Background and Case for Exit Price Accounting

BACKGROUND

This symposium continues the search for the elusive fundamentals of accounting measurement—fundamental foundations that researchers and practitioners have for decades sought in developing a constrained optimal accounting system for reporting assets, liabilities and measuring income. Churchman (1961) noted constraints in deriving those foundations. There, he observed that associated costs and benefits differ with the many possible measurement functions. Myriad panellists' views about the proper function and other issues reveal many contestable implications that flow for measurement.

Below, Mary Barth notes that those forging the various national and international conceptual frameworks (CFs) underpinning contemporary accounting systems have chosen the function of accounting to be the provision of cash flow information relevant for decision making. Given that, standard setters have specified that asset and liability items be reported using a mixed set of attribute measures (costs, market prices, expectations and estimates of various types). This has come to be referred to as a ‘mixed attribute system’. Market based valuations within that set have taken many forms: for example, exit price (selling), entry price (replacement price), deprival value (DV) and a catch-all fair value accounting (FV).¹

Nearly fifty years after most of those measurement bases were first placed in the professional accounting lexicon, accounting measurement remains contentious. This is particularly so given the recent implementation of fair value market measures (FVs) in several of the IFRSs and the contemporary allegations (as noted by Laux and Leuz, 2009) that mark-to-market valuations in accord with FASB 157 have fuelled the 2007–09 sub-prime-cum-global financial crisis (GFC).²

Having organized the structure of the symposium and invited the panellists to participate, I introduced matters by noting the ideas that might be assumed to have underpinned John Bennet Canning's 1929Economics of Accountancy³—a potentially path-breaking book—which, arguably for the first time in the Anglo-Saxon accounting literature, articulated the need to adhere to the prevailing generic canons (foundations) of measurement. In this context correspondence materials previously unpublished from the so-called ‘preparatory phase’ to the ‘Golden Age’ of accounting researchers were deemed particularly relevant. That correspondence addresses issues that had underpinned the early paper on measurement by Ray Chambers, the 1959 ‘Measurement and Misrepresentation’ (a TIMS conference paper), eventually published in 1960 in Management Science. This presaged several ‘accounting measurement’ articles in The Accounting Review, like Homburger's 1961‘Measurement in Accounting’, Harold Bierman's 1963‘Measurement and Accounting’, and books such as Edwards and Bell's classic, The Theory and Measurement of Business Income (1961), Ricco Mattessich's 1964 path-breaking Accounting and Analytical Methods: Measurement and Projection of Income and Wealth in the Micro- and Macro-Economy . . . , Chambers' (1966) magnum opus, Accounting, Evaluation and Economic Behavior and Yuji Ijiri's (1967) axiomatic work, The Foundations of Accounting Measurement: A Mathematical, Economic, and Behavioral Enquiry and George Staubus' (1961) A Theory of Accounting to Investors.⁴

Harold Bierman's ‘new’ focus on measurement provided an opportunity for what he described as a ‘revolution in accounting thought’ and practice. The opportunity lapsed. Supporting this view I specifically outlined evidence from the recent archival research using the Chambers Collection Archive (http://chamberslibrary.econ.usyd.edu.au) into the 1950s and 1960s measurement debates, noting how that work sheds light on recent controversial professional deliberations on identifying, where observable prices are not available, an appropriate measure and presentation of the worth of financial institutions' toxic assets (e.g., ABSs, CDOs or CDSs). Earlier arguments of several Golden Age researchers were recounted, suggesting that identifying and applying the canons of measurement could assist in resolving such issues, whatever the prescribed measurement attribute (given a specified accounting function). Being a member of the Sydney School of Accounting, not surprisingly I argued the case for the current cash equivalent (a form of exit price) attribute to account for assets and liabilities. Emphasis on that attribute differentiates it from the professionally prescribed mark-to-model variant of FV accounting discussed below.

Against that background, three panellists' fifteen-minute presentations followed. These discussed other measurement attributes—notably entry price, replacement costs, FV and DV. Andrew Lennard, Director of Research at the ASB, began by developing the case for entry values; Geoffrey Whittington, former IASB member, spoke about the need to place measurement in the context of the market setting (leading to an informational approach consistent with more eclectic measures such as DV); Richard Macve, from the LSE, presented the case for DV—noting the DV rule for choosing between measurement bases as well as its use for liabilities, its relationship to revenue recognition (as in the latest IASB DP) and addressed whether DV is useful to internal or external users. Finally, the fourth panellist, Mary Barth (a former IASB member), observing the proceedings from a standard setter's perspective, responded briefly to questions raised in the presentations and others from those in the audience.

Each panellist was asked to convert their presentation into an approximately 3,000 word paper, with a view to highlighting critical questions for standard setters to address. Drawing on the panellists' powerpoint presentations, updated for events occurring between the symposium and the time of publication, the papers are reproduced in the order they were presented.

Mark Twain's attributed quotation, ‘History may not exactly repeat itself but it rhymes’, introduced the symposium, conditioning the audience to the argument that resolving the measurement issue would require first agreeing on the function of accounting (as noted also by Whittington, 2008, and Bradbury, 2008). As well, one needs to adhere to the measurement canons or foundations, namely: (a) the need for a common attribute (property) to be measured, (b) measuring change in wealth entails observations of initial and terminal states, and (c) if the conditions upon which measures are taken at those states have changed then those measures need to be standardized through scaling.

The presenters were asked to consider matters related to current measurement issues facing standard setters worldwide. Particularly in focus are FV implementation issues in many IFRSs, and the contemporary dispute regarding the influence of mark-to-market (M-t-M or FV) valuations for financial institutions holding toxic assets underpinning the institutions' (and some commentators') special pleading to have SFAS 157 effectively changed.⁵ Three matters were canvassed briefly in my introductory remarks: (a) canons of measurement, (b) politicization and special pleading—déjà vu; resulting in (c) professional failure to adhere to any canons of measurement.

CANONS OF MEASUREMENT

Since Canning'sEconomics of Accountancy (1929) introduced into accounting many of the ideas of his mentor and leading 1920s economist Irving Fisher (some contiguously being developed in the works of contemporary European business economists like Fritz Schmidt, Die organische Tageswertbilanz, 1921), calls have emerged to break the crust of custom underpinning accounting thought and practice, to adhere more to the foundations of measurement.

To many, like Chambers (1979), Whittington and Skerratt (1980) and Mattessich (1995), Canning proposes an entry price (or a related notion of replacement cost, which is an allocated entry price) as a substitute for true economic present value. Chambers regarded Canning as having introduced other more fundamental measurement issues, in particular a desire that measures were to be constrained by a common property, a common measurement scale, and a common unit.⁶ Chambers referred to these as canons or foundations of accounting measurement.

While not all participants would agree on how to operationalize those foundations, the symposium focused on several different accounting measurement properties that might, given a specified accounting function, be deemed appropriate. Such an overview is critical as the recent CF deliberations demonstrate that the objective (function) of accounting and hence measurement remains in dispute. Panellists' views contrasted with Chambers' preferred measurement property, current cash equivalent (sometimes attributed to Chambers as exit or selling price)—derived from his observed function of accounting, namely, to provide data about an entity's financial capacity to adapt. This position was first systematically outlined in his 1961 monograph, Towards a General Theory of Accounting, and is probably best known from his magnum opus, Accounting, Evaluation and Economic Behavior (1966). Other exit price advocates include Sterling (1970) and Rosenfield (2006).

But for exposition, let us consider several measurement definitions.⁷

For Homburger (1961) ‘requirements of measurement are satisfied by the assignment of numerals and subsequent mathematical manipulation’. Torgerson (1958/1967, p. 14) notes: ‘Measurement of a property . . .  involves the assignment of numbers to systems to represent that property. In order to represent the property, an isomorphism, i.e. a one-to-one relationship, must obtain between certain characteristics of the number system involved and the relation between various quantities (instances) of the property to be measured.’ While others noted that measurement is not merely the assignment of numbers to phenomena according to rules. Chambers (1966, p. 89) pursues this aspect, drawing on Campbell'sFoundations of Science (1957, p. 267): ‘measurement is the process of assigning numbers to represent qualities’—and continuing at p. 101, Chambers observes: ‘Measurement is the assignment of numbers to objects and events according to rules specifying the property to be measured, the scale to be used, and the dimension of the unit’, and on p. 104 in fn 1, drawing on Margenau, The Nature of Physical Reality, pp. 369–70: ‘ “measurement involves (1) an object or system upon which an operation is to be performed; (2) an observable whose value is to be determined; and (3) some apparatus by means of which the operation can be carried out.” . . . [he concludes] In our setting the subject is an individual asset or equity (or collection of assets or equities); the observable is its cash equivalent; the apparatus is the market.’

Mattessich (1995, pp. 49–51) provides an overview of where he (together with the likes of the AIA, Moonitz, Stamp and the FASB) and many others (such as Chambers, Margenau, Spacek, Rosenfield and Sterling) disagree.⁸ Disagreement centres partly on whether expectations are acceptable as measures accounting. Consider Mary Barth's assessment: ‘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’. But also the disagreement is a function of how authors view time in relation to accounting—namely, where a distinction is made between retrospective, contemporary and anticipatory aspects. The last is excluded from consideration in CoCoA, but is not in some other accounting systems.

To finish this definitional perspective, two more (current and proposed) definitions from the IASB are:

While a new definition proposes:

POLITICIZATION AND SPECIAL PLEADING—DÉJÀ VU⁹

As noted, the symposium presentations examined GFC-related measurement issues. Commentators and lobbyists suggested that accounting exacerbated the GFC. The aftermath of many economic booms and busts have presaged similar claims for three-quarters of a century with Fritz Schmidt (1927) and Gottfried Haberler (1937) being two early examples.¹⁰ In the current debates M-t-M or FV accounting has been claimed to have fuelled the GFC, to have been a pro-cyclical agent leading to unexpected corporate collapses and consequential government bailouts of financial institutions.¹¹

The financial sector strongly pleaded that M-t-M prescriptions are not what is required in the current, purportedly unusual, illiquid or inactive markets—proposing that FV accounting embodied in FASB 157 (and its IASB counterpart) be suspended. This was initially suggested at the height of the GFC in September–October 2008 in the U.S. The politicization campaigns were extended in mid-October to the IASB's domain in respect of IAS 39 where carve outs followed representations by French and other European politicians (see Hinks, 2008). Returning to the U.S. and SFAS 157 special pleading resurfaced in early April 2009.¹² Debates on these occasions were vacuous, lacking evidence of any causal links between accounting and similar crises in previous periods. Evidence is available from episodes in the post-1930s Depression period, in the aftermath of the 1965–67 (primarily U.S. and U.K.) conglomerate boom/bust imbroglios, in Australia's 1967–70 mining boom and bust, in the post-1980s U.S. Savings and Loans crisis period, in the U.K. 1970s Secondary Banking crisis, as well as in the early 1990s Japanese banking and property crisis with its associated ‘lost decade’ of economic growth. It reveals some similarities to the present, such as: undercapitalization, excessive leverage, regulatory failures (sometimes due to deregulatory initiatives, while on other occasions the problems were due to poorly considered regulatory interventions) and greed. And there are dissimilarities: the asset classes differed—property, inventories, financial instruments including derivatives, even companies in the case of the merger manias. Yet a pervasive similarity is that accounting is asserted as ‘a’ or ‘the’ problem. But there is little evidence that accounting was ‘a’ or ‘the’ cause of the problems. Interestingly, it should be noted that the accepted or prescribed forms of accounting have differed over those various boom/bust periods.

The latest politicization of accounting has created outrage from those seeing the independence of professional standard setters being eroded. To many, the FASB caved in to pressure on 9 April 2009—issuing FAS FSP 157-4, after allowing only about a week for submissions. Approximately 360 submissions were received—yet, notably, the FASB's decision was opposed by more than half those submissions. Bob Herz, FASB Chairman (26 June 2009) lamented events at that time, specifically noting that some major companies, including federal bailout recipient American International Group Inc., had sought political intervention into accounting standard setting. He observed: ‘While that is their right . . .  politicization of accounting standard setting by special interest risks undermining public confidence’.

That April 2009 FASB saga once again saw pragmatism prevail. It resulted in the revised FASB 157 statement providing U.S. financial institutions with greater ability to exclude losses due to toxic assets' declines from their income calculations and to give banks greater flexibility with the M-t-M models used to determine the reported value of their toxic assets. Accounting rule makers thereby relieved financial institutions of any semblance of serious, verifiable compliance with the M-t-M rule. They have sanctioned unreality in accounts—the ultimate in untrue and unfair disclosures. Institutions will now find it easier to leave non-marketable toxic financial assets in their balance sheets at considerably more than they could fetch in the market. And where there is no market for, say, ABSs, CDOs or CDSs more flexible guidelines allow the models they use to invent a non-existing price. Some commentators have referred to the outcomes as ‘dodgy’ or ‘gimmicky’ numbers—the ultimate accounting humbug.¹³ Further, the FASB's April 2009 revision provides more flexibility for directors/managers in classifying financial assets—whether to be held to maturity or available for sale. It is the latter that attracts the M-t-M valuations, but now with greater flexibility regarding the models used by financial institutions to invent a number where the market is thin, so-called distressed, or where it is deemed that no market exists. How such misleading balance sheets can do other than obfuscate or further mislead the market is anyone's guess.¹⁴

PROFESSIONAL FAILURE TO ADHERE TO CANONS OF MEASUREMENT

As a member of the Sydney School, and accepting the capacity for adaptation function noted above, I (along with my co-author on many exercises, Frank Clarke, and other Sydney colleagues) support in principle the current exit price version of the proposed M-t-M basis for asset valuation embodied in IFRSs. The objective of accounting, at least for those of the Sydney School, is as far as practically possible to ‘tell it as it is’ in terms of expressing assets and liabilities at their current cash equivalents, and to require companies thereby to disclose serviceable indications of their current financial positions. This provides an indication of the entities' capacity for adaptation.

While we contest the latest revisions to SFAS 157 there is a prospective positive point of departure in the current brouhaha. The ongoing converged CF deliberations give international standard setters like the IASB and FASB an opportunity to rail against political pressure. It is time, arguably, that business politicization of accounting through special pleading is, as the FCAG (2009) suggests, rejected—a suggestion also positively avowed by the IASB/FASB in the latest CF Discussion Papers.¹⁵ In particular, the opportunity exists to reconsider the need for an accounting technology to adhere to the canons of measurement—codified into the profession's conceptual framework project on measurement. In this regard the views of Chambers (1974a) ‘Introduction’ to the re-issue of Accounting, Evaluation and Economic Behavior were noted:

It is not suggested that the result of applying such foundations to whatever accounting measurement attribute is specified will result in a true, unique profit figure. But it will result in an accounting technology that attempts to measure things rigorously, as they are measured in the physical sciences. The end result undoubtedly will have errors, but errors in estimation, as opposed to what some commentators infer about the current reported figures, namely that they are at the discretion of management—producing the above-mentioned dodgy or gimmicky numbers. Presently, such claims follow from the figures being the product of mark-to-model calculations under the accepted professionally prescribed SFAS 157 Levels 2 and 3 measurement hierarchy—subjective numbers based on forecasts—expectations-based data that arguably are impossible to verify by external parties like an auditor. M-t-M in this guise is nothing but speculation incapable of corroboration. Accounting standard setters and regulators surely cannot afford much longer to have outsiders deride accounting, by virtue of not having rigorous measurement foundations underpinning its valuation practices.

IN SUM

Attention was drawn to some historical nuggets from analyses of certain 1955–64 correspondence from within the R. J. Chambers Collection (http://chamberslibraryecon.usyd.edu.au). Inferences were made about their significance regarding the measurement dilemmas facing today's standard setters. Ernest Weinwurm, a well-known contemporary engineering economist, but ‘little known academic accounting soldier’—was considered. In a working paper, Clarke and Dean (2009) show that Weinwurm facilitated exposure in the 1950s and 1960s of Ray Chambers' ideas to a wider, especially U.S., audience including the AAA, AICPA and TIMS. This eventually helped him to penetrate the influential deliberations on accounting theory. Regarding similar issues to those currently vexing the FASB/IASB and regulators in determining fair value measures where no direct price exists, a prescient Chambers wrote to Weinwurm (30 July 1964):

Later, Chambers (1970, 1974b: ‘second’ and ‘third’ thoughts, published in this journal) contended that if there was no observable price or reasonable approximation (corresponding to the IASB/FASB Levels 2 and 3) then a zero value should be reported in the balance sheet. But he implied that the existence of the asset or liability should be disclosed. Later, Chambers (1976, p. 45) proposed in principle the use of a double account system for this form of disclosure for several politically intractable accounting issues like the present one.¹⁷ How to report or disclose impairment issues in illiquid or no markets is paramount today in the minds of preparers, regulators and standard setters.¹⁸

The above leads to several questions:

If these questions are not adequately addressed by standard setters, the latest conceptual framework exercise could become another example of Sterling's‘recycling of ideas’ without ‘resolving problems’ (1975, 1979). Matters raised here, in the panellists' presentations and related questions in this symposium, provide the basis for some progress by standard setters.