Abstract
I cannot endorse Donleavy’s conclusion that “fair value” accounting is derived from the eighteen-century economic thinkers: Anne Robert Turgot; Richard Cantillon; and Adam Smith. In his well-written study, Donleavy seems to misperceive fair value accounting as the reporting of accounting items by their market values. However, fair value accounting also advocates a theoretical explanation of how market prices would be formed and why they would be fair values. It is important to clarify this point, because it leads to two subsequent conclusions. First, that the price/value theory in fair value accounting is quite distinct from Turgot, Cantillon and Smith’s theories on the same matter. For instance, Cantillon and Smith suggested cost-based theories of value, where labor was the key element. Second, it is possible to distinguish fair value accounting from periods where the case for reporting accounting rubrics by market values was promoted based on other theoretical motivations.
Keywords
Introduction
There is much to be commended in Donleavy’s (2019) “Inquiry into the origins of fair value accounting,” which contributes to the important effort toward further connecting economic theories to accounting theory and practice (Baker, 2018; Bryer, 2013; Cardao-Pito and Ferreira, 2018a, 2018b; McWatters, 2018; Markarian, 2018; Mattessich, 2003; Mouck, 1995; Oldroyd et al., 2015; Persson et al., 2018). Indeed, fair value accounting needs to be demystified, and its underlying theory must face the degree of scientific inquiry which is regularly faced by other theories (Cardao-Pito and Ferreira, 2018a). Fair value accounting has intellectual origins, rather than being the result of market imperatives or a superior theory as contended by regulators, standard setters, and many scholars (Bhimani, 2008; Donleavy, 2019; Edwards et al., 2013; Lee, 1995; Lee and Williams, 1999; Reiter, 1998; Tuttle and Dillard, 2007; Young, 2006; Young and Williams, 2008). It is these origins which need to be identified better.
For instance, the recent debate regarding the relevance of the economist Irving Fisher (1867–1947) to accounting (in general, and fair value accounting in particular) is proof that contestable and unresolved issues regarding the origins of accounting ideas and practices still remain. Bryer (2013) and Mouck (1995) argued that Fisher is very significant for the understanding of accounting history. However, Oldroyd et al. (2015) suggest that evidence for supporting this claim is missing, given that most papers and books regarding accounting history tend to ignore Fisher. To address this issue, Cardao-Pito and Ferreira (2018a) studied Fisher’s theories regarding market prices and values, written more than a century earlier. Using content analysis, these authors have shown that fair value accounting standards emanating from the International Accounting Standards Board (IASB) and the Financial Accounting Standards Board (FASB) coincide with Fisher’s theories. Markarian (2018) praises Cardao-Pito and Ferreira’s (2018a) findings and he provides further evidence to contribute to this debate. However, Baker (2018) is not yet fully convinced. He suggests that the demonstrated parallelism between Fisher’s ideas and fair value accounting could still be a coincidence or an anomaly (Baker, 2018). In turn, Cardao-Pito and Ferreira (2018b) responded by stating that the parallelism between Fisher’s writings and fair value accounting is undeniable, as conceded by Baker. Nevertheless, the “route followed by Fisher’s ideas into fair value accounting norms may have been quite complex” (Cardao-Pito and Ferreira, 2018b: 200). Indeed, many accounting researchers, standard setters and practitioners tend to use Fisher’s ideas without being aware of this fact. This example demonstrates how complex accounting origin issues can be. Indeed, in order to make an academic contribution, this note is focused on the parts where I disagree with Donleavy (2019). Nonetheless, I trust that it remains clear that I have found his paper to be well-written and interesting. Furthermore, I support his attempt to connect accounting phenomena with the history of economic thought, which I hope will result in further thought-provoking studies.
Fair value accounting must be considered with its three levels
In his analysis, Donleavy (2019) does not establish a clear distinction between “fair value” and “fair value accounting.” The title of his paper mentions fair values. The first-half of the paper discusses theories of value formulated in the eighteen century. The paper then switches to the twentieth and twenty-first centuries during the second part, where Donleavy discusses fair value accounting as similar to the conceptions of value discussed in the first-half. Nevertheless, because the word “fair” is rather ambiguous (Biondi and Suzuki, 2007; Georgiou and Jack, 2011; Ijiri, 2005), the use of the same expression in the term “fair value” can entail very distinct concepts – which have been discussed since the advent of ancient times.
When referring to the expression “fair value accounting,” I denote the accounting methodology predicted in recent standards regarding fair value accounting emanating from the FASB and IASB. Generally, this approach is similar to Donleavy’s. 1 However, it is important to be aware that the expression fair value in accounting has previously been used to imply different meanings and relations. For instance, instead of presenting a market-based alternative to historic cost accounting, fair value is an expression which has also been used in relation to the adequate accounting of historical costs or without reference to market prices (see, for instance, Clarke, 1980; Dean and Clarke, 2010; Persson et al., 2018).
Furthermore, contrary to what appears to be suggested by Donleavy (2019, abstract), fair value accounting is not yet accepted by “the standard setters of the accounting profession as the primary way to value business assets and liabilities.” Fair value accounting has only been applied to a small minority of assets and liabilities (Baker, 2018; Cairns, 2006; Nobes, 2015). It entails merely a small proportion of financial items, together with a small number of circumstances where it is used for the initial determination of cost, when no other sensible alternatives are available. The current dominant accounting system is a mixture of historic costs and current market prices (Baker, 2018; Cairns, 2006; Nobes, 2015). Contemporary systems allow for impairment tests, which prudently enable the registering of accounting rubrics by the lower value between cost and market values, which is also compatible with a cost-based accounting system (Nobes, 2015). In addition, Donleavy (2019) makes little or no reference to liabilities. Nevertheless, many accounting research departments and standard setters demonstrate support for fair value accounting. Therefore, there is an actual possibility that accounting systems will move further in the direction of fair value accounting (Cardao-Pito and Ferreira, 2018b; Georgiou, 2018; Mora et al., 2019; Rayman, 2007).
However, Donleavy (2019) seems to misperceive fair value accounting as being a methodology to register accounting items by market values. He traces the origins of this idea to the French Physiocrat Anne Robert Jacques Turgot (1727–1781), the Irish-French banker and political economist Richard Cantillon (1680–1734), and also the Scottish political economist and moral philosopher Adam Smith (1723–1790). Nevertheless, it is not clear whether the standards and regulations cited by Donleavy (2019: 253) 2 convey the same understanding of market transactions as the one implied by writers centuries ago.
In fact, Donleavy (2019: 254) restricts the concept of fair value accounting to its first level, namely, the market price of the item (mark-to-market). However, fair value accounting has two other levels, namely, the market price of a comparable (proxy) element if Level 1 is not available (Level 2, mark-to-another-market), or a model to predict the market price if Levels 1 and 2 do not exist (Level 3, mark-to-model). The three levels all have to be considered, because the third level is also a theoretical explanation for how market prices occur at Levels 1 and 2. Fair value regulations assume that in order to achieve certain prices, markets are efficient processors of information regarding discounted forecasts of the eventual future economic income. Accordingly, market prices predicted in Levels 1 and 2 could be considered to be rational expectations of uncertain forecasts, just as in the case of those of the models predicted in Level 3. Consequently, to register accounting items by their market prices would be preferable to register them by their historical costs (Cardao-Pito and Ferreira, 2018a, 2018b).
Cardao-Pito and Ferreira (2018a, 2018b) do not say that the third level is the true fair value accounting. However, what they do say is something more sensible – namely that the third level contains a theoretical explanations of how accounting standard setters (and Irving Fisher) describe the formation of market prices and values. Cardao-Pito and Ferreira (2018a, 2018b) make this inference on the grounds that when no market exists, regulators allow the use of discounted cash flow models (the third level) as an approximation of how market prices would be formed. Accordingly, Cardao-Pito and Ferreira (2018a, 2018b) are able to conclude that accounting standard setters assume that discounted cash flow models explain market price formation. Thus, when certain items are registered by market values by means of fair value accounting, these are registered according to the supposition that market agents would establish prices by calculating projections of future cash flows discounted by a rate to proxy for risk and uncertainty, among other factors. In other words, Level 3 applies to the cases where no market exists, on the assumption that if a market was to exist, it would operate as a large discounted future cash flow model processor.
Unfortunately, the integral definition of fair value raises two impediments with regard to supporting Donleavy’s (2019) conclusions. The first setback is that the theories of prices and value formulated by the Physiocrats, Cantillon and Smith, are not consistent with the theory of value which can currently be found in the third level of fair value accounting. Therefore, fair value accounting cannot result, at least directly, from Turgot, Cantillon and Smith. The second setback is that although there have been other attempts to guide accounting into reporting market values in accounting, these are characterized by motivations which are quite distinct from the theoretical motivations which exist in contemporary fair value standards and norms.
Turgot, Cantillon and Smith’s theories regarding prices/value do not match fair value accounting’s theory
The eighteen century writers cited by Donleavy (2019) lived in an era which was very different from ours. They could not have had any knowledge about contemporary market operations. Turgot and Cantillon did not live long enough to see the French Revolution in 1789 and Smith died the following year. The Industrial Revolution had barely commenced. Nevertheless, fair value accounting and its proponents claim that market produce prices/values which are “correct” are, thus, “fair.” The expression “prices/values” is referred to because it contains an important feature in fair value accounting, namely considering that prices would be equivalent to “fair” values. Undoubtedly, most Physiocrats, such as Cantillon and Smith, did not align with this view of things. They agreed that, in the short-term, prices/values can be impacted by market pressures. However, they also argued that in the long-term, prices tend to their intrinsic values, which are based on the costs of production and must be found outside market transactions (Backhouse, 2002; Brue, 2000; Dupont, 2017; Ekelund and Hébert, 1997; Gailbraith, 1991; Screpanti and Zamagni, 2005; Skinner, 1999).
Generally, the Physiocrats were great land owners, who thought that land was the major source of wealth and value (Gailbraith, 1991). During this period, the economy was mainly agrarian. François Quesnay (1694–1774), who was a founder of this group, argued that although value is realized in exchange, it is not determined by that exchange, but rather by the costs of production (Prix Fondamental) (Dupont, 2017; Screpanti and Zamagni, 2005; Spiegel, 1991). Indeed, not all Physiocrats adhered to the same theory of prices/values. Turgot’s can be said to be more close to a subjective value theory, where prices depend on subjective consumer preferences, utility and scarcity (Backhouse, 2002; Brue, 2000; Dupont, 2017; Ekelund and Hébert, 1997; Screpanti and Zamagni, 2005; Turgot, 2011[1769], 2011[1753–1754]). However, there is no evidence that Turgot claimed that market prices are equivalent to rational expectations of discounted cash flows forecasts, as is contended in fair value accounting. In fact, while fair value accounting claims that market prices provide “fair values,” Turgot (2011[1769]) advised against confusing the concepts of price and value: “In the language of commerce, price and value are often confounded without inconvenience, because the enunciation of price always contains the enunciation of value. They are, however, different concepts that must be distinguished.” 3 Cantillon, who was a banker for part of his life, anticipated Smith’s labor theory of value (Backhouse, 2002; Brue, 2000; Cantillon, 2015 [1755]; Dupont, 2017; Ekelund and Hébert, 1997; Gailbraith, 1991; Screpanti and Zamagni, 2005; Skinner, 1999). He proposed that although market prices can be determined by supply and demand in the short-term, in the long-term they tend to their intrinsic value, which are based on land and labor which is necessary for production (Backhouse, 2002; Brue, 2000; Dupont, 2017; Ekelund and Hébert, 1997; Gailbraith, 1991; Screpanti and Zamagni, 2005).
Smith (1776, 1999: 131) made a distinction between “value in use,” and “value in exchange.” He proposed that market values can fluctuate temporarily in the short-term, due to exchange pressures from supply and effective demand. However, in the long-term, value gravitates toward the natural values of wages, profits and rents. These expenditures arise from the costs of elements necessary for production, namely, work, capital, and land. Nevertheless, labor would be the real value of the different components of prices (wages, profits, and rents):
real value of all different component parts of price, it must be observed, is measured by the quantity of labour which, they can, each of them, purchase or command. Labour measures the value not only of that part of price which resolves itself into labour, but of that which resolves itself into rent, and of that which resolves into profit. (Smith, 1776, 1999: 153)
4
Indeed, although David Ricardo and Karl Marx are currently better known for having advocated labor theories of value, Cantillon and Smith also promoted labor (and cost)-based theories of value.
As shown earlier, Turgot, Cantillon and Smith’s theories regarding prices and values are not consistent with the theory of prices/value which is implicit in fair value accounting. Accordingly, on inspection, the evidence fails to support the contention that fair value accounting is a direct result of Turgot, Cantillon and Smith’s writings.
Accounting based on market value that differs from fair value accounting
Another shortcoming in Donleavy’s (2019) study is to assume that accounting based on market values always has the same motivation as fair value accounting. The purpose of this section is not to observe all the historical situations where advocates have presented the case for accounting by market values, but rather, this section aims to demonstrate that the underlying motivations for market-based accounting could be different from those of fair value accounting. Indeed, it is an exaggeration to say that there was a “reign of historical costs for virtually the whole of the nineteenth and twentieth centuries” (Donleavy, 2019: 9; Mumford, 2000). The history of accounting is much more nuanced, and has allowed for certain rubrics to be registered at market value at least since the nineteen century. Furthermore, in many cases, these practices were compatible with historical cost accounting systems.
The debate regarding whether to register accounting items by market values or historical costs is not new. Disputes in France and Germany already existed between 1673–1800 among the advocates of market value-based accounting, who were preferred by creditors and their spokespersons (lawyers), and the proponents of cost-based accounting, who were more aligned with entrepreneurial capitalists (Ding et al., 2008; Georgiou and Jack, 2011; Richard, 2004, 2005, 2012; Zeff, 2013). The period (1800–1890) was characterized as being the static stage of accounting in France, Germany, and to a certain extent in the United Kingdom, due to the fact that accounting items were registered at market values (Ding et al., 2008; Georgiou and Jack, 2011; Richard, 2004, 2005, 2012; Zeff, 2013).
Between 1800–1850, pure market values were advocated for all items. In the context of the fast development of limited joint-stock companies between 1850 -1890, lawyers were obliged to adapt their accounting market-based theory and to adopt a prudent variant, which resulted in a “lower (value of) cost or market” (LCM) static approach. This is distinct from the pure cost-based accounting, where all accounting items are registered by their cost (Ding et al., 2008; Georgiou and Jack, 2011; Richard, 2004, 2005, 2012; Zeff, 2013).
To implement the static stage of accounting, it was not necessary to concur with fair value accounting’s belief that market values are correct and adequate. For during this period, creditors and the lawyers who represented them, won the debate regarding which accounting practices should be followed. They preferred market-based accounting, because it limits the amount of money that can be paid as dividends to owners/shareholders. Therefore, greater financial resources could be retained in the firm, increasing the likelihood of being able to service firms’ debts. Entrepreneurs and investors, however, preferred an accounting system based on costs (“dynamic accounting”), because this enhanced the possibility of increasing firms’ net income and dividend payments (Ding et al., 2008; Georgiou and Jack, 2011; Richard, 2004, 2005, 2012; Zeff, 2013). During the dynamic stage (1890–2000), prudent accounting continued to be used. However, the underlying assumption was no longer maintaining the firm’s liquidation, but rather continuing the ongoing concern (dynamics) of operations. Accounting carried on assuming that assets have a finite life. However, during the period 1890–2000 the amortization of long-term assets over a long period was permitted, which, in turn, enables the possibility of investors/shareholders being paid dividends much sooner, as profitability during the earlier years of an investment is certain to increase (Ding et al., 2008; Georgiou and Jack, 2011; Richard, 2004, 2005, 2012; Zeff, 2013). It can, thus, be observed that it is possible to have accounting systems based on historical costs which are compatible with the registering of certain accounting rubrics by their market values when these are lower than the net value of costs (prudent accounting). Donleavy (2019) appears to have a somewhat underdeveloped perspective when following Mumford (2000), he claims the existence of a “reign of historical costs for virtually the whole of the nineteenth and twentieth centuries” (Donleavy, 2019: 261). The history of accounting is much more nuanced, as can be seen from the three stages described earlier, whereby rubrics within historical cost accounting systems are permitted to be registered by their market values (APB Opinion 4; Clarke, 1980; Ding et al., 2008; Georgiou and Jack, 2011; Richard, 2004, 2005, 2012; Tweedie and Whittington, 1984; Zeff, 2013).
Even though the adoption of accounting standards was far from straightforward, the United States was a different case. For before the stock market crash of 1929 and the onset of the great economic and financial depression, there was clear opposition in the United States to the use of normalized formats and regulations for financial reporting, which would be applicable to the generality of organizations. Such norms were only imposed in the 1930s in response to not only the crisis, but also to the existence of various questionable accounting practices, abuses and scandals 5 (Berle and Means, 1932; Clarke, 1982; Ripley, 1927; Tweedie and Whittington, 1984; Walker, 1992; Walker, 1976; Zeff, 1979). Before the crisis, most organizations were characterized by heterogeneous self-reporting and self-ruling views about financial information, which brought about a large disparity of financial reporting models (Cardao-Pito and Ferreira, 2018a; Clarke, 1980; Georgiou and Jack, 2011; Walker, 1976; Zeff, 2007). After the post-1929 crisis, the United States adopted an accounting system based on historical costs, with the support from the recently formed Securities Exchange Commission (an institution which was created to address the crisis and to prevent further crises). This was supported by SEC Acts, the executive committee of the American Accounting Association and several academics(Clarke, 1980; Markarian, 2014; Paton and Littleton, 1940; Tweedie and Whittington, 1984; Walker, 1992; Zeff, 1979). The historical cost standards were imposed in the 1930s, not only in response to the crisis, but also to mitigate various questionable accounting practices, abuses and scandals (Berle and Means, 1932; Ripley, 1927; Tweedie and Whittington, 1984; Walker, 1992; Zeff, 1979). Indeed, the practice of the upward revaluation of assets tended to be eliminated 10 years after the Great Depression of 1929, as suggested by Donleavy (2019) and Swieringa (2011). However, exceptions can still be found (Walker, 1976). Furthermore, as explained earlier, assets can be revalued downward using market values within historical cost accounting systems.
There is another example where market-based accounting was proposed without the theoretical motivation present in fair value accounting. Chambers suggested a continuous contemporary accounting (CoCoA) theory (Al-Hogail and Previts, 2001; Chambers, 1965, 1966; Clarke et al., 2018; Wolnizer, 1999), which operates by reporting asset prices at their selling/exit value. The updated reporting of asset values would be made every time a new financial statement is produced. According to this theory, readers of financial statements are informed at each point about the organization’s capacity to generate liquidity. Chambers even proposed the necessity to register as an expense every item that fails to have a market where it could be sold, stating that this applies to not just physical assets or merchandise, but also to goodwill and various intangible assets which are currently described by most accounting standards as being assets. Thus, CoCoA presents the case for accounting at market values that is of interest to creditors, as in the static stage of accounting (1800–1890). Chambers has argued that historical cost accounting has had a specific historical path, which he did not see as inevitable (e.g. Chambers, 1971, 1994).
However, Chambers was against adopting fair value accounting methodology for explaining market prices which are found at fair value Level 3. According to Chambers (1965, 1966, 1968), such a methodology confuses measurement issues with valuation issues. CoCoA was the target of various criticisms that are described in Al-Hogail and Previts (2001) and Clarke et al. (2018). Nevertheless, it is very clear that even though he was an opponent of the theoretical explanation for prices/values proposed by fair value, Chambers proposed reporting accounting items by market values. These two examples thus demonstrate that Donleavy (2019) cannot be entirely correct in simply defining fair value as market-based accounting.
Financial economic theory and fair value accounting
During the second part of his paper, Donleavy (2019: 260–264) addresses what he perceives to be a puzzle: That the theory that supports fair value accounting was formulated in the eighteen century, and “the last quarter of the Twentieth century” was the period which saw any real action by accounting standard setters to implement fair value in accounting. Once again, Donleavy assumes that Turgot, Cantillon and Smith’s theories support fair value accounting, which this article argues is not the case. However, if this puzzle is artificial, then another important puzzle remains: Where does fair value accounting result from?
In general, I agree with many of the points raised by Donleavy (and those researchers cited by him) in the second part of his paper. I concur that institutional and political factors have been quite relevant to the implementation of fair value accounting (Donleavy, 2019: 262–264) and I have the same opinion that the empirical support for the theory underlying fair value accounting is rather weak, if not altogether inexistent (Donleavy, 2019: 264). Furthermore, Donleavy (2019: 263) may well be correct when he notes that fair value accounting is somewhat linked to mainstream financial economic theory in accounting research and with the standard setters.
As mentioned earlier in the “Introduction” section, Cardao-Pito and Ferreira (2018a, 2018b) exhibited that the fair value standards mentioned by Donleavy (2019) appear to be aligned with Irving Fisher’s theoretical writings regarding economic and accounting value, and market prices. The content analysis of these authors identifies a literal correspondence of Fisher’s writings to key fair value norms from the IASB and FASB. For instance, these norms mimic Fisher’s claim that cash flow forecasts and discount rates could explain market values/prices. Furthermore, Cardao-Pito (2016, 2017a, 2017b, forthcoming) shows that the seminal text in financial economic theory, namely that of Modigliani and Miller (1958), and financial economic textbooks are also fundamentally aligned with Irving Fisher’s writings.
Nevertheless, these findings need to be further tested and confirmed in future research. For most research on accounting history omit Fisher, or only mention him in passing, which raises the question whether Fisher influenced accounting theory and practice in a long-lasting manner (Baker, 2018; Oldroyd et al., 2015). If confirmed, these findings could provide an interesting, under-researched explanation, and, in addition, they would also connect accounting to the history of economic thought (Bryer, 2013; Cardao-Pito and Ferreira, 2018a, 2018b; Markarian, 2018).
Conclusion
Fair value advocates argue that market values would be superior to historical costs for reporting accounting items. The underlying theory claims that markets would continuously process information regarding discounted forecasts of future cash flows. Thus, market prices/values would be correct, that is to say, they would be “fair.” However, Cantillon and Smith believed that while supply and effectual demand can temporarily impact market prices/values, in the long-term, these prices would tend to maintain their intrinsic value, which is based on the costs of production (where labor is a key element). Turgot, who was more inclined to support that markets are able to attain subjective prices, was never a proponent of the theory that markets compute prices according to the rational expectations of discounted cash flow forecasts. In fact, Turgot advised against confusing the concept of price with the concept of value. Turgot, Cantillon and Smith’s theories of prices/values are, thus, directly opposed to the theoretical explanation presented by fair value accounting.
Furthermore, in previous historical periods, the case was made to register accounting elements by their market values, while invoking different motivations to those of fair value advocates. The argument was that producing accounting statements at liquidation values would better inform creditors and ensure payment of their credit. In this manner, owners/shareholders find it more difficult to use dividend payments as a means to transfer wealth from the organization to themselves. In this sense, a system based on historical costs is compatible, for example, with registering certain asset rubrics by market values, when the market value is lower than the net value of costs. The case for accounting by market values can thus be put forward without demonstrating the motivations presented by many contemporary accounting scholars, regulators, and standard setters. Fair value accounting refers to a specific period in the history of accounting, which should not be confused with the reporting of accounting items by their market values.
Donleavy (2019), however, appreciates the importance of the financial economic theory in underlying fair value accounting, its empirical weakness, and also how the financial economic theory has replaced empirical observation in defining accounting standards. To confuse fair value accounting with the reporting of accounting rubrics by their market values, is to second the claim of the advocates of fair value accounting theory that we fully understand how market prices/values are formed. In future research, the history of economic and accounting thought may indeed help explaining this strange belief in contemporary accounting and economics.
Footnotes
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Joao Silva Ferreira, Julia Smith, Patrick McColgan, Andrew Marshal, Christine Cooper, the ADVANCE Research Center at ISEG, and Portuguese national funding agency for science, research, and technology (FCT) under the Project UID/SOC/04521/2019.
