Constructing the Fair Value of Non-Financial Assets -A Case Study

1 Constructing the Fair Value of Non-Financial Assets – A Case Study Ekaterina Chetyrkina [email protected] João Oliveira [email protected] Maria do Céu Ribeiro [email protected] Abstract The fair value phenomenon has been transformed into a “quasi-philosophical principle at the center of an accounting reform” (Power, 2010, p. 197). However, while there is much quantitative research on fair value, there is scarce research about how this complex and controversial concept is operationalized, within particular contexts of actual organizations. We analyse the fair value phenomenon through a qualitative approach, by researching an actual process of determining the fair value of non-financial assets at a large Portuguese industrial company. Inspired by Actor-Network Theory (Callon, 1986 and Latour, 1997), the study shows that the fair value recorded in the financial statements is not simply calculated, but is “socially constructed”, based on the value perceptions resulting from an agreement reached during interactions within a complex network of human and non-human actors. These actors, of varying centrality, influence directly or indirectly not only the process of determining the fair value, but also the behaviors of every participant of this complex network. Two different concepts of fair value were being applied, not clearly discernible in financial statements and constructed through two different processes, involving a different set of network actors. By showing that understanding accounting figures requires understanding accounting practices, unfolding through particular organizational arrangements, this study highlights the social nature of accounting figures. 2 1. Introduction Over time there have been many discussions about the basis of valuation of balance sheet items and the cost method was the initially preferred in different jurisdictions. However, the use of market values has been growing in the last decades, in particular with the international expansion of IAS / IFRS. These standards are based on an accounting model oriented for the investor and, as such, are based on the fair value and the extensive use of professional judgment by preparers of financial information (Gwilliam, 2008; Hellmann, 2010). These multiple professional judgments are performed in a complex social context, involving a network of different actors that interact in multiple ways, in multiple locations and across multiple tools. So there is a complex social reality underlying the fair value reported in financial statements. Using the lenses of Actor-Network Theory (Callon, 1986), this paper analyzes, through a case study, the multiple social relations in the process of determining the fair value of non-financial assets in a large Portuguese industrial company and, in particular, the interactions of the different agents and how the fair value is based on a consensus that results from a complex and non-deterministic process. The emergence and spread in international accounting regulation of the fair value concept have fueled many discussions about its relevance and reliability, particularly when compared to historical cost. One debated issue is the possibility that fair value measurement has contributed to the aggravation of the 2008 global financial crisis (Laux and Leuz, 2009), given the fair value inherent subjectivity, especially in the absence of an active market 1 , which contrasts the objective measurement by historical cost. Financial reporting aims to provide useful information for decision-making by various users. However, subjectivity and the consequent reduced reliability of fair value, as well as its quantification complexity, call into question its usefulness and therefore represent obstacles to that purpose. Furthermore, empirical evidence has not provided decisive arguments in favor of either historical cost or fair value (e.g. Nissim and Penman, 2008; Laux and Leuz, 2009; Herrmann et al., 2006). 1 When there are active markets, this method becomes as objective as the cost method. 3 There are several studies on the value relevance 2 and on the disclosures of historical cost and fair value in quantitative terms (Barth et al., 2001; Song et al., 2010), but there is little information about the establishment of the fair value’s theoretical concept and the very process of its determination. International accounting regulation is characterized by the complexity of some standards, particularly on fair value, since the regulation is based on principles 3 , not on specific rules. Power (2010) argues that when it comes to quantifying the value, fair value reliability is related to the construction of different perspectives of values by different actors involved in the process. In this case, accounting regulation only establishes parameters to facilitate consensus among actors, but it does not entirely eliminate subjectivity. In this perspective, the items measured at fair value with no active market are shown in the financial statements as an end result of a consensus between different agents, not providing any information about the difficulties of its measurement. During this interactive process, conflicts between actors until they arrive at fair value can be expected (Power, 2010). However, there are hardly any empirical studies on fair value that analyze the specific contexts and organizational processes in which reported calculations are developed and operationalized (Hopwood, 2009). This study fills this gap by allowing to understand how the concept of fair value, abstract in nature (Power, 2010), is operationalized in terms of calculation in the accounting practices of particular organizations (Hopwood, 2009). The case study of a process of determining the fair value in a large Portuguese industrial company reveals, contrary to initial expectations, several procedures for determining fair value, involving networks of different actors, with direct or indirect influence in the process. It is interesting to underline that no actor “calculates" 4 , but rather interacts in the fair value’s 2 These studies use shares value as a benchmark of the company’s value, to analyze to what extent the reported financial information is useful to an investor and is reflected, by him, in the company valuation, i.e., if there is a significant association between this variable (in this case, the fair value or disclosure of the respective hierarchy) and the price (Barth et al., 2001). 3 According to the Principles-Based Approach, first is defined a conceptual basis within the regulated subject, and then there are provided guidelines to explain their goal. In this approach the principles are designed to serve the public interest and are accompanied by a small number of rules that show how these principles should be applied in specific situations. It does not seek to regulate all possible situations, such as in the Rules-Based Approach, because in case of doubt the principle in question is applied. 4 The word "calculate" is here used in a narrow, mathematical sense. 4 co-production process (Callon and Muniesa, 2003). Hence, the fair value should be understood as a "socially constructed" figure based on the perceptions from key actors on the process. The interactions between the key actors are based on the model called "fair value triangle", whose vertices correspond to the preparer of financial information, the appraiser and the auditor. The case study allowed to enrich this “fair value triangle” model, by incorporating the multiple actors that actually constitute each of these vertices, linked by formal and informal relationships inside and outside the entity. Together, these actors form networks whose configuration changes for the construction of the different concepts of fair value. After this initial section, section 2 introduces the fair value’s theoretical framework that underlies this study. Section 3 discusses the problem of calculating fair value and the process of its social construction. Section 4 deals with the qualitative research methodology used in this study, and section 5 presents the case study. The main reflections, conclusions, contributions and limitations are presented in Section 6. 2. Theoretical Framework on Fair Value Fair value has been a hot topic among professionals and academics. Inherent to the measurement of an asset and liability is one of the biggest dilemmas in accounting: reliability versus relevance (Whittington, 2008). Investigation regarding fair value has focused mainly on studies of value relevance (Barth et al., 2001; Beattie, 2005; Basu, 2012; Okamoto, 2014), not capturing the emerging economic reality within the "imaginary" market created by fair value (Bougen and Young, 2012). To relate the measurement of fair value and its imaginative features, a definition of fair value is required, in particular since, unlike historical cost 5 , it is a non-observable value. Fair Value - Development and Controversy 5 Some arguments for and against the fair value can be found in Nissim and Penman (2008), but no evidence has objectively supported the superiority of fair value over historical cost to value assets/liabilities (Okamoto, 2014). 5 Along with a tendency to lower the importance attributed to verifiability in favor of relevance (Barth et al., 2001; Whittington, 2008), fair value expanded regarding financial instruments, particularly derivatives (Power, 2010), and it has been extended by analogy to other assets and liabilities through financial evaluation methods and models. The concept of fair value is currently defined in IFRS 13 as "the price that would be received in selling an asset or should be paid for the transfer of a liability in a current transaction between market participants at the measurement date". Certain market-oriented assumptions are used, in particular the highest and best use of the asset, the asset price, as well as inputs for their measurement, classified as observable or non-observable. This categorization of inputs is related to the company perception on the assumptions that many users of financial information would take into account in determining a price. Thus, they are assumptions of assumptions, which imply that assessments are essentially "simulacra" (Bougen and Young, 2012). Clearly, the most observable and independent inputs are quotations in active markets for identical assets, and the least independent are the non-observable inputs, that Bougen and Young (2012) call the "simulacrum" 6 of assumptions that the market would have, if it existed. In other words, the so-called imaginary markets are created, in which the assets and liabilities are valued in the context of a new reality of economic relations (Bougen and Young, 2012). In those cases where the fair value of an asset or liability may be determined by the market, the price of this asset results from a set of large volume transactions between buyers and sellers 7 . There is a trade-off between relevance and reliability of financial reporting when it comes to the different levels of the valuation inputs (Barth et al., 2001; Whittington, 2008). Relevance does not decrease as markets become less liquid (Song et al., 2010; Christensen and Nikolaev, 2013); however, reliability is reduced as the levels increase, thereby making level 3 inputs less reliable (Song et al, 2010;. Bagna et al., 2014; Chung et al., 2014). The existence of less or non-observable inputs in the market requires the simulation of the asset’s features 6 Bouguen and Young (2012) refer to the purpose of FAS (Financial Accounting Standards) 157, in the US regulation, which is similar to IFRS 13. 7 Even in these circumstances, there will be subjectivity in the process, since there is no guarantee that this asset / liability will be exchanged in the future at this price. 6 to obtain a comparable asset, requiring the creation of another economic reality (Bougen and Young, 2008). This challenges one of the fair value claimed advantages, that by incorporating more information, it is more relevant for decisions (Barth et al., 2001; RayBall, 2009), Two concepts: Fair Value and "Fair Value" There is currently a mixed accounting system in which the fair value and historical cost coexist (Okamoto, 2014). This system involves estimates, and market values may be used to support these estimates, not as an application of the principle of accounting for the fair value, but only as an alternative way to support these estimates. As already mentioned, the fair value is not a synonym of the real market values, but rather of the market price estimates (Power, 2010). In practice, excluding financial assets and liabilities, the concept of fair value is applicable to the measurement of assets acquired in a business (IFRS 3) and to the subsequent measurement of non-financial assets (IAS 16 and IAS 38), if the preparer of financial information opts for fair value. In turn, non-financial assets impairment tests incorporate fair value as a primary determinant of the decision to impair such assets (IAS 36), albeit in a perspective of historical cost. So, there are actually two concepts of fair value: the first, with direct accounting impact; the second, only in case of impairment. Throughout this study, we chose to call "fair value" (fair value, between quotes) the fair value applied to the last situation (for impairment decision), and fair value (without quotes) the one that applies to the first situation. For a better understanding, we present graphically in Appendix 1 the similarities and differences between the two concepts, as well as the connection with the other levels of fair value. According to the economic theory of perfect market, the market value of an asset is equal to the income that it will generate. However, the asset profitability depends on its specificity to the purchaser, i.e., its use value (Bignon et al., 2009). In a perspective of historical cost, in accordance with IAS 36, the value to be considered for impairment decision 7 is the biggest one between the market value (fair value of Level 1, 2 or 3 8 less the costs of the sell) and the value in use, i.e., an estimate of discounted future cash flows that are specific to the organization (firm-specific values), considering possible synergies of that asset. Value in use is tagged with a fair value level 3 where the cash flows are simultaneously firmspecific and market-specific values. Paradigm of Fair Value for the Appraiser a) Convergence of Different Valuation Techniques and Fair Value Concepts IFRS 13 establishes fair value on the basis of exit price, in particular with regard to the reliability of estimates of that price that are based on valuation techniques (Ryan, 2008) and its suitability for the assets in use or held to maturity (Whittington, 2008). If there is no market information or comparable transactions in the market, this exit price should be estimated through various valuation techniques. The valuation techniques to measure fair value should be appropriate to the circumstances, and there must be sufficient data available to ensure its correct application. The IASB (2011) defines three techniques:  market approach (“uses prices and other relevant information generated in a set of transactions in the market of assets and liabilities identical or comparable” - exit price);  cost approach (also called "current replacement cost" – “reflects the amount that would be currently required to replace the service’s capacity of an asset” - entry price that in perfect markets would be equal to exit price);  income approach (consists in discounting the future cash flows generated by the asset to the present moment; in this case the fair value measurement reflects current market expectations about those future cash flows - neither exit price, nor entry price). 8 Although its fair value, the disclosure of the respective levels is not, however, required (IFRS 13). 8 IFRS 13 does not prescribe what kind of technique should be used in specific circumstances, unlike the IVS (International Valuation Standards), standards generally accepted internationally for the professional valuation practice. The IVS define three bases of value: market value, investment value and fair value. IFRS 13 fair value is generally consistent with market value as both concepts involve an hypothetical transaction in the market. The IVS define the following measurement techniques to determine market value: replacement cost approach (depreciated or not), or (depreciated) replacement cost method (reflects the current cost of replacement of the asset or depreciated); income approach or discounted cash flow method (the estimate of the present value of the asset consists in the update of the future cash flows); comparative market approach, or sales comparison method (reflects the price of the similar asset tradable in the market). The IVS framework does not define all bases of value that may be required for all situations around the world. However, as IAS/IFRS are widely used, there was a clear convergence of concepts between both frameworks through the specific basis of value, that is market value. There was, thus, a change in the fair value’s concept with IFRS 13, but at the same time the exit and entry prices converge with the "market value" defined in the IVS framework. b) Fair value and the Appraiser's Role As stated above, there is no single concept of fair value, nor consensus among appraisers as the "best" fair value (Campbell, 2008). Thus, the preparer of financial information must identify the primary market for the asset or liability to be valued, as well as choose the most appropriate valuation approach and its consistent use, even in the case of delegating these tasks to an external appraiser. The decision of delegating the mentioned tasks to an appraiser is sometimes related to the greater confidence and reliability that fair value estimates determined by an independent appraiser transmit to the users of financial information (Cotter and Richardson, 2002). However, in practice, a greater use of appraisers is closely related to the auditor’s support (Campbell, 2008), as we shall see in section 3. 9 Fair Value Paradigm for the Auditor The emergency of the regulation on fair value led to a greater number of estimates and judgment in financial reporting, further hindering the audit work. Griffin (2014) shows empirically how the auditor makes decisions in the absence of reliable market information. In a fair value context, the estimation process involves a greater degree of uncertainty, given the greater inherent subjectivity of the inputs used in the calculation of these estimates and the imprecision of the respective results. It is also found that the less observable market inputs are, the greater the level of proposed adjustments and disclosures of assumptions that formed the basis for determining fair value. The main objective of the auditor is to express an opinion on the accuracy of financial reporting information, even in situations of fair value measurement, which raises problems due to the subjectivity and imprecision of the latter (Smith-Lacroix et al., 2012). The shift from the historical cost paradigm to the fair value altered the decision-making process in auditing, in particular with a lower significance of quantitative materiality in favor of qualitative materiality (Christensen et al., 2012; Griffin, 2014). The concept of materiality, "one of the anchors of audit” (Machado de Almeida, 2014, p. 185), provides a level from which an omission or misstatement could influence the decisions of users of financial reporting. This is a quantitative concept, but it is also influenced by several qualitative factors and is used in the validation of the financial report which is dependent of the auditor’s judgment. There is evidence of the non-existence of agreed limits for materiality, which is thus based on professional judgment and subjectivity. With the expansion of fair value, a subjective and imprecise value, there is a growth of the professional judgment and hence of the qualitative materiality. Because the auditor’s professional judgment is a mental process based on a psychological model made up largely by the auditor’s experience, stimulus and judgment process (Machado de Almeida, 2014), each auditor will have a different opinion on the fair value. On the other hand, the fact that accounting matters have undergone profound changes in terms of their knowledge base, especially in regards to the introduction and spread of the 16 company and covers more actors, in particular the information preparer, the auditor and the appraiser (and additional actors who were identified only during the study). A letter was sent to the company Chief Executive Officer (CEO), presenting the preliminary project and requesting a previous meeting, for a better explanation of the purpose of the study and to define the terms of the proposed collaboration. The collaboration was accepted, with the condition of maintaining the company’s anonymity. The authorization to contact the auditor (a "Big Four" company, auditor of the case company for several years) and appraiser of the company was obtained. The company received a full draft of paper, made comments and corrections and approved it for dissemination. Data Collection Techniques Semi-structured interviews, all carried out by the first author, were combined with document analysis to corroborate and orient interview information (Yin, 2009). A research protocol with the company was established to access to the information disclosed and undisclosed in the report and accounts. As mentioned, the first contact with the company was through a letter sent to the CEO, followed by a meeting with the Accounting Director to explain the study purpose, define the collaboration terms and obtain approval. The first interviews to the three actors, initially identified as being relevant to the determination of fair value (the preparer of financial information, the auditor and the appraiser) had a ripple effect (snowball sampling, Gil, 2007) leading to the identification of other actors to interview in person. Table I below summarizes the interviews. Table I – Interviews Conducted In-Person Entity Position Total Duration No. Interviews Preparer Accounting Director (covering Administrative and Consolidation areas) 2 hours 3 Preparer Consolidation Director 30 minutes 1 Preparer Strategic Planning Director 30 minutes 1 Preparer Technical Director 1 hour 1 17 Preparer Management Control Officer 30 minutes 1 Appraiser Managing Director – Fixed Assets 1,5 hour 2 Auditor Responsible Partner 30 minutes 1 The interviews, all in person, were not recorded, because of concerns that the respondents could feel inhibited due to the possible sensitivity of the subject. However, the content of each interview was summarized and sent to each interviewee to ensure accuracy, or even to allow them to add some information they deemed relevant; subsequently, there were telephone contacts to clarify any doubts. In addition to the data collected through the above interviews, we analysed the proposal to provide appraisal services by the independent appraiser and the evaluation reports of nonfinancial assets revalued in the last reporting period (land and buildings). For confidentiality reasons and information sensitivity, the first author not receive a copy of the evaluation reports, but had the opportunity to consult them extensively for 1 hour, taking notes as appropriate. 5 The case: the fair value Construction Process at “TecnoCorp” 5.1. Company description and the ex-ante conceptual model: the Fair Value Triangle “TecnoCorp” (fictitious name) is a large Portuguese industrial organization and an important player in its market, with production units both in Europe and beyond. Its products are relatively homogeneous, yet at the same time diversified due to numerous variants and customizations. The company has grown mostly through national and international acquisitions throughout the decades. However, the recent global economic crisis has had a negative impact on profitability, and forced the company to close units with worse performance in various countries and to implement various organizational changes. Various disclosures in TechnoCorp’s financial reports signal the usage of market values for various non-financial assets: land, buildings and goodwill related to assets acquired in business combinations. Financial reports also indicated that “Fair Value”, as defined in 18 Section 3, was also used as basis for impairment calculations. The remaining non-financial assets were recorded at historical cost. Disclosures also indicated the usage of an external appraiser. However, regardless of the accounting rule used, the user has no insight about what processes and procedures underlie the reported numbers. Based on the available fair value literature and on TechnoCorp’s financial reports, an ex-ante theoretical framework was developed, identifying three main (collective) actors involved in the fair value determination process: the company itself (the preparer), the auditor and the appraiser This conceptual model, here labeled as the “Fair Value Triangle", guided us when further exploring the case. This ex-ante model already conceptualized fair value as emerging from a process involving a network of actors. However, based on the literature indications about the complexity of determining fair value, the authors anticipated that the empirical study could potentially reveal complex processes and a more complex network. As reported in this chapter, interviews and further documentation analysis revealed two different processes underlying the two types of fair value. The process regarding Fair Value (without quotes) involves an independent appraiser (as expected in the initial ex-ante theoretical framework), but the network of actors within each of the collective actors, and the complexity of the process, turned out be far more complex than anticipated. On the contrary, the process regarding “Fair Value” was substantially simpler than anticipated, with the notable absence of the appraiser from the network and, overall, a less complex network and process; however, this less complex process was actually the one with greater discussions and power struggles between the two involved actors (preparer and auditor). Each process is now analysed separately, culminating in the shift from the single, ex-ante conceptual model presented above to two conceptual models, each applicable to the two concepts of fair value. 5.2. The Fair Value construction process The actors with central role in determining the valuation parameters 19 Before carrying out the land and buildings appraisals, the actors needed to reach consensus on three essential parameters (scope, premise and valuation techniques), on the base of which the appraiser issued the proposal for the valuation engagement. The appraiser is widely recognized in the marketplace and was already in charge of other appraisals at TechnoCorp. is shown below, the centrality and the decisive power of each of the main actors vary according to the parameter at stake. Choice of the valuation scope Firstly, the actors had to define the scope of the appraisal. The valuation of the company’s operating units adopted two different scopes, to be analyzed separately: In Loco Appraisal (in which the appraiser personally visits the units for issuing its opinion of value); and Desktop Appraisal (in which the appraiser does not visit the properties, being the opinion based on information provided by the company and on follow up of previous In Loco appraisals). The consulted proposals and valuation reports showed that the appraiser had carried out in loco appraisals of most industrial units (about 60%) for insurance coverage So, in year N, TechnoCorp decided to measure the fair value of those units under a desktop appraisal basis, using a in loco appraisal scope only for the remaining ones – conditional to the auditor’s agreement. Given some audit comfort obtained from recent external valuations, the auditor agreed that adopting the simpler desktop appraisal scope for these units was acceptable. The auditor thus assumed a central role regarding the adopted valuation scope. Desktop Appraisal On a desktop appraisal basis, the external appraiser draws upon the values determined in previous in loco appraisals and requests the company details about subsequent acquisitions, disposals and write-offs. As confirmed by the limitations stated in the valuation report and during the interview with the appraiser, under such scope, the appraiser did not take any responsibility on issues related to the assets physical condition, utility, level of use and their existence ether. The appraiser then simply adjusts the previous information assuming the land areas given by TechnoCorp and for the buildings considering a normal tear and wear based on the maintenance observed in previous valuations.. In turn, this limitation on the valuation report created another scope limitation on the audit work. This limitation was mitigated by 20 TechnoCorp’s technical team confirming that all lands and buildings under a desktop basis were in a normal condition and experienced no abnormal deterioration. This statement was provided directly to the auditor, helping this actor to support its opinion on the fair value measurement of the assets determined by the appraiser on a desktop basis. In order to mitigate such limitation, the auditor also confirmed with local audit teams that no structural deficiencies existed that might affect the useful remaining life of the appraised buildings. The technical team also contacted with TechnoCorp’s legal department to confirm the properties legal ownership through certificates from the Property Registry. In Loco Appraisal Conversely, in a inlLoco appraisal, the external appraiser physically visits the units, as it was the case for a few TechnCorp’s properties. Before the physical inspection,the appraiser requested to the company the accounting inventory of the assets to be appraised, blueprints, maintenance plans, land and building certificates to confirm ownership, and the buildings’ usage permits, typically issued by the local authority - without such permit, the asset value is either considered null, or the asset is valued as a warehouse. . As noted in the valuation reports, studies about the land area, location, building design, types of building material, number of floors, functionality of each area, physical depreciation, maintenance conditions, the building configuration and on the economic environment were performed by the appraiser. Choice of the valuation premise The second major valuation parameter was the choice of the valuation premise, continued use or not. Estimations of market values arrived at on each premise are quite different. TechnoCorp assumed a central role in this matter as the premise of an asset’s forced sale or continued use was based on TechnoCorp’s strategic decision for that asset at the valuation date. That premise was communicated to the two other actors, However, the appraiser was merely informed by the company about the intended purpose, without intervening in the decision. However, it had to be validated by the auditor through the minutes of the Board of Directors. Under the premise of forced sale, the valuation is always performed by in loco appraisal, while in the other premise both scopes are applicable. As checked in the valuation 21 reports, for most assets fair value measurement was based on the premise of the continued use. Choice of the valuation technique Thirdly and finally, once the valuation scope and premise have been settled, the appraiser choosed a valuation technique to derive an opinion of the market value of the assetAs noted in the valuation proposal and reports, among the valuation techniques (cost approach, market approach and income approach), the most appropriate one was chosen by the appraiser considering the type of asset at stake 10 . However, as confirmed during the interview with the auditor, those techniques were duly discussed and agreed with the auditor, in order to check its adequacy. Land and land improvements As generally appraised, land was valued by the market approach 11 derived through an industrial land market research. The appraiser clarified that each piece of land is classified in the official Territory Plan (the term varies across countries), which then determines the land usage – at TechnoCorp, typically, industrial usage is usually the case. Land value is determined by multiplying total area by square meter price. The total area was confirmed by the blueprint. Square meter price was determined by analyzing market transactions of land with similar areas and characteristics, namely whether construction is planned or not that affect the value. Any work done upon the land as well as infrastructures, such as roads, grounds, water systems, sanitation, and other installations under the ground, were considered land improvements, appraised separately from the land. Since there was no active market, such improvements were appraised by the cost approach.. The market value was derived from technical studies on the construction costs per square meter of of the related improvement less allowances for physical deterioration resulting from wear and tear. Such information is available in databases developed by the appraiser network from other projects, clients’ architecture projects (construction certification) and construction industry publications with 10 In assets under construction, biological assets and forest assets, the valuation was only performed for land. 11 The income approach was never applied at TechnoCorp. 22 construction costs indices – externally drawn inscriptions with a relevant role in the determination process. Buildings The valuation technique applied for most buildings items was the cost approach, since no sales prices judged to be comparable to TechnoCorp’s buildings were available in the market. However, for determining warehouses values the market approach was applied as they were considered a more standardized construction. Moreover, buildings were appraised depending on their usage (administrative or industrial, production or storage unit), discussed with the TechnoCorp’s technical department given their different construction costs. The appraiser drew upon several non-human actors: internal databases, external inscriptions, websites of real estate firms for an indication of comparable asset (with a similar area, year of construction and other parameters) and construction publications, as referred to above for land improvements valuations. In summary, different valuation techniques were used to derive fair value of TechnoCorp’s land and buildings. A market approach was used for land and more standardized constructions, with fair value being estimated as the average price of comparable transactions sales defined by the appraiser. Other buildings and land improvements fair value was derived from a cost approach, given that no sales comparison was possible.. The applicability of each technique is only discussed between the appraiser and the auditor. Unforeseen actors and connections in the Fair Value Triangle The process of estimating lands and buildings fair value thus emerged as more complex than the "triangle" initially anticipated. The ex-ante theoretical framework proved to be simplistic and lacking organizational granularity, given the unforeseen diversity of actors – typically, collective and individual actors within the collective actors initially identified – that influence and need to be considered to understand the process of constructing fair value. These actors may be more or less numerous, depending on whether Desktop or in loco appraisal is at stake, as shown in Figure 1. 23 Figure 1 - Key and Secondary Actors Network in the Fair Value Determination Process Source: Developed by the authors For simplicity, the Portuguese audit team, the Portuguese appraisal team and TechnoCorp’s fair value measurement team will be referred to as "key actors", the others being "secondary actors". The secondary actors had two major roles, as described below, to support key actors to achieve their goals and interests and to help them to mitigate any limitation.  Engineering/Maintenance Department – in both in loco and desktop appraisal, it helps the appraiser to better understand the characteristics of the asset, reducing information asymmetry between the appraiser and the company. TechnoCorp’s employees belonging to this department also assist the appraisal team in verifying the land areas (in in loco appraisal) and in other technical aspects that may help the appraiser in determining the asset’s market value.  Technical team –it stated to the auditor the maintenance conditions of the buildings and other constructions not physically inspected by the appraiser, as well as the, non-existence of any abnormal deterioration. This was only applicable for items appraised on desktop basis and specifically for the actor, auditor, as highlighted by the Technical Director. The 24 statement issued by the TechnoCorp’s technical team was discussed and reviewed with its Technical Director. The technical team only participates in the process regarding buildings and equipment, since depreciation allowances for physical deterioration are only applicable to those assets and not to land. The technical team based its statement on physical verifications of the assets by plant employees, in order to confirm before the type of construction (simple / complex), the building areas and configurations depicted in the maps and drawings, and finally the deterioration level.  Unit’s director – in Desktop Appraisal, it confirms to the technical team the maintenance status of the unit’s buildings; and in both in loco and Desktop Appraisal, it ensures that the assets inventory is reconciled with what physically exists, thus a guidance with the assets to be appraised for the appraisal and linked to the financial reporting for the auditor  Accounting Department – in both in loco and desktop appraisals, it provides the appraiser with the accounting inventory duly reconciled by the unit’s director and for desktop appraisals information about acquisitions and disposals / write-offs occurred since the last in loco appraisal.  Legal Department – provides the appraiser (in in loco Appraisal) and the technical team (in desktop appraisal) land official records to confirm ownership.  Local audit teams – in desktop appraisal, they confirm to the Portuguese audit team the normal wear of buildings, as stated by the Technical team.  Local appraisal teams – in both in loco and desktop appraisal, they help the appraiser and the company to tackle the time and cost problems due to geographical distance. The appraiser has local valuation teams to visit and determine each operating unit fair value, but that are always coordinated by the same responsible in Portugal to ensure knowledge accumulation on valuations carried out. .The field work was carried out by these local teams, usually two or three people accompanied by TechnoCorp’s plant employees, typically belonging to the area of maintenance and engineering. In in loco appraisals, plan drawing is compared with the one available in Google Maps, another nonhuman actor. These teams also tested randomly measurements to scale at some sites in order to ensure the accuracy of the plants scales. 25 As described above, it should be noted that the network of actors has very complex paths and connections and hybrid actants: the same secondary actor can simultaneously help more than one main actor, and can help not only to achieve the objectives, but also to reduce the limitations. So those actors might not be conceived as fixed actants in the network. Interests, constraints and solutions of each main actor The main actors have a common goal (fair value calculation), but each has its own motivations and a set of constraints that might hamper the achievement of that goal. Table 2 reports the details of such interests, constraints and limitations for each key actor, the preparer, the auditor and the appraiser. As the other secondary actors depend upon the key actors, we have not displayed their interest and constraints. Table 2 – Key Actors’ Interests, Constraints and Solutions in the Fair Value Determination Process Company Interests  Preparation of financial information in accordance with accounting principles at reporting date;  Determination of fair value of lands and buildings, assets acquired in a business concentration Constraints and solutions  Existence of IAS / IFRS accounting rules  Cost associated with physical visits to various factories  Independence of the external appraiser to be ensured Auditor Interests  Validate adequacy of the technique, premise and scope approach and the value determined by the appraiser, particularly when the appraiser does not assume responsibility on issues related with physical conditions and maintenance.  Ensure that the company's financial information is true and presents a fair view of its financial position wit accounting principles. Constraints and solutions  Many auditors consider the audit as a commodity, with partners having objectives to meet, customers to serve, proposals to do, businesses more and more lucrative. The auditors behavior is based on the balance between the marginal cost and marginal benefit, i.e. in reducing the likelihood of an unfavorable litigation outcome against the auditor (Machado de Almeida, 2014)  Commercial professional and legal constrictions of the 32 Fair value is determined through an external appraiser – independent and recognized in the market and supporting its work on international valuation regulation (associated with level 2). By contrast, for "fair value" the discounted cash-flows model, developed by the company, is applied, and includes a set of data, projections and estimates (associated with level 3). Therefore, the actors’ network determining fair value (without quotes) is much more diverse than for "fair value". The existence of three main actors in the first case, involving various external entities and secondary actors in the process, creates a set of complex links between them, implying that the behavioral models of all actors suffer transformations and adjustments to the interactions and communications processes. This complexity is in part a consequence of the purpose and potentially greater impact of fair value, since it will be recorded in the accounts regardless of being higher or lower than the historical cost. On the other hand, determining "fair value" is a more complex, discussion-prone process, where conflicting interests and motivations between the preparer and the auditor become more visible. Finally, for "fair value”, the external appraiser does not even participate in the process, while for fair value it is central for several aspects, including determining the evaluation approach. Regarding the power of the auditor, he has a central role in determining the assessment scope in fair value and identifying signs of impairment in "fair value". Finally, the preparer has discretionary power in determining the evaluation’s perspective, budgeting and the business plan. Figure 3 depicts the two process, with red arrows representing the relations regarding fair value and blue arrows representing the relations regarding “fair value”; two actors which only intervene in the “fair value” process are also depicted in red. In general, there is an essential similarity between the two processes, since consensus among all parties on the reported value always has to be achieved. Figure 3 –The two processes for determining fair value and “fair value” 33 Source: Developed by the authors 5.5 Theoretical Development: Connections and Interaction in the Process of Fair Value Since the economic calculation is distributed between human and non-human actors, hence with several possible measurements (Callon and Muniesa, 2003), and since fair value has an economic orientation given its market basis, then fair value will inevitably be inaccurate in its quantification – because no “absolute” accuracy is possible within the adopted framework of analysis. If no active market exists, the reliability of fair value (without quotes) and/or "fair value" (with quotes) of a non-financial asset is based on the model, which should be understood in a broad sense as the evaluation method of that asset, used by the preparer and discussed with the main actors, the auditor and the appraiser (or just the auditor, in the case of "fair value"). The fair value of such assets arises from a specific option of the organization given accounting regulation, and this option determines a specific process through which a figure will be constructed. 34 Inspired on Actor-Network Theory, this case shows that determining both fair value and "fair value" for non-financial assets involves various human actors within the company, in such diverse functions as the financial department, management control, technical and legal department. Understanding these construction processes requires acknowledging this diversity within organizations, with multiple collective and individual actors, rather than conceiving organizations as "monolithic" entities. This network of the company’s internal actors, central and local, contributes in a structured way to the establishment of a value, consistent with Okamoto (2014) concerning the fair value of complex financial instruments. However, it should be noted that local actors only produce the information requested by central actors, adjusting themselves to fair value regulatory requirements; that is, although the process has to necessarily go through these local actors, they are not “obligatory passage points" in the sense of Callon (1986), since the requirement of their intervention does not give them an importance and decisive power in this process. On the other hand, actors outside the company, the appraiser and the auditor, are also an integral part of the process, influenced, respectively, by formal and informal power. In turn, these external actors are also formed by various actors, from different teams to different individuals, and are also not "monolithic" entities. The central role of the auditor, based on its formal power of legal nature, changed with the high degree of complexity associated with fair value, having lost autonomy in relation to the appraiser. The informal power of the latter, through specific knowledge about the valuation techniques and its indirect inclusion as part of the audit process, has been "socially" accepted by the auditor. That is, despite the auditor clearly being an "obligatory passage point", its centrality and power in the process is likely to be reduced compared with other actors (in particular, the appraiser). Finally, we should note that relationships can be defined formally (for example, through a contract or proposal to provide appraisal services) or informally (for example, an internal procedures manual). For non-financial assets, identified non-human actors (e.g., Google Maps) were not considered highly relevant in the process of determining the fair value, unlike Okamoto (2014) found for financial assets, showing different market valuation practices between financial and non-financial sectors. 35 6 Conclusions, Reflections, Contributions and Study limitations The case study researched the social construction process of the fair value of non-financial assets in the particular context of a Portuguese industrial company with characteristics appropriate to the research topic (to have significant non-financial assets, non-existence of active market and lack of comparable transactions in the market). Inspired by Actor-Network Theory lenses, this study shows that the process underlying the fair value figure reported in financial statements is not a calculation. Instead, it is a "socially constructed" value, through highly specific processes, involving a network of actors and based on negotiated “guesstimates” (Smith-Lacroix et al., 2012), although produced by using valuation methods generally perceived and accepted as legitimate. The reliability of fair value based on unobservable inputs (the less relevant for the investor and the most contested in terms of that qualitative characteristic - Bagna et al, 2014; Chung et al, 2014) is thus not based in documental verifiability; instead, its reliability is based on the consensus among actors, as well as on a multitude of assumptions derived from various origins and nature and subject to dispute that the network needs to resolve to achieve one common goal: the presentation of a number. This study illustrated the constructive social process that occurred so that that number can be presented, having also provided information as to the importance of different actors involved in the construction process, and the fair value dependence towards estimates of third parties that result from a focus on qualitative materiality by the auditor (Christensen et al., 2012). We found that in this network of actors the auditor did not assume a central role; instead, the appraiser took an increasingly important role and was the actor where the process ends up being channeled, given its expertise in the evaluation. Therefore the appraiser is considered as the obligatory passage point in this process, as an actor that gained significant power in this relational network of actors. This reliance on appraisers has resulted in that the preparer has become an information compiler and not the appraiser itself (Power, 2010). The appraiser centrality has, however, been challenged by regulators of the audit function, having the PCAOB (Public Company Accounting Oversight Board) recently requested comments on a proposed amendment of the audit approach to the fair value and of the estimates and review procedures of the assumptions and methods used by the appraiser (PCAOB, 2014). This 36 contestation, an arena of power struggles at a supra-organizational level, reveals how the positions of centrality and relational power in specific actors networks are contingent and permanently subject to contestation, including contestations that arise from outside the immediate network. The preparer’s change of "role" described above (from "appraiser" to a mere “preparer" or even “compiler”) is an interesting result, from the point of view of the theoretical model used. On the one hand, the preparer’s presence remains central in the process, since he intervenes in virtually every step of the processes. But on the other hand, its centrality as an “obligatory passage point" (Callon, 1986) becomes smaller, with less relational power to influence the process and the network and therefore to achieve its interests (Oliveira and Clegg, 2014). In fact, the same has been already pointed out above with respect to the auditor and, at a lower level, for some secondary local actors. This finding suggests a potential refinement of concepts used in research inspired by Actor-Network Theory, to be developed in future studies. Finally, we stress that although the final numbers in the financial statements do not reflect the complexity and diversity of the processes that were involved in their collective construction, this study does not suggest that the values in the financial statements are 'wrong', but suggests that they based on a non-deterministic process. In fact, the concept of 'wrong' is not appropriate, because it implies the existence of a 'correct' and unique value, which does not exist. Considering the globalization of accounting and auditing practices, as well as their increased submission to the normative power of fair value, the processes identified in this study are probably not exclusive of the organization under study (Smith-Lacroix et al., 2012). Thus, although acknowledging the generalizability limitations of single case studies, this study contributes to the fair value literature, highlighting three results. The first refers to the focus on organizational and operational process of fair value, because it allows a better understanding of its "calculation" in real contexts, characterized by a process of social construction within a multi-actors network, in contrast to previous studies which assume fair value merely as a given value. The second contribution is a response to Hopwood’s (2009) 37 and Basu’s (2012) calls, through the emphasis of "accounting in action”, in real contexts, little discussed in the literature on financial accounting, and even less in the case of the valuation of non-financial assets. The third contribution is the evidence of change in the centrality of the actors involved. There are ambiguities inherent to the construction process of fair value, as the "singularity" of the evaluated asset (Callon and Minusa, 2003) (here illustrated with the decision of which part of the automated warehouse should be considered as a building). An extension of this research could be the study of the actors network involved in a process of equipment fair value, predicted to exist in a near future at TechnoCorp; in particular, we could expect an expansion of the actors involved, in particular with regard to the identification of assets to combine for valuation purposes. References Arnold, P. J. (2009), “Global Financial Crisis: the Challenge to Accounting Research”, Accounting Organizations and Society, 34, pp. 803-809. Ball R. (2009), “International Financial Reporting Standards (IFRS): Pros and Cons for Investors”, Accounting and Business Research, International Accounting Policy Forum, pp. 5-27. Bagna, E., G. Di Martino and D. Rossi, (2014),”An Anatomy of The Level3 Fair Value Hierarchy Discount”, available at http://ssrn.com/abstract=2367405, Access: 23 June 2014; Barth, M. E, W. H. Beaver and W. R. Landsman (2001), “The Relevance of the Value Relevance literature for Financial Accounting Standard Setting: Another View”, Journal of Accounting and Economics, Nº 31, pp. 77-104. Basu, S. (2012), “How Can Accounting Researchers Become More Innovative”, Commentary, Accounting Horizons, Vol. 26, Nº 4, pp. 851-870. 38 Beattie V. (2005), “Moving the Financial Accounting Research Front Forward: the UK Contribution”, The British Accounting Review, Nº 37, pp. 85-114. Bougen P. D. and J. J. Young (2012), “Fair Value Accounting: Simulacra and Simulation”, Critical Perspectives on Accounting, Nº 23, pp. 390-402. Bruen, C. (2010), 'The System Cannot Fail –Reflections on The Audit Society: Rituals of Verification”, Policy & Practice: A Development Education Review, Vol. 11, pp. 131-133, available at http://www.developmenteducationreview.com/issue11-review2, Access: 14 August 2014. Callon M. (1986), “Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St. Brieuc Bay”, J. Law (Ed.), Power, Action and Belief (pp.196-233), London: Routledge and Kegan Paul. Callon, M. and F. Minuesa (2003), “Les Marchés Économiques comme Dispositifs Collectifs de Calcul”, Réseaux 21, Nº122, pp. 189-233. Campbell, R. L., L. A. Owens-Jackson and D. R. Robinson (2008), “Fair Value Accounting from Theory to Practice: Implementation Tumultuous, Painful and Require Significant Adjustment for U.S. Accounting Professionals”, Strategic Finance, Nº 90, pp. 31-37. Canning, M., and B. O’Dwyer (2013), “The dynamics of a regulatory space realignment: strategic responses in a local context”, Accounting, Organizations and Society, Nº 38, pp. 169-194. Christensen H. B., S. M. Glover and D. A. Wood (2012), “Extreme Estimation Uncertainty in Fair Value Estimates: Implications for Audit Assurance”, Auditing: A Journal of Practice & Theory, Vol. 31, Nº 1, pp. 127-146. Christensen H. B. and V. V. Nikolaev (2013), “Does Fair Value Accounting for NonFinancial Assets Pass the Market Test?”, Review of Accounting Studies, Forthcoming, available at http://ssrn.com/abstract=1269515, Access: 20 June 2014. 39 Chung, S. G., B. W. Goh,, J. Ng, and K. O. Yong (2014), “ Voluntary Fair Value Disclosures Beyond SFAS 157’s Three-level Estimates”, available at http://ssrn.com/abstract=1335848, Access: 23 June 2014. Cotter, J. and S. Richardson (2002), “Reliability of Asset Revaluations: The Impact of Appraiser Independence”, Review of Accounting Studies, Nº 7, pp. 435-457. Gil, N. (2007), “On the value of project safeguards: Embedding real options in complex products and systems”, Research Policy, Vol. 36, Nº 7, pp. 980-999. Griffin, J. B. (2014), “The Effects of Uncertainty and Disclosure on Auditors’ Fair Value Materiality Decisions”, On-Line, Journal of Accouting Research, available at http://dx.doi.org/10.1111/1475-679X.12059, Access: 8 July 2014. Guerrreiro, M. S., Rodrigues, L.L. and Craig, R (2014), “Institutional Change of Accounting Systems: The Adoption of a Regime of Adapted International Financial Reporting Standards”, European Accounting Review, available at http://dx.doi.org/10.1080/097638180.2014.887477, Access: 7 August 2014. Gwilliam, D. and R. H. G. Jackson (2008), “Fair value in Financial Reporting: Problems and Pitfalls in Practice. A case study analysis of the use of fair valuation at Enron”, Accounting Forum, 32, pp. 240-259. Healy, P. M. and K. Palepu (2000), “Information Asymmetry, Corporate Disclosure and the Capital Markets: A Review of the Empirical Disclosure Literature”, JAE Rochester Conference, available at http://ssrn.com/abstract=258514, Access: 5 August 2014. Hellmann, A., H. Perera and C. Patel (2010), “Contextual Issues of the Convergence of International Financial Reporting Standards: The Case of Germany”, Advances in Accounting, incorporating Advances in International Accounting, Nº 26, pp. 108 – 116. 40 Herrmann, D., S. M. Saudagaran and W. B. Thomas (2006), “The Quality of Fair Value Measures for Property, Plant and Equipment”, Accounting Forum, 30, pp. 43-69; Hopwood, A. (2009), “The Economic Crisis and Accounting: Implications for the Research Community”, Accounting, Organizations and Society, Nº 34, pp. 797-802. IASB (International Accounting Standards Board) (2011), International Financial Reporting Standard, IFRS 13 – Fair Value Measurement. Kaplan R. S. (2011), “Accounting Scholarship that Advances Professional Knowledge and Practice”, The Accounting Review, Vol. 86, Nº. 2, pp. 367-383. Latour, B. (1997), “On Actor Network Theory: A Few Clarifications”, available at http://www.nettime.org/Lists-Archives/nettime-l-9801/msg00019.html, Access: 18 August 2014. Laux, C. and C. Leuz (2009), “The Crisis of Fair Value Accounting: Making Sense of the Recent Debate”, Accounting, Organizations and Society, Nº 34, pp. 826-834. Machado de Almeida B. J. (2014), “Significados, Poder e Auditoria: Uma Abordagem Interpretativa”, Revista Portuguesa de Contabilidade, Nº 14, pp. 185-256. Nissim, D. and S. Penman (2008), “Principles for the Application of Fair Value Accounting”, White Paper 2, Centre for Excellence in Accounting and Security Analysis (Columbia Business School), available at http://www.columbia.edu, Access: 16 August 2014. Okamoto N. (2014), “Fair Value Accounting from a Distributed Cognition Perspective”, Accounting Forum, Nº 38, pp. 170-183. Oler, D. K., M. J. Oler and C. J. Skousen (2010), “Characterizing Accounting Research”, Accounting Horizons, Vol. 24, Nº 4, pp. 635-670. 41 Oliveira J., J. Ribeiro and S. Pereira (2009), “Investigação em Contabilidade de Gestão”, In Contabilidade e Controlo de Gestão, Major, M. e R. Vieira (Eds), pp. 63-88, Lisboa: Escolar Editora. Oliveira, J. and S. R. Clegg (2014), “Organizational Control and Circuits of Power”, 37th European Accounting Association Congress, 21-23 May, Tallinn. PCAOB (Public Company Accounting Oversight Board) (2014), “Auditing Accounting Estimates and Fair Value Measurements”, Staff Consultation Paper, available at http://pcaobus.org/Standards/Documents, Access: 22 August 2014. Power, M. (2010), “Fair Value Accounting, Financial Economics and the Transformation of Reliability”, Accounting and Business Research, Vol. 40, International Policy Forum, pp. 197-210. Ryan, S (2008), “Accounting in and for the Subprime Crisis”, The Accounting Review, Vol. 83, Nº 6, pp. 1605-1638. Smith-Lacroix, J., S. Durocher and Y. Gendron (2012), “The Erosion of Jurisdiction: Auditing in a Market Value Accounting Regime”, Critical Perspectives on Accounting, Nº 23, pp. 36-53. Song, C. J, W. B. Thomas and H. Yi (2010), “Value Relevance of FAS 157 Fair Value Hierarchy Information and the Impact of Corporate Governance Mechanisms”, Accounting Review, Vol. 85, Nº 4, available at http://ssrn.com/abstract=1198142, Access: 10 June 2014. Toman C. M (2012), “Global Financial Crisis – An Accounting Literature Review and Overview”, available at http://anale.steconomiceuoradea.ro/volume/2012/n1/141.pdf, Access: 18 June 2014.