IAS 40: fair value or cost? An empirical analysis of the European real estate industry Open Access

The valuation of investment property (hereafter IP) is a critical factor in the real estate industry, as IP often represents the most significant asset on these firms' balance sheets (Christensen and Nikolaev, 2013; Quagli and Avallone, 2010). Accurate IP valuation impacts not only reported earnings and balance sheet values of real estate companies but also the overall quality of financial information used by investors, creditors, and other stakeholders in decision-making. Therefore, understanding how these assets are evaluated is important to both standard setters and financial statement users. Indeed, the primary objective of accounting information is to provide decision-useful information, and the valuation approach adopted for IP directly impacts the relevance and reliability of financial statement numbers.

In particular, the choice between fair value and historical cost accounting for IP, as permitted under IAS 40, directly influences the transparency of financial reporting and affects users' ability to evaluate firm performance and risk (Barth et al., 2008; Muller et al., 2011). Thus, insight into the valuation methods of IP enhances the quality of financial information, aiding users in making informed decisions.

For standard setters and regulators, understanding the factors that drive these valuation choices is essential to promote consistent and comparable financial reporting practices across firms and jurisdictions (Ball, 2006).

The International Accounting Standard 40 (IAS 40), reissued in 2003 and effective for financial statements starting on or after January 1, 2005, allows firms to choose between the fair value and cost models for IP valuation (International Accounting Standards Board, 2003). While IAS 40 expresses a clear preference for fair value accounting ^[1], the choice between the two valuation methods is still permitted. This option was introduced to give financial statement preparers and users time to become familiar with the fair value model and to allow real estate markets and valuation professions, particularly in less developed countries, to mature (IASB, 2003).

In this context it is interesting to analyze the actual adoption of IFRS evaluation options of European real estate firms as it allows to have a deeper insight into the factors influencing these choices and to assess whether the fair value model is, in fact, becoming more widely adopted.

Empirical research on accounting choice indicates that multiple motivations, arising from market imperfections, influence managers' decisions. These include asset pricing motivations (agency costs) and contractual efficiency motivations (information asymmetries) (Holthausen, 1990; Fields et al., 2001; Quagli and Avallone, 2010; Christensen and Nikolaev, 2013; Israeli, 2015; Olante and Lassini, 2022). Additionally, country and industry factors significantly influence IFRS policy choices (Stadler and Nobes, 2014; Isidro et al., 2016), with institutional contexts, including pre-IFRS accounting practices, continuing to play a role in firms' accounting decisions (Kvaal and Nobes, 2010, 2012; Nobes, 2013).

Despite the flexibility offered by IAS 40, the long-term factors influencing the choice between fair value and historical cost remain underexplored, particularly within the unique business ecosystem of the real estate industry. As firms and managers gain more experience with IAS 40, it is unclear whether choices to adopt fair value over cost reflect the intrinsic benefits of the fair value model, or whether they continue to be shaped by external and internal factors, such as contractual efficiency, asset pricing incentives, and the institutional context. This study aims to fill this research gap by examining the factors that drive managerial decisions in IP valuation a decade after the first-time adoption of IFRS in Europe.

This study aims to address the following research question: What factors continue to shape the managerial decision between fair value and cost for investment property under IAS 40 in European real estate firms, ten years after the initial adoption of IFRS? Building on prior literature, I hypothesize that various motivations affect managerial behavior regarding the choice between fair value and historical cost for IP. These include contractual efficiency motives (e.g. firm reliance on debt and political costs); asset pricing incentives tied to information asymmetries and country-specific institutional factors.

Only a limited number of studies (Olante and Lassini, 2022; Israeli, 2015; Christensen and Nikolaev, 2013; Quagli and Avallone, 2010) deals with this issue, focusing specifically on what drives management adoption of accounting alternatives treatments permitted by IAS 40. Moreover, except for Olante and Lassini (2022), these studies focus on early responses to IAS 40, as they analyze the choice between fair value and cost in the first year of IAS 40 enactment or immediately after. In addition, only Quagli and Avallone (2010), focus specifically on the real estate industry. To the best of my knowledge, this paper is the first to provide some insight into the choice between fair value and cost for IP in the real estate industry after more than 10 years from the IFRS first-time adoption and covering data up to 2023. Hence, the observation of likely more expert managers’ preferences, when different accounting treatments are allowed, can provide additional and relevant evidence about the consistency of those choices with the IASB onward shift towards fair value accounting.

To test these hypotheses, I hand-collected IP data for the period 2014–2023 from the annual financial statements of 212 listed real estate companies located in nine major EU economies (i.e. United Kingdom, Ireland, Sweden, Finland, Spain, Italy, France, Germany, and Greece), each prescribing either the fair value or cost model in their pre-IFRS GAAP.

My findings indicate that approximately 45% of the variance in the choice between fair value and cost can be explained by the variables analyzed, suggesting that multiple factors discussed in the accounting choice literature—including contractual efficiency motives, asset pricing incentives, and institutional contexts—still play a role in determining IP valuation choices.

Specifically, the firm’s amount of debt (leverage), is negatively associated with the fair value choice, suggesting that debt holders tend to favor conservatism accounting as it is perceived to reduce agency costs through a greater lenders’ protection (Watts, 2003; Qiang, 2007). Additionally, firms with significant IP relative to total assets are more inclined to use fair value, as this method arguably better reflects performance variations than historical cost for these core assets. Larger firms are more inclined to select fair value, likely due to the distinct characteristics of the real estate industry where markets for IP are typically more liquid than in other sectors. Consequently, fair value estimates are perceived to be more reliable and do not contribute to increased shareholder litigation and associated costs. Moreover, the negative association between the market-to-book value ratio (my proxy for information asymmetries) and fair value choice suggests that by narrowing the gap between book and market values, fair value may contribute to mitigating information asymmetries, allowing a better alignment of financial statements with market expectations and thus ultimately enhancing the relevance and reliability of reported figures. Finally, the country’s institutional context, particularly legal origin, capital market development, and the openness or closeness of a society also strongly influences the fair value choice.

This study makes several contributions to existing literature. First, I integrate indicators from various strands of prior research and propose a multi-motivation model (contractual efficiency, asset pricing incentives, and country factors) to provide a better understanding of the complexity behind managerial decisions regarding accounting choices for IP in the real estate industry. Moreover, while previous studies have focused either on the traditional contrast between the UK and Germany (Christensen and Nikolaev, 2013) or on selections of European countries that mandated only the cost method before IFRS (Quagli and Avallone, 2010; Israeli, 2015), my sample includes nine European countries, each adopting either the fair value or the cost model in their pre-IFRS national GAAP. This allows me to broaden the scope of the analysis and to explore more in depth the complex relationship between institutional context and accounting choice.

Finally, this study not only corroborates but also extends previous findings on accounting choices for investment property, offering new insights based on the post-IAS 40 adoption period and covering data up to 2023. While prior research has primarily concentrated on the first-time adoption of IAS 40 or the immediate aftermath, to the best of our knowledge, no study has yet investigated real estate firms’ actual choices regarding fair value versus historical cost in such a recent and extended time frame.

Lastly, this research provides valuable insights for standard setters, offering an updated perspective on the ongoing learning process associated with the use of fair value accounting within European real estate firms. This knowledge is critical for informing future regulatory decisions that could enhance the transparency and comparability of financial reporting in the real estate sector.

IAS 40 Investment Property establishes the accounting treatment for IP, defined as land or buildings held to earn rentals or for capital appreciation, or both by the owner or the lessee in a finance lease. IPs are properties not owner-occupied, used in production, or held for sale in the ordinary course of business (IAS 40, par. 5).

Initially, IPs must be measured at cost (IAS 40, par. 20). After the initial recognition, IAS 40 permits entities to choose between the fair value model or the cost model (IAS 40, par. 30). The selected method must be applied consistently to all IPs held by the entity. Under the fair value model, properties are measured at fair value, with changes recognized in profit or loss, and no depreciation is recorded (IAS 40, parr.33-55). On the other hand, the cost model requires to measure IP at cost minus accumulated depreciation and impairment losses, with fair value disclosed in financial statement notes (IAS 40, par. 56), except for rare instances where fair value cannot be reliably determined ^[2].

A change in the adopted model is considered a voluntary change in accounting policy which, according to IAS 8 Accounting Policies, Changes in Accounting Estimates and Errors, is allowed only if it results in more reliable and relevant financial information ^[3].

Therefore, after initially deciding to evaluate IP using the cost model, a firm can switch to the fair value model. However, the Standard states that it is highly unlikely that changing from the fair value model to the cost model will provide a more relevant presentation ^[4].

The IASB has debated mandating the fair value model but retained the option to use the cost method to allow countries with less-developed property markets and valuation professions to mature (IAS, 40, 2003, Basis for conclusions). The Board decided to reconsider the option at a later time, presumably mandating the use of fair value for all IP recognized in the financial statements.

Before the IFRS adoption, IP accounting varied across Europe as they were evaluated according to the provisions of each country’s local GAAP. Most EU countries’ rules, including Italy, France, Spain, Germany, Greece, Finland and Sweden prescribed the cost method, whereas in the UK and Ireland accounting standards required IP to be recognized on the balance sheet at open market value ^[5], without depreciation. Changes in value were reported on the statement of recognized gains and losses (and credited to a revaluation reserve) unless a deficit was expected to be permanent ^[6].

The transition to the IAS 40 fair value model, which represents the IAS 40 benchmark model (IAS 40, par. 31), marked a significant change for many EU countries.

Given the different treatments required by each country’s local GAAP and IFRS, my sample of listed companies from nine countries presents a unique opportunity to assess whether different accounting traditions, as an element of the institutional context, influence the choice between fair value and cost.

The accounting choice between fair value and cost is one of the most extensively discussed topics in contemporary accounting debates. The selection of accounting methods by Standard Setters, such as the International Accounting Standards Board (IASB), is driven by the goal of providing useful information to investors, lenders, and creditors, as stated in their conceptual framework. Achieving this goal involves balancing the trade-off between relevance and reliability, which are the primary characteristics of useful-for-decisions information ^[7].

On the one hand, fair value measurements are generally considered more relevant as they reflect specific time and market conditions and therefore financial statements based on fair value provide more pertinent and timely information (Sharpe and Walker, 1975; Standish and Ung, 1982; Easton et al., 1993; Barth and Clinch, 1996, 1998; Aboody et al., 1999; Danbolt and Rees, 2008; Herrmann, 2006; CFA Institute Centre, 2008; Barth et al., 2008).

On the other hand, prior authors (e.g. Martin et al., 2006; Watts, 2006; Whittington, 2008; Christensen et al., 2012; Barker, 2015) have criticized fair value for producing less reliable figures, thereby potentially undermining the stewardship function of financial statements ^[8].

Regarding IP, IAS 40 indicates a clear preference for fair value accounting, which is regarded as the benchmark treatment ^[9], however the choice between the two valuation methods remains permitted. In this context, the analysis of real estate firms’ actual behavior can provide a deeper understanding about the drivers influencing managers’ choice between valuation options, therefore contributing to the advancement of accounting choice theory.

A limited number of studies (Quagli and Avallone, 2010; Christensen and Nikolaev, 2013; Israeli, 2015; Olante and Lassini, 2022) have addressed this topic, specifically examining the factors that influence management’s decision to adopt alternative accounting treatments for investment property under IFRS. Table 1 presents the relevant literature’s focus, main results, and contribution.

Specifically, three studies analyzed the choice between fair value and cost at the first adoption of IAS 40.

Quagli and Avallone (2010) analyzed a sample of 73 real estate companies from European countries where the cost method was mandated under pre-IFRS regulations, focusing on their choices between fair value and cost at the IAS 40 first time adoption. They found that factors such as information asymmetry, contractual efficiency, and managerial opportunism influenced the choice of fair value. Specifically, firm size, a proxy for political costs, was negatively associated with the adoption of fair value, while market-to-book ratio, a proxy for information asymmetry, had a negative relation with fair value adoption.

Christensen and Nikolaev (2013) studied the choice between cost and fair value for investment property and other non-financial assets (i.e. property, plant, and equipment and intangible assets) by a sample of companies based in the UK and Germany at the first adoption of IFRS. Their results show that fair value is more likely to be used when reliable estimates are available at lower cost and when such valuations provide useful information about a firm’s operating performance. Regarding investment property, they found that companies with higher levels of debt financing were more inclined to adopt fair value and that pre-IFRS practices were a significant determinant of the decision to adopt fair value accounting.

Israeli (2015) examined the first-time adoption of IAS 40 in a sample of publicly listed European companies, investigating the decision to recognize or disclose fair values for investment properties in financial statements. His findings suggest that both contractual factors and asset pricing incentives played a significant role in explaining the choice between recognition and disclosure.

Finally, a recent study by Olante and Lassini (2022) analyzed the choice between fair value and cost by European firms located in different countries and operating in different industries, with a particular focus on manufacturing and banking industries, in more recent years. They found that asset pricing incentives and contractual motives as well as country and industry specific characteristics all play a role in explaining the choice.

My study aims to extend this body of literature by identifying multiple factors that influence managers' decisions between fair value and cost for investment property in the real estate industry after ten years have passed from the first adoption of IAS 40, and by proposing a model to explain these choices.

Previous literature on accounting choice identifies several incentives that help explain management’s selection among the different valuation options permitted by accounting standards (Watts and Zimmerman, 1978, 1979; Holthausen, 1990; Fields et al., 2001; Aboody et al., 2004; Choudhary et al., 2009; Israeli, 2015).

The first set of incentives is represented by contractual incentives tied to agency costs. The efficient contracting perspective (Watts and Zimmerman, 1986; Holthausen and Leftwich, 1983) suggests that accounting choices can be explained by contractual incentives arising from agency costs in imperfect markets. These costs are generally linked to contractual characteristics, such as managerial compensation and debt covenants, with contractual arrangements designed to align incentives. However, according to the opportunistic perspective, managers may have incentives to make ex-post accounting choices to influence the firm’s contractual arrangements. In this context, the selection of accounting alternatives may be driven by objectives such as increasing compensation, preventing covenant violations, or achieving multiple goals simultaneously (Fields et al., 2001).

Prior research on accounting choice indicates that specific factors are associated with accounting decisions, such as a firm’s reliance on debt, measured by the leverage ratio, which serves as a proxy for the firm’s proximity to violating its debt covenants. In my context, managers might be more inclined to choose fair value to increase the book value of equity and total assets, thereby avoiding violations of debt covenant linked to financial statement data, but the evidence provided by the accounting literature is mixed.

Watts (2003) highlights that much of the information that enhances the timeliness and informativeness of accounting measures (e.g. earnings and net assets) lacks verifiability. Consequently, estimates are not used in contracts due to their unverifiable nature. Debt contracts employ lower bound measures, which are typically calculated based on orderly liquidation assumptions, and therefore conservatism.

Debt holders favor conservatism (i.e. the cost model) in debt covenants because it reduces the likelihood of management overstating earnings and assets and making dividend payments to shareholders at the expense of debt holders. Thus, conservatism better protects the interests of debt holders by lowering agency costs, leading debt holders to prefer contracts that consider historical cost measures.

Conversely, despite the preference for conservatism in debt covenants, credit agreements often require companies to provide information about the fair value of collateral. Consequently, firms accessing debt markets are likely to face demand for fair value accounting. In such cases, the incremental cost of obtaining reliable fair value estimates for financial reporting is minimal since these estimates have already been produced for financing purposes. Thus, fair value is more likely to be utilized when a firm has a significant reliance on debt. Additionally, fair value accounting might be preferred over cost accounting as it provides a timelier measure of the current value of assets, thereby reflecting the firm’s real solvency capacity more accurately and facilitating more efficient debt covenant negotiations (Christensen and Nikolaev, 2013).

Finally, other authors (Quagli and Avallone, 2010; Olante and Lassini, 2022) show that firm’s leverage does not have any predictive power in explaining the choice between fair value and cost for IAS 40, both in the real estate and in other industries.

Given this mixed evidence, my first hypothesis is:

From a contractual perspective, the choice between fair value and cost has another significant implication. The understatement of net asset value resulting from conservatism (i.e. historical cost) reduces the firm’s expected litigation and associated costs (i.e. political costs). Conversely, the use of fair value is more likely to increase shareholder litigation and associated costs. Previous literature also indicates that conservative accounting practices reduce political costs, while the adoption of fair value tends to increase them due to higher reported profits, which increase company visibility (Hagerman and Zmijewski, 1979; Watts, 2003).

However, according to prior literature, this is true when the markets for an asset are illiquid. In fact, fair value numbers provide verifiable values of assets when their markets are liquid and, in contrast, if an asset does not have a liquid market, it tends to be valued at verifiable historical cost. Therefore, when the market for an asset is illiquid, the doubtful verifiability of fair value compared to cost measures could lead to an increase in litigation and its related costs (Watts, 2003).

Previous studies indicate that the magnitude of political costs is significantly influenced by firm size (e.g. Watts and Zimmerman, 1978). Managers utilize accounting choices to minimize reported earnings to decrease the probability of adverse political actions, as the public often associates high reported profits with monopoly rents, thereby reducing the expected costs of such actions.

Therefore, I consider firm size as an independent variable to examine the impact of political costs on the choice between fair value and cost. Some previous literature shows that political costs increase with firm size. Consequently, larger firms face higher political costs, which decreases the likelihood of selecting fair value accounting ^[10].

However, in the real estate industry, markets for investment property are usually more liquid than in other contexts, thus fair value estimates are perceived to be more reliable and hence we could expect that even larger firms do not perceive the choice of fair value as increasing the likelihood of incurring higher political costs.

Given those competing arguments, I formulate my second hypothesis as follows:

In addition to contractual incentives, prior literature identifies asset pricing motivations (information asymmetry) that can drive management choices among different accounting methods.

Asset pricing incentives are generally related to information asymmetries between better-informed managers and less-informed investors. Managers’ accounting choices, influenced by information asymmetry, aim to impact market prices or returns (Fields et al., 2001). In this context, managers may opt for fair value to accurately convey information about the firm’s real value to the market. Fair value is preferred because it provides a higher, timelier, and thus more relevant level of information in financial statements (Barlev and Haddad, 2003; Ball, 2006; Danbolt and Rees, 2008; Whittington, 2008). Additionally, enhanced information helps reduce information asymmetry, potentially lowering the cost of capital (Armstrong et al., 2011).

If the cost method is selected, the fair value of IP must still be disclosed in the financial statement footnotes. However, previous studies indicate that recognition differs from disclosure, and investors assign lower valuation weights to disclosed amounts compared to recognized amounts (Schipper, 2007; Israeli, 2015). Therefore, I presume a positive relationship between the choice to recognize rather than disclose fair value for IP and asset pricing motivations influenced by information asymmetries.

Many researchers view the market-to-book value of equity ratio as a proxy for information asymmetry (Smith and Watts, 1992; Amir and Lev, 1996; Quagli and Avallone, 2010; Israeli, 2015; Lin et al., 2017; Olante and Lassini, 2022), where market value reflects a firm’s future growth prospects and book value represents the current value of its assets. A higher difference between these values indicates greater information asymmetries, leading managers to potentially choose fair value to mitigate them and thus reduce the cost of capital.

Consequently, if fair value effectively reduces information asymmetries, a firm should observe a decrease in the gap between its market and book values following the choice of fair value.

Therefore, I formulate my third hypothesis as follows:

Another body of accounting literature indicates that significant differences persist among countries applying International Financial Reporting Standards (La Porta et al., 1996, 1997; Ball et al., 2000; Ball, 2006; Nobes, 2006; Zeff, 2007; Muller et al., 2008, 2011; Kvaal and Nobes, 2010, 2012; Christensen and Nikolaev, 2013; Stadler and Nobes, 2014; Olante and Lassini, 2022). Taken together, the findings of these studies demonstrate that the institutional context significantly affects firms' accounting choices, as national accounting practices result from diverse institutional contexts, leading to significant international variations.

Isidro et al. (2016) examine country attributes identified by previous literature to explain global financial reporting differences. Specifically, they analyze 72 variables used to measure differences in economic, cultural, institutional, and societal development across 35 countries. The authors employ factor analysis to empirically derive four latent factors that capture the combined explanatory power of these country-level institutional characteristics for reporting outcomes worldwide and compute standardized factor scores for each analyzed country.

The study demonstrates that these four latent factors, collectively explain a significant portion of the observed cross-country variation in financial reporting outcomes.

Thus, analyzing the accounting choices of firms located in different countries can contribute to the literature on the impact of institutional contexts on accounting decisions. My dataset of European companies, each with distinct institutional backgrounds, provides a unique setting to examine whether these backgrounds still influence the accounting choice for IP, ten years after the initial adoption of IFRS. Specifically, I expect a shift from cost to fair value after the first adoption of IFRS, potentially due to increased managerial familiarity with fair value. Nonetheless, I expect that the characteristics of different national institutional contexts continue to influence firms' accounting choices to some extent.

To account for the institutional context, I follow the approach of Isidro et al. (2016), using the latent factors identified by the authors as proxies for the institutional characteristics of the countries included in my sample.

Eventually, previous research suggests that fair value is more commonly adopted when it facilitates performance measurement. This is particularly true for firms whose main business is real estate, where realizing capital gains from trading IP is part of the business model and thus fair value provides a more accurate measure of periodic income than historical cost (Christensen and Nikolaev, 2013; Olante and Lassini, 2022).

Therefore, I consider the potential impact of the materiality of a firm’s IP relative to its total assets on its accounting choice, expecting a positive relationship between the selection of fair value and the materiality of IP.

I formulate my fifth hypothesis as follows:

I test my hypotheses by analyzing IAS 40 accounting choices of a sample of European real estate firms from 2014 to 2023.

Data for my analyses was collected from the ORBIS Bureau van Dijck database. My initial sample included all companies for the selected ten-year period, that met the following criteria: the company’s registered office is in Finland, France, Germany, Greece, Ireland, Italy, Spain, Sweden, United Kingdom; the company is active, listed, classified in Nace Rev. 2 section L (Real estate activities) and the value of investment properties is known for at least one of the selected periods.

The selection of these specific countries is directly informed by the study conducted by Quagli and Avallone (2010), which examined the accounting choices of real estate firms regarding fair value versus historical cost for investment property at the time of the first adoption of IAS 40. Quagli and Avallone focus on Finland, France, Germany, Greece, Italy, Spain, and Sweden, where pre-IFRS national GAAP mandated the exclusive use of historical cost for investment property. My study aims to extend their analysis by evaluating the accounting choices made by real estate firms in these same countries a decade after the IFRS first-time adoption, providing an opportunity to assess whether managers' preferences have evolved with increased experience in using IAS 40.

Additionally, I have included the United Kingdom and Ireland in the sample to capture the impact of different accounting traditions, as these countries historically mandated fair value under their pre-IFRS GAAP.

Incorporating countries with distinct institutional backgrounds and accounting practices enables a broader analysis of how institutional contexts and regulatory backgrounds may influence the choice between fair value and historical cost, therefore offering insights into the long-term effects of IAS 40 on accounting practices within the European real estate sector.

To validate the information obtained from the ORBIS database I checked the accounting policy section of firms’ annual reports to verify the accounting standards followed by each company and the IP valuation practice.

Finally, data needed to build my independent variables are hand-collected from financial statements.

This resulted in a final sample of 212 real estate listed firms and 1.513 firm-years observations.

In all, I identified 1.275 firm-year choices for fair value (84,3% of the final sample) and 238 firm-year choices for historical cost (15,7% of the final sample).

First, I present descriptive information on the hypothesized explanatory variables. Then, I perform an independent sample t-test to analyze the differences in these variables between firms opting for far value and those opting for historical cost.

Then, to understand the impact of independent variables on the firms’ choice, I estimate the following model:

My model assumes as a dependent variable a dichotomous variable (FAIR_i), that is the decision to recognize IP in the balance sheet using either fair value or cost. Specifically, the dependent variable assumes the value of one if the company i used the fair value model under IAS 40 in the analyzed year, and zero if the company i adopted the historical cost model. I define SIZE_i as the log of total assets of firm i at the end of the analyzed year; LEV_i is the debt to total assets ratio of firm i measured at the end of the examined year; MTBV_i is the market-to-book value of firm i calculated over the last month of the analyzed year; finally, IP_TA_i is the ratio of IP over total assets at the end of each year of the selected period. In addition, I include the four latent factors identified in Isidro et al. (2016) as proxies for the institutional context. The first factor (FAC_1) comprises a mix of measures related to a country’s legal and governance systems, economic welfare and other social attributes; FAC_2 captures the development of capital markets (e.g. creditor and investor rights, securities regulation, capital market size and legal origin) and social characteristics associated with capital market development such as English proficiency; the third factor (FAC_3) encompasses variables related to the political process and financial and tax reporting system characteristics; finally FAC_4 incorporates the openness or closeness of society particularly in relation to external investment (Isidro et al., 2016).

In line with prior literature (Quagli and Avallone, 2010), I control for three variables I hypothesize can influence the choice of fair value. First, I control for firm activity (ACT_i), calculated as the ratio between total rents and total operating income at the end of each year, in order to discriminate companies whose core activity is renting out IP from those predominantly operating in trading of IP, services and other business (Quagli and Avallone, 2010). Secondly, I consider whether a company is a member of the European Public Real Estate Association (EPRA), including a dummy variable (EPRA_i) coded 1 if firm i is an EPRA member in year t and 0 otherwise (Quagli and Avallone, 2010). Given that EPRA’s Best Practices Committee supports the adoption of fair value to improve uniformity, comparability and transparency of financial reporting by real estate companies (EPRA, 2006), I expect a positive impact of being an EPRA member on the fair value choice.

Eventually, I include a dummy variable (REIT_i), assuming the value of 1 if the company is a Real Estate Investment Trust (REIT), and zero otherwise (Israeli, 2015).

I include in the model year fixed effects. All continuous variables are winsorized at the top and bottom 1%.

Table 2 presents the distribution of the sample by country and shows the number and proportion of companies that opt for either fair value or historical cost for IP in the years examined.

The data seem to display a different behavior regarding the choice of firms established in different nations. On the one hand firms located in the UK, Finland, Ireland and Sweden choose the fair value model for IP; contrastingly, firms from Italy and Spain seem to prefer the historical cost, whereas French, Greek and German companies show mixed evidence.

Table 3 reports the yearly distribution of the sample according to firm-years valuation model choice for investment property. The data show a consistently high prevalence of fair value adoption over the entire period, with the proportion of firms using fair value ranging between 80% and 87%. While the absolute number of fair value adopters increases slightly over time, the adoption rate remains relatively stable, suggesting that the majority of listed real estate firms consistently favor fair value accounting under IAS 40, though historical cost remains still adopted.

Overall, data in Tables 2 and 3 highlight the widespread use of fair value accounting for investment property among listed European real estate firms over time. Nonetheless, the country-level distribution indicates that the historical cost model continues to be broadly applied within certain institutional environments—such as Italy and Spain. These findings suggest that institutional factors at the country level may still play a role in shaping firms’ accounting choices under IAS 40 and shed some light on whether managers' preferences have evolved with increased experience in using fair value.

Table 4 reports the descriptive statistics for explanatory variables included in my model.

In order to investigate the different characteristics of companies that select the fair value model as compared to those that choose the cost model, I run an independent sample t-test.

Panel A of Table 5 presents the descriptive statistics on continuous explanatory variables for fair value and historical cost samples.

The mean size of firms adopting fair value is larger (13.56) than that of firms using the cost model (12.43). Similarly, the incidence of IP over total assets (IP_TA) is on average 74.8% for companies in the fair value sample whereas it is about 60% for firms in the cost sample. The average ratio between total rents and total operating income (ACT) is higher for firms adopting fair value (39.1%) as compared to firms adopting historical cost (1.3%).

Conversely, the mean amount of leverage (LEV) of firms in the historical cost sample, is only slightly higher (32.5%) than that of firms in the fair value sample (32.1%). Finally, the average ratio between market and book value (MTBV) of firms adopting historical cost is higher (1.646) than that of firms adopting the fair value model (0.976).

Table 5, Panel B reports the results of the independent samples t-test I run using the dummy variable FAIR to assess whether the differences between mean values of independent variables for firms in the two samples are statistically significant. As the table shows, the difference between mean values of SIZE, MTBV, IP_TA and ACT are all significant.

First, there is a positive relation between the size of the firm and the likelihood of selecting fair value. Firms opting for the fair value model are on average significantly (at 0.01% level) larger than firms using the historical cost.

In addition, I find a statistically significant difference (at the 0.01 level) in the ratio of market to book value (MTBV) between firms in the two samples. The analysis of mean values discloses a negative relationship between the market-to-book ratio and the choice of fair value (mean value of 0.98 for the fair value sample versus a mean value of 1.65 for the cost sample).

Moreover, there is a statistically significant difference (at 0.01 level) in the incidence of IP over total assets (IP_TA) between firms in the two samples. The mean values suggest a positive association between the relative weight of investment property on total assets and the choice of fair value, with an average of 0.75 in the fair value sample compared to 0.60 in the cost sample.

Furthermore, the two groups differ significantly (at 0.01% level) with respect to the firm’s activity (ACT). Specifically firms whose prevalent activity is IP rental seem to prefer valuing them at fair value, with respect to firms more involved in other businesses, like trading or services.

As a final point, differently from some previous research ^[11] I did not find a statistically significant difference between the two samples with respect to leverage (LEV) in the univariate analysis. Firms in the historical cost sample show a slightly higher mean leverage (32.5%) with respect to firms in the fair value sample (32.1%), but the difference is not significant.

Table 6 provides the Pearson correlation matrix for both dependent and controls variables. The table shows a relative weakness of the correlation coefficients for independent variables indicating that multicollinearity is not likely to be a significant issue in my sample ^[12].

Results of my regression model displayed in Eq. (1) are presented in Table 7. First, looking at the Pseudo R-squared (Table 7, Panel A) we observe that the model explains a significant portion (44.18%) of the difference in the choice to opt for fair value or cost. With reference to my independent variables, I notice that the company size (SIZE), the ratio of market to book value (MTBV), the incidence of investment property on total assets (IP_TA), the amount of firm’s debt (LEV) my descriptors of the institutional context (FAC_1, FAC_2, FAC_3, FAC_4) and, among control variables, membership of the European Public Real Estate Association (EPRA) are all significant.

More specifically, in line with existing literature (Christensen and Nikolaev, 2013; Olante and Lassini, 2022) the coefficient of IP on total assets (1.448; p-value: 0.002) shows a positive and significant (at 1% level) relationship with the dependent variable. The estimated exponential coefficient, or odds ratio of IP_TA (4.257) indicates that for a 10% increase in IP_TA, the odds of choosing fair value increase by 16%, therefore firms with a significant proportion of total assets represented by IP are far more likely to choose fair value than firms with a lower percentage. This finding confirms the hypothesis that the use of fair value is more common when it facilitates performance management and this is the case of firms whose primary activity is real estate, as realizing capital gains from trading of IP is part of the business model and thus fair value allows a better measure of periodic income than historical cost. Additionally, in the real estate industry IP are likely perceived to have more liquid markets, and therefore reliable fair value estimates are available at a lower cost, thus managers could be more prone to show them at market value in the balance sheet to further reduce information asymmetries both for trading activities (assets that will be sold in a short time) and for rental activities (assets that will typically be realized in a longer time) (Christensen and Nikolaev, 2013).

In addition, unlike some previous studies (e.g. Olante and Lassini, 2022; Quagli and Avallone, 2010), which generally document a lack of predictive power, my results show that the incidence of firm’s debt (LEV) has a strong and negative association with the choice of fair value (coefficient: −2.498; p-value: 0.000). In particular, the odds ratio (0.082) suggests that for each 10% increase in leverage, the odds of selecting fair value decrease by 22%. This result is in line with prior research on accounting choice (Watts, 2003; Qiang, 2007) and confirms that debt holders tend to prefer contracts based on historical cost figures (i.e. conservatism) as they are probably perceived to reduce agency costs and in fact to better protect their interests.

The company dimension (SIZE) has a positive relation with fair value accounting, significant at 5% level (coefficient: 0.160; p-value: 0.015). This finding is in contrast with previous studies, which generally report a negative relation between a company’s size and the choice of fair value (e.g. Quagli and Avallone, 2010; Christensen and Nikolaev, 2013; Olante and Lassini, 2022) and may be particularly significant, as it highlights a distinctive aspect of the real estate industry. In fact, if it is true that the uncertain verifiability of fair value, compared to cost numbers, in the presence of illiquid markets, could lead to increased litigation and associated costs (Watts, 2003), this is probably not the case for real estate. In fact, in the real estate industry markets for IP are usually more liquid than in other contexts, and therefore fair value estimates are perceived to be reliable. Hence, the use of fair value likely does not contribute to increase the perceived political costs.

Moreover, consistent with H3, the market to book value ratio (MTBV) has a weak negative relation with the fair value choice (coefficient: −0.190; p-value: 0.060). The variable’s odds ratio (0.827) indicates that for each 10% increase in MTBV, the odds of choosing fair value are reduced by approximately 2%. Therefore, the selection of fair value is associated with lower MTBV values. This finding aligns with prior research on both first-time (Quagli and Avallone, 2010; Israeli, 2015) and subsequent adoption of IAS 40 (Olante and Lassini, 2022). While I acknowledge that fair value accounting brings book values mechanically closer to market values, thus reducing the MTBV ratio, I believe this effect remains economically meaningful. In fact, by narrowing the gap between book and market values, fair value may contribute to mitigate information asymmetries, as financial statements become more aligned with market expectations—ultimately enhancing the relevance and reliability of reported financial numbers.

Regarding the effect of the institutional context, I notice that the four latent factors are all strongly significant in explaining the accounting choice for IP. In particular, FAC_1 1 and FAC_2 have a positive impact on the choice of fair value (coefficients: 4.996 and 3.738; p-values: 0.000 and 0.000, respectively). These findings suggest that country institutional characteristics, such as legal and governance systems, economic welfare (Factor 1) and country’s legal origin and the development of capital markets, together with social characteristics such as English proficiency and uncertainty avoidance, also associated with capital market development (Factor 2), positively affect the choice of fair value. That is, when the general institutional context is more developed (FAC_1), as well as are capital markets (FAC_2), the probability of choosing fair value increases.

On the other hand, FAC_3 and FAC_4 have a negative impact on the choice (coefficients: −9.326 and −2.427; p-values: 0.000 and 0.000, respectively). Factor 3 (FAC__3) captures attributes related to the political process (e.g. legislative competition and number of veto players), and some attributes of the financial and tax reporting system (e.g. book tax independence, tax compliance, and enforcement of accounting standards). As expected, lower levels of FAC_3 representing, for example, low legislative competition, high number of veto players, scarce book tax independence and tax compliance, negatively affect the probability of choosing fair value. Finally, the results suggest that societies that are relatively closer, particularly in relation to external investments (FAC__4) tend to prefer historical cost.

Overall, these findings contribute to deepening our understanding of the impact of the institutional context on accounting choice. Indeed, while previous studies (e.g. Christensen and Nikolaev, 2013; Olante and Lassini, 2022) usually employed dummy variables representing companies’ legal domicile as broad proxies for the institutional setting, the incorporation of the four latent factors included in my analysis offers a more refined perspective on the institutional elements affecting fair value selection for investment property in the real estate industry. In fact, by examining more detailed attributes of the institutional context, this approach allows a better understanding of specific dimensions affecting accounting choices within this industry.

With respect to control variables, the output shows a strong positive relation between the EPRA membership (EPRA) and fair value (coefficient: 0.986; p-value: 0.001). This finding confirms the results of prior literature (Quagli and Avallone, 2010), showing a positive impact of EPRA membership on fair value choice, at the IAS 40 first time adoption. This result suggests that the EPRA Best Practices Committee’s encouragement to adopt fair value in order to enhance uniformity, comparability and transparency of financial reporting, in fact continue to play a role in the development of financial reporting, producing incremental effects in the years following the first IAS 40 adoption. On the other hand, the firm’s main activity (ACT) as well as the status of Real Estate Investment Trust REIT do not show any predictive power (p-values: 0.115 and 0.142, respectively).

The aim of this study was to analyze managerial decisions between fair value and cost for IP under IAS 40 in the European real estate industry. Adopting a multi-motivation approach, I explored the factors driving this choice across different countries a decade after the first adoption of IAS 40 and up to recent years.

Following the framework of accounting choice theory, I hypothesized that managers might commit to fair value for several reasons, including contractual efficiency motives, asset pricing incentives, and institutional country factors.

I tested these hypotheses on a sample of publicly traded real estate firms from nine European countries with different pre-IFRS standards for IP. I conducted a logistic regression analysis to determine whether the choice of fair value can be explained by these motivations.

The results of the analysis show that, regarding contractual incentives, firm’s leverage is negatively associated with the fair value choice, suggesting that debt holders favor conservative accounting as a means to reduce agency costs through stronger lender protection (Watts, 2003; Qiang, 2007). Additionally, firms with a higher proportion of IP are more likely to adopt fair value supporting the hypothesis that the use of fair value is more common when it facilitates performance management as in the case of real estate firms, where realizing capital gains from trading of IP is part of the business model and thus fair value allows a more accurate measure of periodic income than historical cost.

Larger firms also tend to choose fair value, likely reflecting the higher liquidity of real estate markets and the perceived reliability of such valuations, which possibly mitigate shareholders’ litigation risk and their related costs. With respect to asset pricing motivations, the market-to-book value ratio, a proxy for information asymmetries, is negatively associated with fair value suggesting that fair value may help reduce such asymmetries by aligning book values more closely with market expectations.

Finally, institutional factors, such as legal origin, capital market development and political process characteristics significantly affect the fair value choice.

Overall, my results indicate that all proposed motivations contribute to the decision to choose fair value, predicting approximately 45% of the variance in this choice.

This study offers several contributions to existing literature. First, it integrates indicators from diverse strands of prior research to develop a comprehensive multi-motivation model offering a deeper understanding of the complexities behind managerial accounting decisions for investment property (IP). This is particularly relevant for the real estate industry, where the valuation of IP plays a critical role in financial reporting and decision-making. Furthermore, while previous studies have focused either on the traditional comparison between the UK and Germany (Christensen and Nikolaev, 2013) or on selections of European countries that mandated only the cost method before IFRS adoption (Quagli and Avallone, 2010; Israeli, 2015), this study broadens the scope of analysis, by including nine European countries that adopted either the fair value or cost model in their pre-IFRS national GAAP, and therefore provides a more comprehensive examination of the relationship between institutional context and accounting choice. Third, this study extends the scope of prior research on accounting choices for investment property. By focusing on the post-IAS 40 adoption period, it provides fresh insights into managerial behavior in recent years, filling a gap in the literature, as no prior research has specifically examined the actual choices of real estate firms in an updated context.

My findings could be of interest for standard setters as well as for financial statement users, as they suggest that whereas many companies in fact recognize IP at fair value in the real estate industry, some firm-specific characteristics still significantly influence the choice, raising concerns about its general acceptance and, indicating that experience with fair value may not be sufficient to ensure its widespread use for IP recognition. In addition, the study highlights the critical role of the institutional development—such as strengthening legal frameworks and improving capital markets—in promoting fair value adoption. In fact, the observed negative impact on fair value choice of certain institutional characteristics (e.g. weaker governance or tax compliance systems) highlights the need for regulatory interventions to overcome these specific challenges.

While my study possibly enhances the understanding of the complexity of managerial decision-making, it is subject to some limitations and raises some issues for future research.

First, as suggested by the analysis, the institutional context plays a significant role in accounting choice, thus further investigation into country-specific institutional factors could provide a deeper comprehension of firms’ behavior. Moreover, analyzing a larger sample size across more industries could lead to enhanced results. Finally, whereas I conduct my study on real estate firms located in nine European countries, expanding the study to firms in other IFRS-adopting countries would help achieve more generalizable outcomes.