Abacus
Volume 46, Issue 1 pp. 120-127
Full Access

Comments on Panellists

MARY E. BARTH, 

MARY E. BARTH

Graduate School of Business, Stanford University

Search for more papers by this author
First published: 24 February 2010
https://doi.org/10.1111/j.1467-6281.2010.00311.x
Citations: 6
Mary E. Barth ([email protected]) is Joan E. Horngren Professor of Accounting, Graduate School of Business, Stanford University.

I appreciate the helpful comments of Hilary Eastman, Wayne Landsman and Henri Venter.

I appreciate the opportunity to contribute to this measurement symposium overview and provide my perspective—as an academic and former standard setter—on some of the issues that arose during the symposium and in the discussions by the symposium panellists. This symposium comes at an opportune time because the current debate relating to measurement in financial reporting could benefit from academic input.

Graeme Dean appropriately reminds us of the scholarly thought that has gone into the issue of accounting measurement in the past, which resulted in the development of the canons of measurement. I do not believe that most practising accountants were taught to think about accounting measurement in this way. Thus, summarizing this body of literature and relating it to the questions standard setters are grappling with today should help standard setters, practising accountants and academics to think conceptually about measurement. Such conceptual thinking about measurement is a contribution academics can make to this debate.

The remaining panellists raise many issues and concerns about the current measurement debate. My goal is to provide my perspective on some of these issues, at times cross-referencing my thoughts to those of the panellists. The standard-setting debate about measurement is taking place in two parts—one is in the Measurement Phase of the International Accounting Standards Board's (IASB) Conceptual Framework (CF) project and the other is in the Fair Value Measurement (FVM) project. The Measurement phase of the Conceptual Framework (MCF) project addresses issues related to developing concepts to identify measurement bases that are appropriate to use in financial reporting and to select a particular measurement basis in a particular standard-setting circumstance. In contrast, the FVM project addresses issues related to defining the term ‘fair value’ as it is used in accounting standards. The FVM project does not deal with the circumstances in which fair value should be used. To clarify the difference between these two sets of issues, my comments focus on these two parts of the debate separately. A productive debate requires all who participate to use the same terms in the same way; doing otherwise obfuscates the debate. Thus, in an effort to enhance the productivity of the debate, I provide some observations about when I believe that participants in the debate, including the panellists, ‘talk past each other’ because they use the same words to mean different things, or use different words to mean the same thing. I also provide some concluding observations.

MEASUREMENT PHASE OF THE CONCEPTUAL FRAMEWORK PROJECT

The existing Framework contains very little discussion on measurement. The discussion it contains simply lists some of the measurement bases or techniques used in existing standards. The list is not complete and there is no discussion of criteria that should be used when selecting a measurement basis in any particular situation. Thus, a measurement section in the Framework is sorely needed.

I believe that it is fair to say that the IASB, jointly with the Financial Accounting Standards Board (FASB), is struggling to craft a measurement section for the Framework. The joint IASB/FASB project staff attempted to reason about measurement concepts from first principles, including considering whether something is, in fact, a measure. Although from a theoretical point of view, one can legitimately ask whether accounting measurement can depict quantitatively objects, events, and expectations, as a practical matter, specifying such measurements is what accountants do. Thus, the focus of the Measurement phase of the CF (MCF) shifted to adopting a more practical approach to measurement, despite the fact that the outcome of approach would be part of the ‘conceptual’ framework. I believe that the change in focus reflects several factors: the difficulty of dealing with accounting measures from a theoretical measurement point of view, the wide differences of views as to which measures are appropriate in financial reporting and for what reasons, and the concern that if a conceptual approach were adhered to, the outcome could be the selection of a single measurement basis and that basis would be fair value.

As a result, the IASB and FASB are developing guidance within a mixed-measurement system. However, this mixed system envisions a mix of some form(s) of modified past price, that is, modified historical cost, and some form(s) of current value. The panellists in this symposium seem to accept that a modified past price—or, equivalently, modified historical cost—typically does not provide relevant information to users of financial statements. Thus, most of the symposium discussion focuses on which current value measure should be used, for example, fair value, replacement cost, value-in-use, or deprival value. The MCF project is now focusing on identifying suitable measurement bases and developing guidance that explains how the objective of financial reporting and qualitative characteristics of financial reporting information apply to the selection of a measurement basis. Beginning with the objective of financial reporting as articulated in the CF is critical because the objective specifies what financial reporting is meant to achieve.

Applying the qualitative characteristics (QC) also is important in selecting a measurement basis. As the Exposure Draft (ED) of the first two chapters of the improved CF (IASB, 2007) explains, the two fundamental QCs are relevance and faithful representation. These fundamental QCs are supported by enhancing QCs—comparability, verifiability, timeliness and understandability. These are enhancing QCs because their existence enhances the fundamental QCs; it does not result in relevance or faithful representation. For example, although verifiability can enhance accounting information, information does not become relevant or faithfully represented simply because it can be verified. Thus, verification is not a fundamental QC. In fact, many accounting amounts in financial statements today are based on non-verifiable information, for example, bad debt allowances, loan loss provisions, and useful lives of property, plant and equipment.

As noted above, the symposium panel implicitly seems to accept that some type of current value is more relevant to financial statement users than is modified historical cost. Thus, in their thinking, they implicitly—or explicitly—apply one of the fundamental QCs, relevance. Regarding faithful representation, the objective of any measurement is to obtain a faithful representation of the asset or liability being measured. However, there could be differences of opinion about which measure is more representationally faithful. Some believe that if two entities hold the same asset, then that asset should be reflected in the two entities' financial statements in the same way. Those who hold these beliefs view such similar reflection as increasing comparability, which is another enhancing qualitative characteristic of financial reporting information. They would reflect any entity-specific economic constraints and opportunities in some other way in the financial statements. Others believe that it is more representationally faithful to portray the economic position of the holder of an asset, rather than to portray the position of a, perhaps hypothetical, market participant who ‘steps into the shoes of the entity’. Those who hold this belief view ‘economic position’ as including the effects of any economic constraints and opportunities of the entity, although they would perhaps exclude the entity's beliefs, expectations and other psychological factors. Regardless, in this view, economic position is entity-specific and, thus, the two assets are not the same and should not be reflected in the same way in the two entities' financial statements (Barth and Landsman, 1995).

Missing from the analysis, so far, is explicit consideration of the fact that the reason that the Framework focuses on measuring assets and liabilities is to measure profit or loss. Thus, any discussion of measurement bases for assets and liabilities should consider the implications for profit or loss. This issue is touched upon by symposium panellists, but they recommend disconnecting the statements of financial position and income (see, especially, Richard Macve's conclusion). That is, they recommend adopting different views of measurement for assets and liabilities and for components of profit or loss, presumably with the difference being reflected in some balancing account, such as other comprehensive income. However, disconnecting the statements loses the benefits of using changes in measures of assets and liabilities to measure profit or loss. It ignores the reason for taking the asset and liability approach in the first place. That is, it ignores the consistent conclusion of standard setters that no conceptual way to measure profit or loss exists in the absence of measuring assets and liabilities. Thus, disconnecting the two statements results in an ad hoc approach to measuring the line item amount of profit or loss, which in the view of many financial statement users is the most important line item in financial reports.

An alternative approach is to select measurement bases for assets and liabilities that result in measures of profit or loss that comport with the concepts in the CF. Because financial reporting is a social science, not a pure science, the objective of financial statements is our collective choice. That choice has been made by the IASB (and FASB) after extensive due process and public debate and deliberation. That we have statements of financial position, comprehensive income and cash flow also is the result of our collective choice. Thus, the roles of these three statements and the complementary information they contain need to be considered when deciding how to measure assets and liabilities. Whatever we include in (exclude from) asset and liability measurement affects profit or loss now (later) (see, e.g., Barth, 2006, for an explanation of this effect when using fair value and value-in-use as the measurement basis).

Using entity-specific measures for assets might provide more information about the value of the asset in the hands of the entity, for example, reflect its management skills and synergies. However, doing so results in recognizing in profit today the effects of such management skills or synergies that have not yet been demonstrated. Using fair value limits today's profit recognition to amounts market participants would assess, and leaves to the future the recognition of profit associated with management skill and synergies. Sometimes advocates of entity-specific asset measures also eschew ‘upfront’ profit recognition. Yet, to others, using entity-specific asset measures would do just that by recognizing the value of management skills or synergies before they are demonstrated. Of course, we might conclude that such profit recognition is appropriate. But the issue has not yet been debated in those terms.

FAIR VALUE MEASUREMENT

The Fair Value Measurement (FVM) project addresses a different issue relating to measurement. That issue is: What does the term ‘fair value’ mean when it is used in financial reporting standards? The FVM project, and the standard resulting from it, will not address the question of when fair value should be used to measure a particular asset or liability. As explained above, that is the issue being dealt with in the MCF project and, ultimately, is a standard-setting issue as measurement is considered in each standard. Because the FVM standard will be a definitional document, it is not fruitful to debate what fair value should mean from a conceptual point of view. These words simply comprise a term of art that needs clear definition so that everyone understands what they mean. Once the definition is clear, then the question becomes whether such a measure should be used in any particular situation. This latter decision is outside the scope of the FVM project.

The FVM ED defines fair value as ‘the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date’ (IASB, 2009, para. 1). Thus, the objective of fair value measurement is to determine a market exit price. The ED explains that if no market price is available for the asset or liability in question, then fair value needs to be estimated. Such estimation typically involves discounting expected future cash flows to present value, where both the cash flow expectations and the discount rate reflect assumptions market participants would use in pricing the asset or liability (IASB, 2009, para. 15). The ED does not consider the resulting estimates as necessarily inferior to observed prices—the estimates and observed prices are merely different means of achieving the fair value measurement objective. However, because estimated prices often require the use of unobservable inputs, the ED proposes note disclosures about how the fair values were estimated to aid financial statement users in interpreting recognized and disclosed amounts (IASB, 2009, paras 38 through 55).

The FVM ED attempts to clarify a point that has confused some in thinking about what the IASB means by fair value—that is, the ED is clear that fair value is not entity-specific, yet fair value can be estimated using the entity's own information and data. To clarify this point, the ED makes clear that ‘entity-specific’ refers to something that is particular to the entity that is not available to other market participants. This does not preclude an entity from using some of its own information in estimating a fair value. In fact, in some cases the entity is the only source of the information necessary to make the estimate. This seeming contradiction is resolved in the ED when it explains that when an entity uses its own information to estimate fair value, the entity needs to adjust the information if reasonably available information indicates that market participants would use different data or there are entity-specific factors that need to be removed. That is, the entity can use its own information as long as the entity believes other market participants would use the same information. The objective of fair value—which does not incorporate entity-specific factors—remains the same regardless of the source of information used to estimate it. See IASB (2009, paras 53–4).

Some are sceptical of adopting a market participant view because they believe that to do so, the reporting entity would need to identify market participants—or imagine some hypothetical ones—and then determine how those market participants would use the asset and what assumptions they would make when valuing the asset. Although fair value measurement requires taking a market participant view, the FVM ED acknowledges that most entities would use a particular asset in the same way, and that the entity's assumptions are likely to be those of a market participant in that situation. Thus, the FVM ED indicates that when estimating the fair value of an asset, the entity assumes that market participants would use the asset in the same way the entity does, unless there is evidence to indicate that the entity's use is not the highest and best use of the asset. If the entity's use is not the highest and best use, then the highest and best use must be the basis for the fair value estimate because that is what would be reflected in a market price for the asset. Similarly, as noted above, the FVM ED indicates that when estimating fair value, the entity uses its own assumptions, unless there is evidence that market participants would make different assumptions. The objective of taking a market participant view is not to require identification of particular market participants and divining their assumptions. Rather, a key objective of the market participant view is to help ensure that the entity presents an unbiased view of the asset's value.

To aid readers' understanding of the fair value definition, the FVM ED compares fair value to other current value measurement bases. For example, it explains how in some situations, replacement cost is appropriate for measuring fair value (IASB, 2009, para. 38(c)). This is not because replacement cost represents an entry value per se, which would seem inconsistent with fair value defined as an exit price. Rather, this is because a market participant would not pay more for an asset than the amount for which it could obtain (or replace) the service capacity of that asset. This matter is discussed by the panellists. Thus, in some situations and considering transaction costs separately, replacement cost can be equal to an exit price for the entity's asset.

The FVM ED also explains that exit value and entry value are the same for the same asset in same market on the same date (IASB, 2009, para. 35; see also Geoffrey Whittington's comments in this symposium regarding the law of one price). Thus, the FVM ED tries to avoid miscommunication that is inherent in framing the debate as between exit value and entry value. As Geoffrey Whittington points out, a difference between exit and entry is transaction costs. Transaction costs are the incremental direct costs incurred to sell an asset or transfer a liability. The Boards now reason that transaction costs are incurred to pay for services rendered—such as brokers' compensation reflected in a bid/ask spread and sales commissions. Thus, they do not relate to the asset acquired or liability assumed. However, the FVM ED distinguishes between transportation costs that are part of the value of the asset if location is a characteristic of the asset and, thus, reflected in the asset's market price and transaction costs (IASB, 2009, para. 16).

The objective of the FVM ED is to define the term ‘fair value’, not to consider all possible measurement bases. Thus, the discussion of other measurement bases in the FVM ED is intended only to address some common questions about the valuation premise that underpins the notion of highest and best use (see, e.g., IASB, 2009, paras BC58–BC66). The discussion is not intended to describe comprehensively the features of each measurement basis. Such a discussion is not appropriate for a document whose objective is to define the term fair value.

CONCLUDING REMARKS

There are genuine differences in views regarding measurement in financial reporting. I am realistic enough not to expect full agreement or resolution to these differences, but they need to be aired and considered by the IASB and FASB. The time is right for such discourse because the Boards are actively developing the Measurement section of their joint conceptual framework. That section will guide the Boards in crafting financial reporting standards for the foreseeable future. The IASB also has a Fair Value Measurement project, the objective of which is to define the term fair value when it is used in financial reporting standards. I am hopeful that the Fair Value Measurement standard will clarify what a fair value measure includes and excludes, which should increase understanding and facilitate the debate. This measurement symposium has provided an opportunity for the panellists—and me—to contribute to the measurement discourse. Academics have much to offer to the debate, and I hope this symposium encourages other academics to weigh in on these issues as well.

Footnotes

  • 1 The use of a mixed measurement system, in which different assets and liabilities are measured using different measurement bases, results in an additivity problem. That is, adding together amounts based on different measurement bases is analogous to adding apples and oranges. As Graeme Dean observes, the additivity problem has been recognized in the literature for many years—but the problem persists. The existence of this additivity problem is one reason many believed that the MCF would result in identifying a single measurement basis. The IASB's and FASB's decision not to pursue that goal at this time means that a mixed measurement system will persist for the foreseeable future. The Boards' joint Financial Statement Presentation project will deal with this additivity problem arising, for example, by requiring presentation of amounts disaggregated by measurement basis. Although the term ‘mixed measurement’ typically refers to a mixture of modified historical cost-based measures and fair value, using modified historical cost-based measures alone also would result in an additivity problem. This is because historical costs are incurred—and modified—at different times, which means they reflect prices at different dates. Thus, they, too, are apples and oranges. The extent to which inflation contributes to this additivity problem is not currently addressed in financial reports. However, if all assets and liabilities were measured using some form of current value, and thus measured in the same currency units, it would virtually eliminate additivity effects associated with inflation.
    Another potential additivity problem is that the sum of the firm's recognized assets minus its recognized liabilities does not equal the value of its equity. This is because recognized assets and liabilities do not include, among other items, assets related to synergies, management skill, and expectations about profitability growth. However, as the Exposure Draft of the first two chapters of the improved CF explains, financial reports are not designed to show the value of the entity (IASB, 2008, para. OB16). Thus, the Boards are not concerned with this second additivity problem.
  • 2 Comparability is also a key motivation for the adoption of International Financial Reporting Standards around the world (e.g., U.S. Securities and Exchange Commission, 2008).
  • 3 In addition, to the extent that fair value—or any other current value measurement basis—is used to re-measure assets and liabilities, profit or loss reflects holding gains and losses related to those assets or liabilities. These components of profit or loss are not triggered by transactions. This is the same as when assets and liabilities are re-measured based on any other measurement basis, for example, when they are amortized or impaired based on a modified historical cost measurement basis. The key difference in profit or loss relates to what the re-measurement represents. In the case of a current value re-measurement, it represents an economic holding gain or loss; in the case of a modified historical cost re-measurement, it represents an allocated amount.
  • 4 Some disagree that entity-specific measures provide more information about an asset. They view the difference between fair value and an entity-specific value as representing a different asset—such as management skill or synergies—not as part of the value of the asset to which the skill or synergies relate.
  • 5 This potential conflict between measurement bases for assets and liabilities and for measuring profit or loss is evident in the IASB and FASB's Revenue Recognition Discussion Paper (IASB, 2009) and in the Boards' current project on Insurance Contracts. In both cases, the Boards have decided tentatively to modify liability measurement because of the effects of the measurement on profit or loss. In the Revenue project, the Boards have decided tentatively to not measure performance obligations directly—the amount assigned to the obligation would be an allocated portion of the customer consideration. This allocation ensures there is no revenue recognized at the time the contract is signed. In the Insurance Contracts project, the Boards have decided tentatively to recognize enough revenue at contract signing to offset the costs of obtaining the contract. This avoids recognizing expenses without revenue, but it also results in not measuring directly the insurance contract liability.
  • 6 One reason a clear definition is needed is that, as Clarke et al. discuss in their 2004 comment letter to the FASB on the Exposure Draft of what became Statement of Financial Accounting Standards No. 157, the term ‘fair value’ was in use before the 1970s and did not have exactly the same meaning as it has today. For some examples see Chambers (1996, especially s. 548, pp. 520–2).
  • 7 Consistent with this distinction, the Basis for Conclusions (IASB, 2009, BC64) explains that fair value differs from value-in-use as described in IAS 36 (IASB, 2004, paras 30–1). This is because value-in-use reflects the future cash flows the entity expects to receive. Although these expectations must be based on reasonable and supportable assumptions, with greater weight given to external than internal evidence, the expectations are not intended to reflect market participant expectations as would be required for a fair value measure.
  • 8 The entry value versus exit value miscommunication also contributes to misunderstandings about the fair value of inventory. In particular, some believe that using fair value, that is, an exit price, for inventory means the inventory would be measured based on retail selling prices rather than based on wholesale buying prices, that is, an entry price. However, this is not the case. The key is to determine the asset whose fair value is being estimated. Inventory that is not in the condition or location to be sold to customers, for example, raw materials, work in process, or finished goods in a warehouse, is not the asset sold to customers in the retail market. Thus, potential buyers for such inventory are other entities that can complete the production and sales processes and sell the products to customers, not retail customers who are not equipped to do this. Although one can estimate the fair value of such inventory based on the retail selling price, that price needs to be adjusted for the additional costs and compensation for risk that the entity bears, including, for example, to convert the inventory into cash and the risk of obtaining a customer for the inventory. For example, the fair value of work-in-process inventory can be calculated as the retail price minus remaining production costs, compensation for risk/holding costs, and selling costs, or as the wholesale price plus value of production to date, and compensation for risk/holding costs incurred to date. Unless there is a market imperfection, one would expect the adjusted wholesale and adjusted retail prices to be equal, which is the exit price, that is, fair value, for the inventory—in its current location and condition (FVM ED, para. B3(f)).

      • The full text of this article hosted at iucr.org is unavailable due to technical difficulties.