A note on auditing fair value of investment properties

Purpose – The purpose of this article is to study how fair values in financial reports are audited. Design/methodology/approach – The study is a qualitative case study based on in-depth interviews. Findings –One important finding is that auditors anchor in the figure presented by the company, and despite the auditing efforts, there is a substantial risk ofmanagement bias in the fair values reported. There is a risk for confirmation bias. Research limitations/implications – Relatively, few respondents were employed in this study, but their background and competence lead to the assessment that the study provides a representative picture of what is being investigated. Practical implications –Auditors may need to develop ways of performing auditing of fair values to reduce the risks identified in this study. Social implications –This study presents a perspective of the auditing process enabling an evaluation of the quality of fair value estimates regarding investment properties in the financial reports. This study also provides users of financial reports as investors, bankers and other institutions with an enhanced understanding of reported estimates of fair (market) values. Originality/value – Very few studies have investigated how auditors evaluate fair values of investment properties. This study contributes by giving users of financial reports an enhanced understanding of the quality of reported estimates of fair (market) values.


Introduction
The quality of property valuations is important for the function of property markets and for financial stability. A number of mechanisms can contribute to higher quality, e.g. different kinds of regulation and authorization of property valuers and/or various forms of government supervision. In this note, another mechanism is in focus, and that is how accountants audit fair values of investment properties when they audit financial reports. In Sweden, where this study was carried out, supervision of auditors is handled by the Swedish Inspectorate of Auditors (SIA). In other countries, there are corresponding monitoring functions.
Over the past 10-20 years, there have been some confusion regarding the definition of fair value. Such confusion may stem from different sources. One possible explanation could be that, in the context of accounting theory, fair value has been explained either as an entry price (e.g replacement cost) or an exit price (see, e.g. Penman, 2007). Another source of confusion could be the description of fair value in the International Valuation Standards IVS (2011), where fair value in the IVS framework was described as a value concept that could be distinguished from market value when estimated for other purposes than use in financial reporting. In 2014, the International Valuation Standards Council (IVSC) and International Financial Reporting Standards (IFRS) Foundation published a protocol where the aim of the agreement was ". . .to ensure that both organizations are able to cooperate effectively in this important area. . ." and to help ensure consistent application regarding measurement of fair value for financial reporting (see IVSC, 2014).
Auditing can be described as a process of quality assurance of information in financial reports. A number of studies have described client pressure and management bias in relation to property valuations (e.g. Achu, 2013;Selling and Nordlund, 2015), and questions have been raised about audit quality regarding accounting estimates (see for instance IAASB, 2019; Ahn et al., 2019 andJoe et al., 2017). The International Forum of Independent Audit Regulators (IFIAR) has since 2012 published an annual report that includes a list of the most common findings among its members. IFIAR (2014) and IFIAR (2015) are examples where both report shortcomings found in audit inspections regarding fair value measurements.
The International Audit Standards (ISA) state how audits are to be performed in different situations. ISA 540 (2020) is particularly interesting when it comes to reviewing estimates of, e.g. fair values in the financial reports. Among other things, this audit standard (para A139-A143) prescribes how auditors should evaluate whether there is an incorrect estimation of fair values or not: (1) The audit evidence can provide support for a point of estimate that differs from the point of estimate made by management. In these circumstances, the difference between the auditor's point of estimate and the management's point of estimate is defined as an error.
(2) The audit evidence may support an interval that does not include the management's point of estimate. In these circumstances, the inaccuracy is the difference between the management's point of estimate and the nearest point in the auditor's interval.
(3) When an estimate in the accounting is subject to a higher degree of uncertainty in the estimate, the auditor may decide that additional information is needed to obtain a true and fair view. If management does not include such additional information, the auditor may conclude that the financial statements contain a material misstatement.
The starting point for this study was the question if auditors work in such a way that bias in estimates will be found and corrected or reported. The purpose is to investigate how fair values of investment properties are audited. To fulfill this purpose, semistructured interviews were carried out with leading real estate auditors in Sweden. The method and the questions asked are described in Section 2 below, and the results are presented in Section 3. Analysis and conclusions made are presented in Section 4.

Selection of auditors
In total, five leading Swedish auditors from the audit firms PWC, EY, KPMG, Deloitte and Grant Thornton have been selected for the interviews. These are the leading auditing firms in Sweden and are working with real estate companies that report in accordance with the IFRS.
The respondents are leading auditors from each firm and are responsible for auditing the accounting and valuation issues of investment properties in Swedish real estate companies. The answers obtained give a good picture of how the auditors in Sweden are working and reasoning regarding the questions asked. Our assessment is that adding more respondents from each firm would not have contributed significantly to the results as firms can be expected to have internal guidelines and that the senior auditors set and interpret the rules. The respondents have been numbered 1-5 randomly in the presentation below.

Aspects studied
The following areas were covered in the interviews: (1) Value concepts: Their view of the meaning of fair value according to the definition in IFRS 13, the relation between fair value according to IFRS 13 and market value according to the IVS and the relation between entry price and fair value and also the relation to the investment value for a specific actor. Valuation methods: What are the auditors views on the methods applied in valuations and the reliability of different methods? How are the inputs in different methods checked? What are their views of uncertainty in valuations? What type of inputs was used in relation to the fair value hierarchy in IFRS 13 -Fair Value Measurement. The hierarchy is summarized below: Level 1 5 Inputs require no adjustments, based on quoted prices (for example, if the valuation object is a share in a listed company, it can be monitored on the stock exchange. It can be observed what exactly the same instrument is worth on the market).
Level 2 5 Inputs which are directly or indirectly observable from the market but require different degrees of adjustments (e.g. comparable properties sold in the property market, close in time in relation to the value date and sold objects whose other characteristics [1] are not significantly different from the valuation object). Level 2 inputs require less significant adjustments to be relevant.
Level 3 5 Unobservable inputs. This group may also include input from the market that requires significant adjustments. Examples of the latter might be comparable objects sold on the property market, which occurred further back in time or when the sold objects' other characteristics differ significantly from the valuation object. Level 3 inputs are usually determined based on management's assumptions.
We were also interested in the auditor's view on some more general issues, e.g. differences between markets with many transactions and thin markets.
(1) The auditing process: For example, do auditors start with the assessed value presented by management or do they, in a first step, make independent valuations of some properties?
(2) The disclosure of information: What information, connected to fair value estimates of investment properties, should, according to the auditors, be presented in the financial report?
2.3 How the interviews were conducted All interviews were conducted by two of the authors face-to-face with the respondents and with digital tools such as Zoom and Teams. All of the interviews were conducted by the same two authors.
The questions to the respondents were sent to them in advance, in order to give them the opportunity to prepare for the interviews.
Compiled responses were sent to the respondents to verify that we had understood their answers correctly.
3. Results from the empirical study 3.1 Value concepts Concerning value concepts, all respondents agreed that fair value according to IFRS 13 and market value according to the IVS should give the same value figure. One of the respondents (No 1), however, expressed uncertainty about the exact meaning of market value according to the definition in the IVS. The respondents also agreed that the entry price at the time of acquisition normally corresponds to the exit price, as defined by fair value in IFRS 13.
One of the respondents (No 4) commented that the fair value of the properties in the financial reports is not really to be regarded as a sales price. It could be difficult to see the entire property portfolio's fair value in a company as a good approximation of what could be obtained from a sale at the time of the report. Another respondent (No 3) stated that it is important to have a long-term perspective on the value, even if the market is volatile and prices/values move up and down in a short-term perspective. This could be described as a "smoothing effect" where the estimated values do not "take the turns" in either ups and downs. One view (No 3) was that valuations should be "relevant" rather than "correct". It was also argued that in any case, the values are never exactly correct.
When it comes to the evaluation of the estimated final value figure, there seem to be some differences. One respondent (No 1) was very clear that the value for each property is evaluated, while another (No 4) expressed that in the end it is most important that the estimated value is "correct" from the perspective of the property portfolio as a whole.
The most common view among the respondents (3, 4 and 5) when evaluating whether there could be an error in the estimates regarding fair values was that the evaluation was made in relation to the aggregated value of the property portfolio as a whole. Some respondents also mentioned that an uncertainty interval of approx. ± 10% was common in fair value estimates made for investment properties. One respondent (1) made it clear that potential errors in this respect were evaluated for each property alone.

Valuation methods
Concerning the valuation methods, the common view among the respondents was that the properties are appraised applying a discounted cash flow (DCF) method. One stated (No 2) that the usual method is a comparable sales method and then added that DCF method is used when valuing more unique objects [2]. All respondents expressed the view that the comparable sales method is used or used to some extent, e.g. when extracting the direct yield demanded in the market. They were not clear about how this yield actually was used in the DCF method.
In total, two follow-up questions were added. The first concerned special transactions like exchange transactions, other kinds of off-market deals, deals driven by synergies for a certain buyer or situations where special financing terms could have had impact on prices. The other follow-up question was about what method to apply if there are very few or no transactions on the market.
Regarding special transactions, there was not a clear view among the respondents how to handle price information from such transactions. The most common answer was that if it was clear that such transactions were not expressing market views, they would be used with care or be put aside.
Regarding what method to use when there are few transactions and what kind of information needed in these situations, the common view (No 2, 3 and 4) was that the DCF method was the first choice in such situations. None of the respondents clearly discussed the need for other market information or described in more detail what additional information would be needed in such situations. One respondent (No 4) clearly stated that it always poses a problem if the market is illiquid.
All respondents agreed that level 3 inputs, according to the fair value hierarchy in IFRS 13, were used in the valuations.

The auditing process
When asked to describe how audits were performed, it was first noted that in most cases a valuation specialist was connected to the audit team. This specialist was in all cases from the same firm as the auditor, but one respondent (No 4) also said that the specialist could be from an outside firm.
The auditors typically start with the fair value estimation presented by the company and then review this estimate. Some of the respondents mentioned that the owners of the properties either "know their properties very well, which is critical for a correct valuation" (No 5) or that it is "hard to see what value it could add if the auditor first make an own estimation since the owner of the properties knows their properties best" (No 3).
In order to check the accuracy of the reported fair values, most auditors (No 1, 4 and 5) make a follow-up-analysis on how selling prices of previously held properties were in relation to the latest valuation. In that way, they get an idea of the quality of the fair value estimates presented by the company. A majority of the respondents (No 1, 2 and 5) also emphasized the importance of reviewing the valuation model applied by the client. One respondent (No 5) underlined the importance of looking at rental income, net operating income, direct capitalization yield, etc. in accordance with accepted and/or international definitions of these concepts in the valuations.
It is common (No 1, 2 and 5) that auditors focus on verifying various individual parameters used as input data in valuation models (such as direct yield requirements, rents and net operating income, etc). The respondents emphasized that it is important that such inputs in valuation models can be verified by comparing them to actual outcomes in financial reports. Of the respondents, two (No 1 and 2) clearly stated that inputs like market requirement for yield is checked against "Datscha" [3] which is the dominating market information system in Sweden.
None of the respondents explicitly stated that there is a normal procedure in the auditing to check whether presented valuations are in line with valuation standards, e.g. the IVS or RICS Red Book. However, some of the respondents were of the opinion that most of the requirements in these standards were checked up anyway, one way or another. However, they did not explain exactly how this check-up was carried out.

Disclosure in financial reports
A common view among the respondents was that the requirements in so called "disclosure checklists" were central. All respondents agreed on the importance of these checklists. Disclosure checklists contain the minimum requirements in IFRS standards, e.g IFRS 13 or IAS 40.
Companies normally report a sensitivity analysis regarding the estimated fair value and different inputs in valuations. Most of the respondents (No 2, 4 and 5) had the opinion that on many occasions these sensitivity analysis and other disclosures were presented on a "too high" aggregation level of the companies' property portfolios. The analyses were looked upon more as mathematical illustrations. The relevance of outcomes in disclosed sensitivity analysis was not clear.
When asked to give an example of critical disclosures in connection to fair value estimates of investment properties that was not explicitly listed in the standards (e.g IFRS 13 or IAS 40), very few suggestions were made. One respondent (No 4), however, stressed that important assumptions also need to be justified and motivated. Another respondent (5) emphasized that the disclosures about valuations should clearly mention which of the inputs were most important for the fair value estimations. The respondent also emphasized the importance of an analysis of actual outcomes of net operating incomes for a number of years in the company, and that this was showed for a like-for-like portfolio. One respondent (No 1) mentioned European Public Real Estate Association (EPRA) definitions of yield and vacancy rates as important.
It was also argued (No 3) that the disclosure requirements in IFRS 13 and IAS 40 are far reaching and contain more than the information that a user of financial reports needs.

Analysis and implications
This study shows that the starting point for the audit is the companies' estimates and disclosures. This anchoring procedure means that the burden of proof is on the one who wants to change the value. Various inputs in the calculation were checked against other sources by the auditors, and in this way large discrepancies can be detected, but as there are uncertainties in all these inputs the auditors may hesitate to question the value presented by the company. This opens up for management bias as the auditors would not react if the difference is say ± 10%. The respondents also regard ± 10 % deviance as an acceptable margin of error. Uncertainty in valuations have been discussed in many studies (e.g. Crosby, 2000;Bretten and Wyatt, 2001;RICS, 2002;French and Gabrielli, 2004;Nellessen and Zuelch, 2011). The answers further indicate that the auditors believe that companies possess expert power and that the auditors need strong evidence to question the values set by the companies. These results indicate that there is a need to strengthen the independence of auditors in relation to their clients.
Another observation is that there are divergent views on how the values in the balance sheet should be interpreted. According to the IFRS (IAS 40), the fair value of an investment property should reflect the value of the single property and not include a portfolio perspective. It can, however, be questioned to what extent this requirement is fulfilled by the auditors in this study. One respondent argued that the fair value is not to be regarded as a probable sales price at a specific date. Another respondent stated that it is important to have more of a long-term perspective of the value even if the market has its ups and downs. It was also argued that it was more important that the value of the portfolio as a whole was correct than that the value of each property was correct which is not in line with IAS 40. Prior studies also indicate that the reported values may not meet the definition of fair value (Hitz, 2007;Lorentzon, 2011). All together the results of this study clearly show that among the respondents in this study, there are opinions about the relevant value that is not in line with fair value as defined by IFRS 13.
Our impression is also that the auditors did not focus on potential problems related to the nature of different capital expenditures (capex) and/or free cash flow yield or free operating cash flow. Such a free cash flow yield can be necessary to evaluate as a complement to yields extracted from the net operating income/net rental income in income statements and their relation to fair values (see Young et al., 2017;Reid, 2019;Nordlund, 2014). It can be important for users of financial statements to be able to evaluate the cash flow capacity in relation to fair values.
We believe that in order to reduce the anchoring effect, the auditors may need to employ independent valuers on their behalf, performing some valuations of investment properties in audit engagements before they look at the valuations presented by management.
The findings in this study also implies that there is a need for auditors to be more aware of the meaning of the relevant value concept as defined in the context of financial reports (i.e in IFRS 13). According to our understanding of ISA 540, referred to in the Introduction above, the auditors are expected to have their own opinion about a correct fair value to compare with the fair value estimates presented by the audit client. Otherwise, it will be difficult for the auditors to have an opinion whether "the difference between the auditor's point of estimate and the management's point of estimate is an error". What has been brought forward here regarding value concepts and adequate disclosures also imply that there is a need for more efforts from surveillance and supervisory bodies when it comes to fair values in both financial reporting and auditing regarding such values. On many occasions, auditors are probably also in need of more proper training about requirements in, e.g. IFRS 13 and IAS 40 when it comes to auditing of property companies. A correct understanding and interpretation of the fair value concept in IFRS 13 is critical to be able to have an opinion about whether presented values in financial statements ore correct or not. It is also important that when valuation specialists are connected to the audit team, such specialists should be specialized on the real estate market and not be general valuation specialists. The specialists those focus on property valuation also need to be aware of the relevant value concepts.
A final comment is that there is a need for more research on various aspects of the quality control process for property valuations. The focus would then be not only on the role of auditors but also on how different types of regulations and authorization systems really work. Notes 1. Important characteristics for properties could be e.g. site/location, type of property, rental income level, vacancy level or technical condition.
2. The Swedish use of the DCF method is analyzed in Lind and Nordlund (2014), and they conclude that it can be interpreted as an income-based comparative sales method.