CHAPTER 3 BACKGROUND TO VALUE RELEVANCE
The purpose of this chapter is to review the value relevance literature to set the scene for the South African study, which follows in Chapter 4. I start the review of the literature by discussing the basic principles that underlie value relevance research, as well as some of the criticism which has been levelled against such studies. I then turn my attention to the value relevance of aggregate earnings and earnings components. Of particular importance to the study that follows in Chapter 4 is the review of the literature relating to the value relevance of non-recurring earnings components. The discussion of the literature includes references to studies based on data from the United States as well as from other countries. Next, I comment on the role of the efficient market hypothesis in value relevance research. Finally, I highlight the key findings from the literature review and their implications for Chapter 4.
Background to value relevance
Value relevance research investigates the association between accounting amounts and share prices or returns (Hellström, 2006). This type of research routinely employs regressions of contemporaneous share prices or returns on accounting variables to evaluate how accounting information maps onto those measures (Aboody, Hughes and Liu, 2002). An accounting amount is considered value relevant if it has a predicted association with share prices or share returns(Barth, Beaver and Landsman, 2001; Holthausen and Watts, 2001). This approach assumes that an important function of accounting numbers is to reflect economic income as represented by share returns, and economic value as reflected by share prices (Hellström, 2006).
Value relevance research has a well-established history. One of the first papers to investigate the association between accounting amounts and share prices/returns is that by Ball and Brown (1968). Amir (1993) was one of the first researchers to use the term “value relevance”. Since then, numerous studies have employed the value relevance methodology. More recent examples include studies investigating issues such as fair value pension accounting (Hann, Heflin and Subrumanayam, 2007), fair value accounting for liabilities and own credit risk (Barth, Hodder and Stubben, 2008) and corporate restructuring charges (Jaggi et al., 2009).
Value relevance research uses accepted valuation models to assess the relevance and reliability of accounting amounts for equity investors who want to value firms‟ equity (Barth, Beaver and Landsman, 2001). Relevance and reliability are two of the qualitative characteristics of financial statements as identified in the International Accounting Standards Committee‟s (IASC‟s)10 Framework for the Preparation and Presentation of Financial Statements (the “IASC Framework”) published in 1989. This Framework has been adopted by the International Accounting Standards Board (IASB)11 since 2001. Relevance and reliability are also two of the primary criteria that the Financial Accounting Standards Board (FASB) uses to choose among accounting alternatives, as specified in its Conceptual Framework (Barth, Beaver and Landsman, 2001). According to the IASC Framework, information has the quality of relevance when it influences the economic decisions of users by helping them to evaluate past, present or future events (predictive value), or to correct their past evaluations (confirmatory value) (IASC, 1989). Information has the quality of reliability when it is free from material error and bias, and when users can depend on it to represent faithfully that which it purports to represent.
Value relevance studies typically involve joint tests of relevance and reliability; and it is difficult to attribute the extent of (or the lack of) value relevance to one or the other attribute (Barth, Beaver and Landsman, 2001). However, when support for the value relevance of an accounting figure is found, it indicates that the figure contains – or is associated with – information that is relevant to investors and that the amount is measured sufficiently reliably to be reflected in share prices (Barth, Beaver and Landsman, 2001). Hence, value relevance research attempts to operationalise the criteria of relevance and reliability empirically (Barth, Beaver and Landsman, 2001).
Another important feature of value relevance studies is that they focus on equity investments (Barth, Beaver and Landsman, 2001). This is in line with the primary objective of financial statements, as established by the IASB and other standard setters. For example, the IASC Framework lists many users of financial statements, but recognises that financial statements that meet the needs of investors are likely to meet the needs of most other users too (IASC, 1989).
Investors gain their information from numerous sources. Barth, Beaver and Landsman (2001) argue that information does not have to be new to a financial statement user to be relevant, because summarising or aggregating information that might be available from other sources is an important role played by financial statements. This comment highlights the distinction between value relevance and decision relevance: accounting information may be value relevant without being decision relevant if it is superseded by more timely information (Barth, Beaver and Landsman, 2001).
Hellström (2006) distinguishes between value relevance studies with a signalling perspective on the one hand and those with a measurement perspective on the other. Value relevance studies with a signalling perspective investigate whether there is a reaction to the announcement of accounting information, in other words, whether financial statements communicate “new” information (Hellström, 2006). By contrast, value relevance studies with a measurement perspective investigate the explicit relationship between the market indicators of the value of a firm and accounting measures (Hellström, 2006). Such value relevance studies are designed to establish whether particular accounting amounts reflect information that is used by investors in valuing a firm, regardless of where the investor obtained the information. Hence, value relevance research often adopts the perspective that accounting data may be regarded as a summary of the events that have affected the firm, irrespective of whether or not the accounting data contain any new information (Easton, 1999). Value relevance implies that equity values reflect an accounting amount if the two are correlated (Barth, Beaver and Landsman, 2001).
Criticism of value relevance research
Holthausen and Watts (2001) question whether value relevance studies are able to contribute to standard setting decisions. In particular, they criticise three assumptions that are often made by value relevance studies, but that are inconsistent with FASB statements:
- that equity investors are the dominant users of financial reporting and the valuation of equity is the dominant purpose of financial reporting;
- that share prices adequately represent equity investors‟ use of information in valuing equity securities; and
- that share-price based tests of relevance and reliability measure relevance and reliability, as defined in the FASB‟s statements.
Holthausen and Watts (2001) argue that the inconsistency of the tests of value relevance studies on FASB statements limits the ability of these studies to make a meaningful contribution to standard setting decisions. Holthausen and Watts (2001) thus hold a contrary view to that of Barth, Beaver and Landsman (2001) (see Section 3.2.1).
During September 2010, the IASB published the Conceptual Framework for Financial Reporting 2010 (the “revised Framework”), while the FASB published similar content as Concepts Statement No. 8 to replace Concepts Statements No. 1 and No. 2. The revised framework contains the boards‟ updated views on the concepts that underlie financial reporting and that should guide them in their standard setting decisions. In this section, I provide an updated view on the consistency of value relevance research related to the statements of the boards.
Holthausen and Watts‟s (2001) first point of concern in respect of value relevance research is its exclusive focus on equity investors. The objective of general purpose financial reporting, according to the revised Framework is (IASB, 2010: par OB2), is to provide financial information about the reporting entity that is useful to existing and potential equity investors, lenders and other creditors in making decisions about providing resources to the entity.
The stated objective of financial reporting is narrower and more direct than previous statements by the boards in that it identifies both the users and the uses of financial statements.
The revised Framework identifies capital providers, namely equity investors, lenders and other creditors, as the primary group of financial statement users. Value relevance research generally does not consider lenders and other creditors. However, in many instances, equity investors are the primary and major capital providers. Hence, by focusing on equity investors, value relevance research provides useful insights for standard setting.
Capital providers use financial reporting in making decisions in their capacity as capital providers (IASB, 2010: par OB3). In this regard, the revised Framework states:
Decisions by existing and potential investors about buying, selling or holding equity and debt instruments depend on returns that they expect from an investment in those instruments, for example dividends, principal and interest payments or market price increases. . . Investors‟, lenders‟ and other creditors‟ expectations about returns depend on their assessment of the amount, timing and uncertainty of (the prospects) for future net cash inflows to the entity. Consequently, existing and potential investors, lenders and other creditors need information to help them assess the prospects of future net cash inflows to an entity.
This statement by the boards provides a clearer link between the use of accounting information and equity valuation. Nichols and Wahlen (2004) summarise the theory that links financial reporting to equity values and returns. The theory establishes three links. First, consistent with the boards‟ statements, the theory establishes that financial reporting provides information to equity investors about current and expected future profitability. Second, the theory asserts that current and expected future profitability provides equity investors with information about the firm‟s current and expected future dividends to shareholders. Finally, the theory assumes that share prices equal the present value of expected future dividends to shareholders.
Holthausen and Watts‟s (2001) second point of concern regarding value relevance studies relates to the fact that, in their view, the statements of the FASB suggest that the FASB is interested in individual investors, not in investors in the aggregate, as represented by the stock market. This concern no longer appears valid, as the boards assert that their basic mission is “to serve the information needs of participants in capital markets” (IASB, 2010: par BC1.23). This mandate is clearly focused on capital markets.
Holthausen and Watts‟s (2001) third point of concern relating to value relevance literature is that the attribute of verifiability in the FASB‟s definition of reliability is not captured by value relevance research. They present two arguments relating to value relevance studies that investigate accounting figures before the disclosure of the amounts are required by standards (“pre-standard data”). Firstly, they argue that pre-standard data may display value irrelevance, due to a lack of verifiability, but may subsequently become value relevant once the data are required by accounting standards and once the data have been audited. Alternatively, pre-standard data may initially show value relevance, but may subsequently become value irrelevant once the data are required by the accounting standards, due to incentives to management to misrepresent amounts, reducing verifiability.
Verifiability refers to one of the qualitative characteristics of financial information. The qualitative characteristics of financial statements are those attributes that make financial information useful (IASB, 2010). The revised Framework distinguishes between “fundamental qualitative characteristics” and “enhancing qualitative characteristics”, depending on how these qualitative characteristics affect the usefulness of information (IASB, 2010). Fundamental qualitative characteristics are essential: in other words, they represent the characteristics that financial information must possess to be useful. Qualitative characteristics that enhance the usefulness of financial information are complementary to the fundamental qualitative characteristics; they distinguish more useful information from less useful information (IASB, 2010).
The revised Framework identifies relevance and faithful representation as the two fundamental qualitative characteristics, whereas comparability, verifiability, timeliness and understandability are identified as enhancing qualitative characteristics. It is worth noting that reliability is no longer listed as a separate qualitative characteristic, but is subsumed under faithful representation. The boards define “faithful representation” as follows (IASB, 2010: par QC12):
Financial reports represent economic phenomena in words and numbers. To be useful, financial information must not only represent relevant phenomena, but it must also faithfully represent the phenomena that it purports to represent. To be a perfectly faithful representation, a depiction would have three characteristics. It would be complete, neutral and free from error. Of course, perfection is seldom, if ever, achievable. The Board‟s objective is to maximise those qualities to the extent possible.
Regarding the requirement that financial statements be “free from error”, the boards state (IASB, 2010: par QC15):
In this context, free from error does not mean perfectly accurate in all respects. For example, an estimate of an unobservable price or value cannot be determined to be accurate or inaccurate.
Regarding the enhancing qualitative characteristic of “verifiability”, the boards state (IASB, 2010: par BC3.36):
Some respondents pointed out that including verifiability as an aspect of faithful representation could result in excluding information that is not readily verifiable. Those respondents recognised that many forward-looking estimates that are very important in providing relevant financial information (for example, expected cash flows, useful lives and salvage values) cannot be directly verified. However, excluding information about those estimates would make the financial reports less useful. The Board agreed and repositioned verifiability as an enhancing qualitative characteristic, very desirable but not necessarily required.
Hence, it is evident that the boards are downgrading verifiability from a key concept that is supposed to guide standard setting to an enhancing qualitative characteristic. However, verifiability may affect investors‟ consensus beliefs – hence, as Holthausen and Watts (2001) point out, value relevance research investigating pre-standard data needs to be interpreted with caution, because the value relevance may be affected once the information is mandatory in terms of the particular accounting standards. This concern is not relevant to this study, because the study investigates the value relevance of information that is already required by regulation.
Despite Holthausen and Watts‟s (2001) criticism of value relevance research, the boards have given the following recognition to this area of research, while also acknowledging its potential shortcomings (IASB, 2010, par BC3.30):
Empirical accounting researchers have accumulated considerable evidence supporting relevant and faithfully represented financial information through correlation with changes in the market prices of entities‟ equity and debt instruments. However, such studies have not provided techniques for empirically measuring faithful representation apart from relevance.
In the boards‟ joint project on financial statement presentation, they also cite a number of value relevance studies in support of their decisions (IASB, 2008). For example, citing Dechown(1994), the boards claim that “academic research demonstrates that accrual accounting produces income numbers that are more highly associated with stock returns than are cash flows from operations or the change in cash during the period” (IASB, 2008: par 4.26). The boards also cite Barth (1991) and Landsman (1986), who argue that “research indicates that assets and liabilities recognised using accrual accounting are significantly associated with stock prices at the end of the period” (IASB, 2008: par 4.26). The boards therefore conclude that these “findings are consistent with the view that accrual accounting provides useful information in the statements of financial position and comprehensive income that is not provided by cash-basis accounting” (IASB, 2008: par 4.26).
These quotes not only provide evidence that the boards do take cognisance of the value relevance literature in making standard setting decisions, but also show that the boards regard an association between accounting amounts and share prices/returns as evidence that the information is useful.
Recent value relevance studies continue to make claims about providing feedback to standard setters. In a recent study on the market valuation of accrual components, Francis (2008: 151) comments:
The study also provides feedback to the Financial Accounting Standards Board (FASB) concerning the usefulness of accounting information by highlighting equity investors‟ perceptions of the cash flow and accrual components of earnings.
Value relevance of earnings
In a landmark study in accounting research, Ball and Brown (1968) assess the usefulness of accounting earnings by investigating its information content and timeliness as evidenced by information contained in share prices and returns. They assume that capital markets are efficient and unbiased, in other words, that if information is useful in forming share prices, the market adjusts quickly to reflect that information (Ball and Brown, 1968). Their study has resulted in a paradigm shift in accounting research, because research prior to their 1968 study was generally based on a theoretical analysis of the usefulness of accounting information, as opposed to an empirical enquiry (Ball and Brown, 1968; Nichols and Wahlen, 2004). The results from their study suggest that the information captured in accounting earnings is useful, because, when the actual income differs from the expected income, the market typically reacts in the same direction (Ball and Brown, 1968).
Since Ball and Brown‟s (1968) study appeared, researchers have documented evidence that suggests that various factors affect the strength of the returns-earnings association. Collins,Maydew and Weiss (1997) list various factors that decrease the value relevance of accounting earnings, such as membership of intangible-intensive industries and small firm size. The inclusion of transitory earnings components in earnings also appears to reduce the strength of the returns-earnings association (Collins et al., 1997; Dechow, 1994; Kothari, 2001).
The issue of earnings persistence is considered further in Section 3.2.5. Basu (1997) shows a non-linearity in the returns-earnings association, because losses (“bad news”) are included in earnings more quickly than gains (“good news”). In a reverse regression of earnings on returns, the slope coefficient and the adjusted R-square of returns are higher for firms with negative returns than for firms with positive returns. This is similar to the findings of Hayn (1995), who shows that the slope coefficient and adjusted R-square of earnings in a regression of returns on earnings are higher for firms with profits than for firms with losses.
Harris and Ohlson (1992) argue that there are “errors” in earnings, as share prices lead earnings in capturing value relevant information, resulting in a mismatch between information contained in earnings and information contained in contemporaneous returns. When they increase the return period, they find a significant increase in the adjusted R-square of the returns-earnings regression. Similarly, Easton et al.‟s (2000) study shows that the effect of the accounting recording lag is to draw the estimate of the earnings coefficient towards zero. Ramakrishnan and Thomas (1998) argue that earnings management may also lower earnings response coefficients. Another factor that may have an impact on the value relevance of accounting information is market efficiency (Hellström, 2006; Holthausen and Watts, 2001). This is considered in more detail in Section 3.2.10 below.
An important empirical question addressed by the literature is whether accrual earnings is a better summary measure of firm performance than cash flows, or vice versa. This issue is addressed in an early study by Dechow (1994), but the issue remains relevant, as is evident from a more recent study by Penman and Yehuda (2009).
Dechow (1994) investigates whether accrual earnings or cash flows have a stronger relative association with contemporaneous returns. She argues that the primary role of accruals is to overcome problems associated with cash flows in measuring firm performance when firms are in continuous operation. Realised cash flows have timing and matching problems that make it a “noisy” measure of firm performance. Accruals address the problems associated with realised cash flows so that earnings reflect firm performance more closely. Her evidence shows that the returns-earnings association is stronger than the returns-cash flow association over short measurement intervals, when aggregate accruals are of a large magnitude, and the longer a firm‟s operating cycle.
Penman and Yehuda (2009) argue that finding a positive slope coefficient on a returns-cash flow regression, as is the case in Dechow (1994), is at odds with economic theory, which predicts that, given earnings, free cash flow is value irrelevant. They specify a regression of returns on earnings, dividends and the book value of equity, and find that the coefficient on earnings is positive and significantly different from one. When free cash flow is included in the regression, the coefficient of earnings remains fairly unchanged, while the free cash flow variable is not significantly different from zero. In their view, these findings confirm the notion often held in accrual accounting that accrual earnings and not free cash flow affects the value of equity. Although Penman and Yehuda (2009) analyse the components of free cash flow, they do not analyse the components of earnings. Cash flow is a component of earnings (Sloan, 1996); hence,the positive association between returns and earnings may in part be attributed to the cash flow component of earnings.
Chapter 1: Introduction
1.2 The study problem
1.3 The South African case
1.4 Summary of finding
1.6 Organisation of the thesis
Chapter 2: Background to headline earnings
2.2 The conceptual basis of headline earnings
2.3 Theoretical links
2.4 The history of headline earnings
2.5 The definition of headline earnings
2.6 Alternative earnings measures
2.7 Headline earnings and comprehensive income
Chapter 3: Background to value relevance
3.2 Prior research
Chapter 4: Mandatory earnings disaggregation and the value relevance of earnings components
4.2 Prior research and hypotheses
4.3 Sample selection
4.4 Research design
4.5 Descriptive statistics
4.7 Sensitivity analyses
4.8 Alternative explanation
Chapter 5: Value relevance: Additional analyses
5.2 Prior research and hypotheses
5.3 Research design
5.5 Additional ex post analyses
Chapter 6: Background to the accrual anomaly
6.2 Prior research
Chapter 7: Mandatory earnings disaggregation and the persistence and pricing of earnings components
7.2 Background to the JSE
7.3 Prior research and hypotheses
7.4 Sample selection
7.5 Research design
7.6 Descriptive statistics
7.8 Sensitivity analyses .
Chapter 8: Conclusion
8.6 Suggestions for future research
8.7 Concluding remarks
GET THE COMPLETE PROJECT
MANDATORY EARNINGS DISAGGREGATION AND THE VALUE RELEVANCE, PERSISTENCE AND PRICING OF EARNINGS COMPONENTS: SOUTH AFRICAN EVIDENCE