflect accounting information. Similarly, the study’s replication of Beaver (1968) showed increased trading volume and price residuals on the earnings announcement day, particularly in the later years (since 2000 ie 2001–2012).²

A study by Professor Mary Barth from Stanford University and her colleagues applies a new design approach (ie statistical technique) to a long time series of US data (1962–2014) to investigate the issue of the declining relevance of accounting information. The authors consider a larger set of accounting amounts than in prior work. They include accounting amounts that could reflect information about intangible assets (R&D expense, advertising expense and recognised intangible assets), growth opportunities (cash and revenue growth) and other performance measures (operating cash flow, revenue, special items and other comprehensive income)—all relevant to new economy companies.

They find no evidence of a decline in the value relevance of accounting information across all sample years or in any decade except the 1990s (tech stock bubble). In addition, they find increases in the value relevance of the accounting amounts for intangible assets, growth opportunities and other performance measures. The combined increase in relevance of these items offsets the decline in value relevance of earnings.

Although the trends are more pronounced for ‘new economy’ companies they are present for other companies too, suggesting the evidence is economy-wide. The authors conclude there is an evolution in value relevance of accounting information to a ‘more nuanced not declining’ relation between accounting amounts and share prices that reflects equity valuation in the new economy.³

Professor Barth’s evidence relates to US companies and markets. Is there evidence from other markets, and for IFRS firms? A recent study by Davern, Gyles, Hanlon and Pinnuck (University of Melbourne and Monash University) finds no decline in the value relevance of earnings for Australian firms from 1992 to 2015.⁴ The authors conclude that both shareholders’ equity and net income are important for investor decision making.

The empirical findings are consistent with evidence gathered in interviews with investors, regulators and practitioners. Interviewees note that the audited financial statements are an important foundation in investor decision making, having a confirmatory role and providing the initial input to investment models. The researchers also investigate the usefulness of summary measures including EBIT and EBITDA. They find both measures are value relevant and conclude that they are complements to net income. Interviewees state that these measures are used by investors when predicting and entity’s future performance.

Marvin Wee, Greg Clinch and I have also been investigating summary performance measures presented by IFRS adopting companies. We studied the non-GAAP disclosures by 400 IFRS adopting firms in eight countries in four years between 2005 and 2013. We find that the disclosure of non-GAAP earnings reconciled to a measures of management defined operating profit are strongly associated with share prices, suggesting the disclosure is useful to market participants. In addition, the adjusting items are not significant, consistent with them not being relevant to determining price.

The evidence is consistent with investors using multiple measures of earnings, not just net profit.⁵ The evidence of these studies is relevant to the Board’s work on the Primary Financial Statements project, which I will mention shortly.

Standard-setter response—Lev 1: Balance Sheet focus/IFRS 3/Conceptual Framework

I will return now to Professor Lev’s criticisms of the shortcomings of accounting information and consider how professional accountants and standard setters should respond.

Overall, we can be encouraged that there is evidence of the continuing value relevance of accounting information, albeit that some items of information have increased in importance (ie information about intangibles and disaggregated amounts) while others (such as earnings, a high-level summary measure) have decreased.

In light of the evidence and the criticisms, what should the Board be doing - remembering of course that the Board works in consultation with preparers, auditors, regulators and investors, so we need your input and advice about any changes we are considering.

An important Board project is the Primary Financial Statements project. We are working on requiring changes to the Statement of Financial Performance that will require entities to present more subtotals and more specific classifications of items. The requirements will encourage disaggregation to ensure more information about items by nature is available.

In addition, we will propose that if management performance measures are provided (ie additional performance measures calculated by an entity’s management) then they are presented within financial statements and subject to reconciliation and audit.

Considering the research findings I have presented this evening, we are confident that these changes will further enhance the accounting information provided that is relevant in today’s capital markets.

But what about Lev’s claim that standard-setters have a misguided focus on asset valuation and that they neglect the matching of revenue and costs? We could look at IFRS 3, the Standard that applies to acquisition accounting, as an example. On acquisition, the assets of the acquired company are recorded by the acquiring company at their fair value, which represents the value of the business acquired. Surely that makes good sense?

The objective of accounting standards, which underpins the standard-setter’s work and is familiar to many of you, is to provide information that is relevant and representationally faithful and thus useful to decision makers who rely on that accounting information.

The acquirer is recording the assets acquired at their cost to the acquirer, and of course recording transactions at historical cost is a fundamental and longstanding principle of accounting. Assets acquired that have a finite useful life are amortised over their useful life, reflecting consumption of the asset.

However, this accounting leads to an issue for analysts that makes comparison of some companies difficult. Companies that grow through acquisitions may show higher cost of goods sold and more amortisation expense relating to the value of inventory and identifiable intangible assets acquired. The postacquisition earnings are reduced by these charges, which, being non-cash items, are not reflected in cash flows.

If a company grows organically, that is, not by acquisition, and creates intangible assets such as brands and customer lists through its own business activity, the company will not record the charges for amortisation that are incurred by the acquirer. Therefore, many analysts adjust the reported earnings of the acquiring company (ie add back to earnings the effects of fair value adjustments and amortisation) so they can better compare companies and investment opportunities.

Companies may provide clear disclosures to allow investors to make these adjustments, but the accounting is part of a friction that imposes costs and noise into the information system. An outcome could be that, for acquirer companies, earnings becomes less relevant as a summary measure and operating cash flows become relatively more relevant. A University of Western Australia doctoral student, Jian Liang, has found evidence that this is the case for IFRS adopting companies.

The above example of IFRS 3 illustrates the logic of the asset valuation model and of using fair value measurement in acquisition accounting. Some critics of the Board state that the standard setter has a fair value bias, perhaps based on their experience with IFRS 3 and the financial instrument standards IAS 39 (and now IFRS 9).

But a careful review of IFRS Standards (which has been undertaken by some academics) shows that fair value measurement is required in only a small number of Standards and there is a clear rationale behind the requirements. In particular, derivatives do not have an historical cost. Fair value measurement is the measurement base which provides relevant and representationally faithful information.

In March 2018, the Board issued the revised Conceptual Framework for Financial Reporting. An important new feature of this document is the measurement chapter. It presents discussion of several measurement bases and the factors to be considered when selecting a measurement base. The chapter is not ‘biased’ in favour of fair value measurement but rather the chapter states that the measurement base will be selected to maximise the qualitative characteristics of financial information (relevant and representationally faithful information).

In addition, the Conceptual Framework is clear that the Statement of Financial Position and the Statement of Financial Performance are of equal importance. It explicitly states that information about income and expenses is just as important as information about assets and liabilities. The structure of asset and liability definitions, followed by definitions of income and expenses, provides the best approach to measuring financial performance and presenting financial position.

If you think about the issues, you can see that the definitions must start somewhere. Standard-setters have spent considerable time and effort attempting to define ‘profit’, but without success. This suggests that we would be unable to build a conceptual framework from the definition of profit.

Also relevant to tonight’s topic is the statement in the Conceptual Framework that understanding an entity’s financial performance requires an analysis of all recognised income and expense, that is, not just a summary total of profit or loss, as well as other information in the financial statements. This approach is consistent with the value relevance research I have been discussing that points to the importance of a range of performance measures within IFRS reporting.

The Concpetual Framework captures a great deal of thinking about contemporary accounting issues and will provide a basis for Board decisions in the future. For the Board, it is already applicable and you can see reference to the Conceptual Framework in our current discussions. The Framework is also being used by the Interpretations Committee.

Standard-setter response—Lev 2: Recognition of intangible assets

The second criticism from Lev is that standard setters have failed to adjust the asset recognition rules to accommodate the changes in the economy related to generating value through intangible assets. Lev’s criticism has more weight in relation to US GAAP where all research and development expenditure (except expenditure on software) must be expensed.

In contrast, IFRS Standards allow the capitalisation of development expenses when certain criteria are met. Lev cites research on IFRS reporting companies that shows firms’ capitalised development costs are highly correlated with market values and that disclosures about the process of testing for capitalisation is useful for investors.⁶

Nevertheless, there is criticism of accounting for intangible assets under IFRS Standards and calls for activity in this area. IAS 38 is one of the oldest IFRS Standards and reflected conservative international practice—recognition of purchased intangible assets at cost, revaluation of intangibles only in the presence of an active market and no recognition of a company’s internally generated intangible assets.

This was somewhat at odds with practice in Australia at the time and some of you may recall a flurry of activity prior to the adoption of IFRS standards in 2005, by both preparers who were concerned about removing internally generated and revalued assets from their balance sheets and academics who considered IFRS accounting requirements as a backward step in relation to providing useful information about intangible assets.

Lev’s criticism is that standard setters do not permit recognition on the balance sheet of the intangible assets that companies are creating. Lev suggests that expenditure on intangible assets should be capitalised and then assets would be recognised on the balance sheet. The relevance of earnings would increase because expenditure that generates future benefits is not being written off. This is consistent with the idea behind capitalisation of development expenses under IAS 38. Lev argues that his proposal is not pursuing an asset valuation approach, but rather better matching of revenue and expenses.

Recent discussions of the Board have raised the question as to whether IAS 38 will serve us well in the future. In addition to being a core part of IFRS 3, intangible assets are important in the extractive industry. The Board has added a research project on Extractive Industries to its work plan so I look forward to hearing your views on this topic in the future.

Another issue that touches IAS 38 is commodity trading, including accounting for cryptocurrencies. The IFRS Interpretations Committee and the Board have had several discussions on this topic over the last year and our research and exploration of issues is continuing.

So, in summary, I find several benefits in the points raised by Professor Lev. He leads us to consider the way the Standards require assets to be recognised (the asset valuation approach) and whether our Standards result in useful information for decision makers.

He also challenges us on the topic of intangible assets which now represents a key issue for standard-setters, without a clear way forward. Therefore, his work shows some of the benefits we can receive from academics: valid criticisms and challenges but also suggestions for changes to address the problems identified.

Does this mean we should pursue more historical cost measurement? In some cases this is not possible. Accounting practice has used a mixed measurement model for decades, reflecting attempts to capture economic reality within accounting information. The Board’s recent Post-Implementation Review of IFRS 13 included a literature review of the impact of IFRS 13, the Standard about fair value measurement. Many studies in that review (using US and IFRS company data) pointed to the relevance of fair value measurement for market participants.

Lev suggests it is time for standard-setters to rethink their approach to accounting for intangible assets. He suggests the capitalisation (and amortisation) of many expenditures on intangible assets (for example internally generated patents, copyrights, software, customer lists) to improve matching of revenue and expense and thus the quality (and value relevance) of earnings. Such a proposal is not presently on our agenda and a project would only be initiated with support from constituents.

We are very aware that at present our community has a substantial amount of work to complete with the implementations of IFRS 15, 9, 16 and 17. We are supporting implementation in a variety of ways, including the work of the Board and the IFRS Interpretations Committee issuing agenda decisions and narrow-scope amendments, transition resource groups, educational material on our website (webinars and articles), speaking at conferences and meetings with regulators, standard-setters and audit firms.

Chair of the Board Hans Hoogervorst has recognised that a period of calm is important as the new Standards are implemented. The future work direction of the Board will be considered in depth through the next agenda consultation.

Analyst response to Lev—Nick Anderson

As I conclude my talk, I would like to mention some insights provided by my colleague Nick Anderson about Professor Lev’s views.⁷ Nick is a Board member with 30 years’ experience as a buy-side equity investor and of course analysts are the primary users of listed company financial information. Nick recognises that the environment for company operations and for financial reporting has changed dramatically during his working life and that the importance of other information (that is, information derived from outside a company’s financial reports) has become more plentiful and more useful.

However, Nick is clear about the vital importance of the foundation provided by audited financial statements in the investment decision making process. Nick also notes that financial statements are not designed to show the value of a reporting entity. As our colleagues at the Board’s Capital Market Advisory Group remind us, valuing an entity is their job!

Overall, Nick points to the range of information that an analyst may use, and also the skills and experience necessary to make wealth enhancing decisions from the available information. Although the Board does not have any specific projects at present that address the broader issues of recognition of more intangible assets, we do have a project to revise our Management Commentary Practice Statement.

We know that investors and others make use of a wide range of information and that many entities seek to provide information to complement the financial statements through their management commentary reports. Accordingly, we consider that updating our guidance for narrative reporting, following the many changes in the economy and in entities' business models and strategies, may encourage better reporting on a wide range of issues, including the value generating potential of intangible assets.

Conclusion

In conclusion, I return to the title of my talk this evening. I once wrote a paper with Philip that referred to accounting information as ‘the life-blood of capital markets’ or similar. (Incidentally, those words feature prominently on this website). I haven’t changed my view and I expect that Philip hasn’t changed his either. Accounting information matters – to entities seeking to raise capital, to accountants working in the preparer, auditor and regulator communities and to analysts and investors.

Yes, the economy has changed in ways we may not have predicted even a short time ago and the range of potentially useful information available to investors has expanded enormously. Yes, there are many intangible assets that are not recognised on the balance sheets of our listed companies and accounting will continue to fail to capture aspects of companies that are essential to their future success, such as the quality of management and the skill of staff.

However, accounting provides a record of past activity and provides the starting point for predicting a range of earnings measures. The Board seeks to improve financial reporting by developing IFRS Standards that promote trust, growth and long-term stability in the global economy.

Please continue to assist us by sharing your knowledge and expertise in the standard-setting process. In the conclusion of his paper, Professor Lev calls researchers to pursue that ‘exciting and potentially influential research area: accounting standard setting’. I fully concur with Professor Lev in this view and look forward to seeing the outcomes of your research endeavours.

¹Lev, B., 2018. The deteriorating usefulness of financial report information and how to reverse it. Accounting and Business Research, 48(5), pp.465–493.

²Dechow, P.M., Sloan, R.G. and Zha, J., 2014. Stock prices and earnings: A history of research. Annual Review of Financial Economics, 6(1), pp.343–363.

³Barth, Mary E. and Li, Ken and McClure, Charles, Evolution in Value Relevance of Accounting Information (May 7, 2018). Stanford University Graduate School of Business Research Paper No. 17–24. Available at SSRN: https://ssrn.com/abstract=2933197 or http://dx.doi.org/10.2139/ssrn.2933197

⁴Is financial reporting still and effective tool for equity investors in Australia? Working paper, University of Melbourne and Monash University. https://www.ifrs.org/- /media/feature/meetings/2018/april/asaf/asaf-08a-research-paper-april-2018.pdf. Also https://www.ifrs.org/-/media/feature/meetings/2018/april/asaf/asaf-08-is-financial-reporting-still-an-effective-tool-for-equity-investors-in-australia-april-2018.pdf

⁵Clinch, Greg and Tarca, Ann and Wee, Marvin, The Value Relevance of IFRS Earnings Totals and Subtotals and Non-GAAP Performance Measures (March 8, 2018). Available at SSRN: https://ssrn.com/abstract=3178567 or http://dx.doi.org/10.2139/ssrn.3178567

⁶Chen, E., Gavious, I. and Lev, B., 2017. The positive externalities of IFRS R&D capitalization: enhanced voluntary disclosure. Review of Accounting Studies, 22(2), pp.677–714.

⁷Anderson, N., 2018. ‘The deteriorating usefulness of financial report information and how to reverse it': a practitioner view. Accounting and Business Research, 48(5), pp.494–496.