Christensen and Nikolaev 2011 Does Fair Value Accounting for Non-financial Assets Pass the Market Test

7/24/2019 Christensen and Nikolaev 2011 Does Fair Value Accounting for Non-financial Assets Pass the Market Test

1/47

Does fair value accounting for non-financial assets

pass the market test?

Hans B. Christensen and Valeri V. Nikolaev

The University of Chicago

Booth School of Business

5807 South Woodlawn AvenueChicago, IL 60637

Abstract: We study managers revealed preferences for fair value or historical cost accountingfor non-financial assets when market forces, rather than regulators, determine the choice. We

document that almost all managers pre-commit to historical cost accounting for plant, equipment,and intangible assets, suggesting that fair value for illiquid non-financial assets is associated with

net firm-specific costs. However, for the more liquid assets groups, property and investment

property, the observed choices suggest that fair value is often associated with net benefits.Indeed, the majority of real estate companies choose fair value over historical cost for investment

property. We also find that fair value use is positively associated with reliance on debt financing.

However, unlike prior studies, we conclude that this result is likely attributable to lowerincremental costs of fair value reporting rather than an attempt to avoid covenant violations. Our

findings contribute to the policy debate over fair value accounting for non-financial assets by

documenting that the firm-specific cost of establishing reliable fair value estimates represents abarrier for fair value to become the primary valuation method on a voluntary basis.

Keywords: Fair value, IFRS, non-financial assets, illiquid assets.JEL Classification: M4, M41

This paper previously circulated under the title: "Who uses fair-value accounting for non-financial assets after

IFRS adoption?". This research was funded in part by the Initiative on Global Markets at the University of Chicago

Booth School of Business. We benefited from helpful comments from Ray Ball, Philip Berger, Alexander Bleck,

Christof Beuselinck, Johan van Helleman, S.P. Kothari, Laurence van Lent, Christian Leuz, Paul Madsen, Karl

Muller, Edward Riedl, Douglas Skinner, Abbie Smith, Ross Watts, Li Zhang, and workshop participants at the EAA

2009 Annual Meeting, University of Chicago, University of North Carolinas GIA Conference, Harvard

Universitys IMO Conference, ISCTE, and Tilburg University. Michelle Grise, SaeHanSol Kim, Shannon Kirwin,

Ilona Ori, Russell Ruch, and Onur Surgit provided excellent research assistance.


2/47

1

1. Introduction

Academics, standard setters, and practitioners actively debate the use of fair value

accounting for illiquid assets (e.g., Schipper 2005a, 2005b; Ball 2006; Watts 2006; Herrmann et

al. 2006; Kothari et al. 2009; Hail et al. 2009; Laux and Leuz 2009). The debate dates back to the

1930s (e.g., Paton 1932, pp. 739-743) and has recently been fuelled by the financial crisis and

the potential adoption of International Financial Reporting Standards (IFRS) in the United States.

We contribute evidence to this debate by studying managers preferences, revealed by their

choice between fair value and historical cost accounting for non-financial assets when market

forces, rather than regulators, determine the outcome. Examining managers actual choices is

useful in light of the policy debate as it reveals whether the firm-specific benefits of fair value

accounting exceed the firm-specific costs. However, we do not evaluate the net social costs or

benefits of fair value accounting due to potential externalities, which may be relevant to

regulators decisions. While we are not the first study that examines the choice between fair

value vs. historical cost (Brown et al. 1992; Whittred and Chan 1992; Cotter and Zimmer 1995),

our setting is likely to be more revealing about the costs and benefits of fair value accounting

than prior settings because of the requirement to pre-commit to one valuation practice, as

discussed next.

We exploit the recent IFRS adoption in the European Union (EU) and focus on major and

arguably the most controversial non-financial asset groups: (i) property, plant, and equipment

(PPE), (ii) investment property, and (iii) intangible assets.1 IFRS requires that companies state

their valuation method in the accounting policy section of their annual reports and apply the

1 In this paper, we use the term asset group to describe the three types of assets we examine. Intangible assets,

investment property, and property, plant, and equipment each constitute one asset group. We use the term asset class

to describe a subsection of an asset group. For instance, property constitutes an asset class under the asset group

property, plant, and equipment. Our definition of an asset class is consistent with IAS 16.37.


3/47

2

chosen method consistently over time. This implies that companies need to pre-commit ex ante

to one valuation practice, which differs from prior studies where managers had discretion to

revalue selectively (from time to time). Pre-commitment is likely to reduce incentives for

opportunistic use of fair value and, in turn, allows us to focus on more fundamental economic

trade-offs. Out of the 28 EU countries adopting IFRS in 2005, we select the United Kingdom

(UK) and Germany because they have the largest financial markets in the EU and, historically,

are at the opposite ends of the spectrum in terms of using fair value accounting under local

GAAP. Specifically, for non-financial assets, German GAAP allows only historical cost

accounting, whereas UK GAAP either allows (for PPE) or mandates (for investment property)

fair value accounting. IFRS expands the available valuation practices in both the UK and

Germany. Indeed, under IFRS both fair value and historical cost are allowed for each of the three

asset groups, which enables managers to reveal their preferences.2

Throughout this paper, we assume that managers ex antechoices of valuation practices

primarily reflect capital markets reporting demands. From a capital markets viewpoint, the

choice between historical cost and fair value involves a cost-benefit tradeoff. On the benefit side,

fair value represents more relevant information used by investors in their capital allocation

decisions (e.g., Barth and Clinch 1998; Schipper 2005a; Herrmann et al. 2006). On the cost side,

construction of reliable fair value estimates is expensive due to their inherent lack of verifiability

(Watts 2006). Subjectivity in non-verifiable fair value estimates can be exploited

opportunistically to manipulate reported performance, and, therefore, translates into agency costs

borne by shareholders (Jensen and Meckling 1976). To alleviate these agency costs managers

need to pre-commit to accounting methods that reduce accounting discretion (Watts and

2We use the term historical cost to describe accounting treatment under which assets are recognized at historical

cost less subsequent depreciation (amortization) and/or impairments.


4/47

3

Zimmerman 1979, 1986). For example, choosing historical cost serves as a commitment against

discretionary upward asset revaluations. This tradeoff between relevance and reliability lies at

the center of the controversy over fair value accounting (e.g., Sloan 1999; Wahlen et al. 2000;

Dietrich et al. 2001; Schipper 2005a; Herrmann et al. 2006; Laux and Leuz 2009) and hence we

argue that this tradeoff should largely determine the choice between the two accounting

principles that we study.

We make three empirical predictions. First, the recent move towards fair value in

accounting standards (e.g., Johnson 2005) suggests that standard setters believe the efficient

solution to asset measurement has shifted towards the relevance side of the tradeoff (Watts and

Zimmerman 1986). In other words, the relevance benefits of fair value are on average expected

to outweigh the cost of lower reliability. Hence, we predict that IFRS adoption is associated with

a significant shift towards fair value accounting for non-financial assets among firms that were

constrained to historical cost accounting under local GAAP. Costs of constructing reliable fair

value estimates, however, are expected to be an important cross-sectional determinant behind the

choice to adopt fair value. Given this, our second prediction is that fair value accounting is more

likely for assets that exhibit relatively more liquid markets (which serves as a source of

verification and hence reduces the cost of establishing reliable fair value estimates). Property is

more likely than other non-financial asset classes to be re-deployable by other firms and

therefore has relatively liquid markets (Shleifer and Vishny 1992). Hence we expect a more

frequent use of fair value for property. Our third prediction is that fair value accounting is

positively associated with reliance on debt financing. Companies that frequently and more

heavily rely on debt are commonly required by creditors to invest in construction of reliable fair

value estimates for the purposes of debt contracting and reporting to creditors. Given this, the


5/47

4

marginal cost of recognizing these estimates in financial statements is low (Holthausen and

Watts 2001).

Another distinctive feature of our study is that we avoid drawing conclusions based on a

particular industry or asset group. While prior studies mainly focus on a selected sample (e.g.,

investment property), preferences towards fair value use are likely to vary considerably across

assets and industries, and understanding this variation can be useful from the policy perspective.

We examine this variation to establish for which assets and in which industries fair value is

associated with net firm-specific benefits. Our hand-collected sample consists of 1,539

companies, which approximates the population of publicly traded companies in Germany and the

UK. We read the accounting policy sections in annual reports to identify the valuation practices.

We find that only 3% of the sample firms use fair value accounting for at least one asset

class under the PPE asset group following IFRS adoption. With very few exceptions, these

companies use fair value accounting for the property asset class only; members of the plant and

equipment asset classes are valued at historical cost in almost all cases.3An even more striking

observation emerges when we examine the post-IFRS choices of companies that recognized at

least one PPE asset class at fair value under local GAAP (i.e., under UK GAAP). We find that

44% of these companies in fact switch to historical cost accounting upon IFRS adoption. In

contrast, among companies that recognized all PPE asset classes at historical cost under local

GAAP, only 1% switch to fair value for at least one asset class.

3Total assets and shareholders equity are, respectively, 31% and 88% higher on average for companies that apply

fair value than for a matched sample of companies that use only historical cost accounting. This suggests that the

choice between the valuation methods is not random and is economically important. This result cannot be

interpreted as causal, however, because incentives to use fair value depend on how using fair value accounting

versus historical cost accounting affects the outcome. See appendix C for the results.


6/47

5

We find that companies are equally likely to use historical cost and fair value accounting

for investment property (i.e., property held for the purpose of earning rental income or for capital

appreciation). The strongest determinant of fair value use for this asset group is whether real

estate is one of the company's primary business activities. Namely, German companies, all of

which were constrained to historical cost before IFRS adoption, are significantly more likely to

switch to fair value accounting for investment property when real estate is among their primary

activities. In contrast, in the UK, where all companies were constrained to fair value for

investment property prior to IFRS, we observe more frequent switches to historical cost

accounting when real estate is not among their primary activities. The results are consistent with

managers revealing their preferences and switching accounting treatments once their choice is

not constrained by accounting regulation. The preferences of real estate companies managers in

the UK and Germany indicate that fair value accounting for investment property in this industry

is generally associated with net firm-specific benefits, consistent with Muller et al. (2008). Since

the real estate industry exhibits relatively liquid markets, the costs of constructing reliable fair

value estimates are lower. In addition, changes in the value of investment property are directly

linked to the performance of real estate business. This implies that benefits of fair value are more

likely to outweigh its costs for real estate companies. Finally, we find that nocompanies in our

sample use fair value accounting for intangible assets.

Logistic regression analysis of the decisions to use fair value reveals for both investment

property and PPE that reliance on debt financing is positively associated with the use of fair

value. This finding holds both when measuring reliance on debt by leverage and the frequency of

accessing debt markets. Further analysis reveals that short-term debt is more important than

long-term debt in explaining the fair value use. Given that we study pre-commitments to fair


7/47

6

value, and that accounting-based covenants are less common for short-term debt, the results are

inconsistent with the conclusion that companies use fair value opportunistically to avoid

covenant violations (opportunism is one proposed explanation for the results in prior literature,

see Cotter and Zimmer 1999 for discussion).

Overall, our results do not support our first prediction that a significant fraction of firms

shift towards fair value accounting upon IFRS adoption. Rather, the evidence indicates that

almost all firms pre-commit against upward asset revaluations for the majority of illiquid (non-

financial) asset groups and therefore stands in contrast to the standard setters' enthusiasm for fair

value accounting. The cross-sectional tests of our second and third predictions suggest that fair

value is used when the costs of obtaining reliable estimates are relatively low. Hence, the

resistance to fair value does not appear to be due to managers disagreement with standard setters

on the conceptual merits of fair values in decision making, but rather due to the costs of

establishing reliable fair value estimates.

Our paper makes three contributions to the literature. First, it documents that, in a setting

where companies must adhere to their choice in the future, fair value accounting for non-

financial assets is crowded out by historical cost accounting, arguably with the exception of

property. This finding implies that despite the conceptual appeal of fair value, the costs of

establishing reliable estimates prevent fair value from becoming the primary valuation method

for non-financial assets. In light of the policy debate over whether to expand mandatory fair

value accounting for non-financial assets, our evidence on managers actual choices cautions that

the firm-specific benefits are unlikely to exceed the firm-specific costs. However, mandatory fair

value accounting for illiquid non-financial assets may still be socially optimal if fair value


8/47

7

accounting is associated with positive externalities. More research is needed to understand

potential externalities.

Second, this is the first study to exploit revealed preferences for pre-commitment to fair

value accounting. Our results contrast sharply to the prior evidence on ex post revaluations. For

example, Brown et al. (1992) report that about two-thirds of Australian companies revalued

assets upwards (at least once) over three- and four-year periods. Similarly, studies on intangibles

such as brand names (e.g., Barth and Clinch 1998; Muller 1999) report relatively frequent ex

post revaluations. We, however, find that few companies pre-commit to the use of fair value for

PPE and no sample firms make such a pre-commitment for intangibles. These differences

suggest that the requirement to pre-commit affects managers choices and also highlights the

difference between our setting and the settings studied in prior literature.

Third, our study contributes to the policy-related debate on the consequences of adopting

IFRS in the US (e.g., Hail et al. 2009). Under US GAAP, fair value accounting is not allowed for

non-financial assets, and this difference between US GAAP and IFRS amounts to an important

point of disagreement among the standard setters. It is often argued that a consequence of

mandatory IFRS adoption is a significant increase in the use of fair value (see Cairns 2006 for

examples). Our evidence generally does not support this conjecture for non-financial assets. One

potential scenario is that fair value accounting becomes a popular and widely used vehicle for

misrepresentation of financial reports. For example, recent evidence in Ramanna and Watts

(2009) suggests that mandatory application of fair value to assess goodwill impairments is

consistent with opportunistic representation of financial reports. In our voluntary setting,

however, most companies choose to pre-commit against the use of fair value for non-financial

assets. This suggests that most managers are unwilling to subject shareholders to the agency


9/47

8

costs associated with fair value measurement for non-financial assets. Interestingly, our finding

echoes accounting practice in the United States before the Securities and Exchange Commission

(SEC) banned upward revaluations in 1940: downward revaluations were much more common

than upward revaluations and the latter were almost never performed on intangible assets

(Fabricant 1936; Paton 1932). The consistency of accounting practice across time and different

institutional settings speaks to the existence of an economic mechanism that governs companies

choice to use fair value.

Section 2 describes the accounting traditions in the UK and Germany, as well as the

valuation methods available to companies under German GAAP, UK-GAAP, and IFRS; Section

3 develops testable hypotheses; Section 4 describes the sample selection procedure and presents

our results; and Section 5 concludes.

2. Accounting in the UK and Germany

Despite EU accounting harmonization that began decades prior to IFRS, the UK and

Germany arguably had the two most distinct asset valuation traditions in Europe at the time of

IFRS adoption. The differences in their accounting traditions are due to institutional differences

in legal systems and ownership structures. Germany has traditionally been characterized by the

existence of private companies who raise capital from banks and communicate via private

information channels (Leuz and Wstemann 2004). Accounting was mainly used to establish

taxable income; hence, accounting regulation was codified and focused mainly on legal entity

statements. As revaluations are often in conflict with the objectives of tax authorities, German

GAAP only allowed historical cost accounting. Today in Germany, there is no formal link

between legal entity reports, which are still used primarily for tax purposes, and consolidated

statements. Thus, companies financial reporting choices in the consolidated statements,


10/47

9

including their valuations of non-financial assets, have no tax consequences. In contrast, UK

accounting has historically developed separately from tax accounting and in the private sector

rather than in company law. UK ownership is dispersed and even middle-sized companies are

commonly listed on the London Stock Exchange. Such ownership structure requires that

financial reporting reduce information asymmetry, and in this context revaluations may serve the

purpose of conveying information about assets' current values.

Next, we discuss the accounting treatment of long-term non-financial assetsinvestment

property, PPE, and intangible assetsunder UK-GAAP, German GAAP, and IFRS.

2.1 Accounting for investment property

IAS 40 defines investment property as land or buildings not currently occupied by the

owner that are held to generate rental income or capital appreciation. Under German GAAP,

companies must value investment property at historical cost, while under UK-GAAP, companies

are required to use fair value. Net income is unaffected by upward revaluations of this asset

group under UK-GAAP, as they are credited to the revaluation reserve. IFRS offers companies

the choice between recognizing investment property at historical cost or fair value. Thus a

significant shift towards fair value is expected among German companies while not for UK

firms. Under IFRS, if a company chooses to recognize investment property at historical cost, it

must systematically depreciate the acquisition costs and disclose the investment property's fair

value in the notes accompanying the financial statements. In contrast, if a company chooses to

apply fair value, changes in the investment property's value become part of operating income and

the assets are not subject to depreciation. We assume that investors are rational and cannot be

misled by whether fair value changes affect net income or directly go to shareholders equity.

Thus we rule out this consideration as a determinant of the fair value choice.


11/47

10

2.2 Accounting for property, plant, and equipment (PPE)

The only valuation method for PPE permitted under German GAAP is historical cost.

Under both IFRS and UK-GAAP, PPE is initially recognized at cost but at each subsequent

balance sheet date is valued at either historical cost or fair value. In either case, these assets are

subject to depreciation. When fair value is applied, positive changes in an asset's value are

credited to the revaluation reserve, which constitutes part of shareholders equity. Revaluations,

therefore, only affect net income through future depreciation charges (unlike for investment

property). Finally, under IFRS, the choice of valuation method must be consistent for all assets in

the same asset class (IAS16.29).

2.3 Accounting for intangible assets

Under German GAAP, historical cost is the only valuation method permitted for

intangible assets. Under both UK-GAAP and IFRS, however, intangible assets are to be carried

at either historical cost or fair value less any amortization and impairment charges. Under fair

value, the accounting treatment is similar to that of PPE; that said, a company may only apply

fair value to an intangible asset if an active market exists for that asset (IAS38.75). The

definition of an active market is very narrow, and for many intangible assets, such as brands,

patents, and trademarks, it is nonexistent, due to their uniqueness and the specificity of their

application (IAS38.78).

2.4 Fair value under IFRS vs. settings in prior literature

The IFRS setting is different from the Australian and UK settings used in prior research

in that companies must ex antestate their choice between historical cost and fair value in their

accounting policy.4,5

If a company decides to apply fair value under IFRS, it must revalue assets

4Note that both the UK and Australia adopted accounting standards in 1999 and 2000 that are similar to IAS 16,

however, the prior literature we refer to rely on data before this change.


12/47

11

every timethe book value is materially different from the market value (IAS 16 and IAS 40). If

the same company instead decides to use historical cost, it cannotperform upward revaluations

in the future. A switch between historical cost and fair value is considered a voluntary change in

accounting principles and needs to be justified to auditors, lenders, equity investors, and

potentially to regulators. Therefore, the choice between fair value and historical cost in our

setting represents an ex ante commitment and hence is unlikely to be driven by instantaneous

earnings management considerations. Indeed, the early studies argued that discretionary

revaluations are related to contracting motives consistent with this view leveraged companies

in danger of violating covenants are more likely to revalue assets (Whittred and Chan 1992;

Brown et al. 1992; Cotter and Zimmer 1995).6

The problem with discretionary revaluations is that managers decide whether to revalue

assets ex postafter they know the effect of the fair value estimate on the financial statements. For

instance, managers may only revalue assets when they need to manipulate reported performance.

Alternatively, managers may revalue assets when reliable fair value estimates are available. Our

setting isolates this issue as we examine ex antechoices to use fair value with limited ex post

discretion to change valuation methods. Thus, examining ex ante choices is more likely to be

informative about the economic trade-off between fair value and historical cost than examining

ex postrevaluations. Furthermore, the ex anterequirement in IFRS aids firms in committing to

non-opportunistic use of fair value accounting, which may imply that fair value accounting under

IFRS is associated with greater benefits than ex post revaluations. Consistent with this view,

5Analogous to, e.g., inventory valuation methods.

6 Whittred and Chan argue that asset revaluations reduce underinvestment problems that arise from contractual

restrictions, while Cotter and Zimmer argue that upward revaluations increase borrowing capacity. While debt

contracting is the main explanation for asset revaluations, Brown et al. also find that bonus contracts, as well as

signaling and political cost explanations, play an important role.


13/47

12

Muller et al. (2008) find lower bid-ask spreads for fair value companies and Cairns et al. (2008)

find that fair value accounting increases international comparability in the IFRS setting.

3. Background and hypotheses

Relevance and reliability are the most basic and important attributes of accounting

information. The tradeoff between relevance and reliability is recognized in the conceptual

frameworks of both the Financial Accounting Standards Board (FASB) and the International

Accounting Standards Board (IASB).7 Thus, when standard setters choose between fair value

and historical cost, the relevance-reliability tradeoff is at play. In recent years, both FASB and

IASB have placed more emphasis on relevance, not reliability.

8

This change in focus is reflected,

for example, in the conclusion of a discussion of relevance and reliability by L. Todd Johnson, a

senior project manager at FASB:

The Board has required greater use of fair value measurements in financial statementsbecause it perceives that information as more relevant to investors and creditors than

historical cost information. Such measures better reflect the present financial state of

reporting entities and better facilitate assessing their past performance and futureprospects. In that regard, the Board does not accept the view that reliability should

outweigh relevance for financial statement measures. (Johnson 2005).

FASB's and IASB's shift towards fair value accounting suggests that they believe the

efficient solution to asset measurement has shifted closer to the relevance side of the tradeoff

(Watts and Zimmerman 1986). The standard setters position is justified on the grounds that fair

value measurement aids financial statement users' decision making. Fair value is also argued to

improve transparency, comparability, the timeliness of accounting information, and relative

performance measurement (e.g., Schipper 2005a). In line with benefits of fair value, a large

stream of value relevance studies on asset revaluations indeed finds that fair value possesses

7See IASB's Framework paragraph 45 and FASB's Conceptual Statement 2 paragraph 15.

8See for example IASB Discussion Paper July 2006, paragraph BC2.62.


14/47

13

superior relevance. These studies find that upward revaluations are positively associated with

equity returns in the month of the revaluation (Sharpe and Walker 1975; Standish and Ung 1982)

and that they are associated with longer period stock returns, future cash flows, and the market

value of equity (e.g., Easton et al. 1993; Barth and Clinch 1998; Aboody et al. 1999; Danbolt and

Rees 2008).

Barth et al. (2001) argue that the value relevance tests are joint tests of relevance and

reliability, because a certain degree of reliability is also established by rejecting the null of no

association. This conclusion increases confidence in fair value and (combined with the argument

that the reliability of fair values is similar to that of other accounting estimates, e.g., allowance

for uncollectible accounts) potentially influences the regulators views. A caveat that is in order

is that the previously documented associations between revaluations and equity values are

largely conditional on a companys discretionary choice to revalue assets. These choices are

non-random as, for example, managers may revalue assets because they anticipate a markets

response to the reported numbers or know how reliable the revaluations are. Alternatively, as the

association is measured over a relatively long window, it could be that managers choose to

revalue when the firm does well, hence has higher returns. Thus, the association could capture

the underlying change in firm performance and hence does not establish that the markets trust the

re-valued numbers. These possibilities potentially limit the generalizability of value relevance

tests for companies that do not use fair value and underscore the need to study the choice

between fair value and historical cost.

IFRS offers an opportunity to test whether the move toward fair value in accounting

standards is supported by the accounting practice choices that managers are making. Managers

have incentives to make ex antevaluation choices that reflect the interests of firms stakeholders


15/47

14

(otherwise stakeholders will price protect or impose costs on management), and hence are cost-

justified for a particular company. The observed choices can be used to infer whether managers

agree with IASB (and FASB) that the efficient solution to the relevance-reliability trade-off has

indeed shifted towards relevance. If the firm-specific benefits of fair value exceed its costs we

hypothesize:

H1: IFRS adoption is associated with a shift towards fair value accounting for non-financial

assets.

The effort and resources a company needs to expend in order to obtain reliable fair value

estimates are likely important in determining a managers choice of valuation practice. We next

consider how the costs of obtaining reliable estimates affect the choice between fair value and

historical cost. The ability to obtain reliable fair value estimates is closely related to the existence

of liquid markets for assets, which provide an independent source of verification (Watts 2006).

Property is the only non-financial asset class for which a relatively liquid market with official

statistics exists. Therefore, our second hypothesis proposes:

H2: Fair value accounting is more likely for asset classes for which liquid markets exist, i.e., for

property as opposed to plant, equipment, and intangibles.

Similarly, the existence of reliable fair value estimates for purposes other than financial

reporting affects the marginal cost of recognizing fair values in financial statements. Companies

that access debt markets are commonly required under their credit arrangements to provide

valuations of collateral. The fact that lenders are willing to lend against these valuations implies

that a company invests in measuring them reliably (e.g., independent valuation and

certification).9Given this, recognizing the fair values of these assets in financial statements is

9 Muller and Riedl (2002) document that fair value estimates produced by independent valuators are viewed by

capital markets as being more reliable than fair value estimates produced by managers.


16/47

15

associated with low incremental costs (Holthausen and Watts 2001). Therefore, our third

hypothesis is:

H3: Fair value accounting is positively associated with reliance on debt financing.

Earlier studies have documented a positive correlation between ex postrevaluations and

leverage (Brown et al. 1992; Whittred and Chan 1992; Cotter and Zimmer 1995). However, this

positive correlation is often attributed to opportunistic behaviour where managers perform

upward revaluation to avoid covenant violations when they move closer to technical default (see

discussion in Cotter and Zimmer 1999). If this hypothesis is indeed true, then we do not expect

to find a significant association between leverage and the use of fair value since avoiding default

on covenants is unlikely to drive managerial choice (i.e., pre-commitments) in our setting.

Furthermore, if opportunism was to explain the association between leverage and the use of fair

value, we would expect to find the association stronger in case of long-term debt, where

accounting-based covenants are more common compared to short-term debt. We test this

implication empirically.

4. Results

4.1 Sample selection and descriptive statistics

We manually verify the accounting standards that a given company follows by looking at

either the accounting policy section or the auditors opinion section of its annual report(s). To

identify the asset valuation practice a company follows, we read the accounting policy section of

its annual report(s). We begin with all UK and German companies (active and inactive) in

Worldscope and further restrict to those complying with IFRS in either 2005 or 2006. For

inclusion in the Germanand UKcross-sectional samples, we further require that a company has

an annual report under IFRS in Thomson One Banker. We construct a cross-sectional sampleto


17/47

16

examine valuation practices after mandatory IFRS adoption and a switch sample (UK only) to

examine whether companies use mandatory IFRS adoption to switch their accounting practices.

For inclusion in the UK switch sample, we also require that a company has an annual report

(prepared according to UK-GAAP) before IFRS adoption.10

Table 1, Panels A and B present the distribution by industry of companies in Worldscope

as well as in the German sample, the UK cross-sectional sample, and the UK switch sample. The

industry distribution in each sub-sample approximates that of Worldscope.

4.2 Valuation practices

In this section, we provide evidence on hypothesis H1. A company is classified as

applying fair value accounting if it recognizes at least one asset class (within an asset group) at

fair value.11

Similarly, a company is classified as applying historical cost if it recognizes at least

one asset class (within an asset group) at historical cost. Appendix A presents examples of fair

value accounting and historical cost accounting for PPE.12

4.2.1 Valuation practices in the UK

Table 2 documents the valuation practices in the UK cross-sectional sample. We identify

no use of fair value accounting for intangible assets; instead, all companies in our sample rely on

historical cost for this asset group. For PPE, 5% of companies use fair value accounting while all

10For companies both in Germany and the UK, we obtain their first annual report under mandatory IFRS, which is

typically for fiscal year 2005. In addition, for companies in the UK, we look for their last UK-GAAP annual report,

which is typically for fiscal year 2004. In the rare cases where we cannot find these annual reports, we take the nextannual report available in Thomson One Banker (e.g., for fiscal year 2006).

11Note that this is a conservative way of defining the use of fair value because the fraction of assets recognized at

fair value can be relatively small. Yet, even this definition generates pronounced economic differences across the

two groups of companies.

12Panel A of Appendix A provides an example of a company that switched from fair value to historical cost, Panel

B provides an example of a company that used fair value under both UK-GAAP and IFRS, and Panel C provides an

example of a German company that uses fair value.


18/47

17

companies use historical cost for at least one asset class within this asset group. We observe that

fair value use differs across industries, with higher concentration in the financial sector.

Table 3 presents the results from the UK switch sample. For PPE, we find that 6% of

companies use fair value under UK-GAAP and 5% use fair value under IFRS. A large number of

switches occur for this asset group. Specifically, 44% of companies that use fair value for at least

one asset class in PPE under UK-GAAP switch to historical cost (for all asset classes) upon IFRS

adoption. In contrast, only 1% of companies using historical cost for all asset classes under UK-

GAAP switch to fair value for at least one asset class upon IFRS adoption. The joint evidence in

both tables does not support H1 and implies that only a small number of companies find fair

value to be cost-justified.

What made a high portion of UK companies switch to historical cost on IFRS adoption?

IFRS and UK-GAAP are very similar when it comes to the valuation of PPE (see Section 2.2).

Thus, the switches observed upon IFRS adoption are voluntary in the sense that IFRS did not

force these companies to switch to historical cost. If historical cost maximizes net benefits, why

did these firms not switch to historical cost under UK GAAP? 13 One explanation is that

switching accounting principles is uncommon in practice because it is costly.14

The costs of

switching accounting principles include renegotiating contracts that require consistency in

GAAP, convincing auditors that the new practice better reflects the underlying economics of the

company, and communicating the change to financial statement users. Most of these costs are

13

We contacted those non-financial companies that switched to historical cost and received several repliesindicating that IFRS was a convenient opportunity for them to make the switch.

14Consistency in accounting policies across time is highly regarded by the accounting profession. Comparability is a

qualitative characteristic expressed in IASBs Framework (paragraph 39): . . . the measurement and display of the

financial effect of like transactions and other events must be carried out in a consistent way throughout an entity and

over time for that entity. In U.S. literature, consistency is expressed in several places, including the Accounting

Research Study No. 1 of the American Institute of Certified Public Accountants (postulate C-3). See Ball (1972) for

an extensive discussion of the accounting professions reliance on consistency.


19/47

18

fixed (i.e., they are independent of the number of changes) so the incremental cost of voluntary

changes is substantially lower when combined with a mandatory change such as IFRS adoption.

Even if these companies did want to switch to historical cost before IFRS adoption, the

associated costs could have made switching unattractive. However, the observation that a switch

is more likely from fair value to historical cost than from historical cost to fair value implies that

firms revealed preferences do not echo with the standard setters' enthusiasm over fair value

accounting, in contrast to hypothesis H1.

After IFRS adoption, fair value is more common for investment property, for which UK

companies had to use fair value under local GAAP, than it is for PPE. Nevertheless, 23% of

companies reveal preferences for historical cost by switching from fair value to historical cost

once they are no longer constrained to the use of fair value by accounting regulation. Significant

industry variation is present: whereas only 2% of financial companies switch to historical cost,

45% of non-financial companies switch.

4.2.2 Valuation practices in Germany

Table 4 documents the valuation practices in the German sample. We find no use of fair

value accounting for intangible assets in Germany, similar to the UK. For PPE, 1% of companies

switch to fair value for at least one asset class upon IFRS adoption (note that under German

GAAP fair value was not allowed). Only one company applies fair value to all asset classes in

PPE, while all other companies use historical cost for at least one asset class. These findings

approximate those we observed in the UK and indicate that most managers reveal preferences for

historical cost.

For investment property, we find that 23% of German companies reveal preferences for

fair value by switching from historical cost to fair value once they are no longer constrained to


20/47

19

historical cost by accounting regulation. However, we also observe substantial industry variation.

Among financial companies, 49% switch to fair value, while only 6% of non-financial

companies switch.

In summary, we find that a small number of companies use fair value accounting for at

least one asset class under PPE after IFRS adoption. The absence of fair value accounting for

intangibles and its limited use for PPE in both the UK and Germany suggests that only a small

subset of companies perceive net firm-specific benefits of fair value accounting. Counter to

hypothesis H1, there is no general shift towards fair value accounting for non-financial assets.

The exception is investment property among German financial institutions where approximately

one half of them shift to fair value. The evidence that shifts towards fair value occur for

investment property rather than PPE and intangibles is consistent with H2. Next, we provide

further evidence on H2 by exploiting the differences in asset liquidity within the PPE group.

4.2.3 Assets recognized at fair value

In this section, we examine which asset classes in PPE are recognized at fair value. Table

5 presents the distribution of fair value use across the three asset classes within the asset group.

Sixty-nine companies in the sample use fair value accounting either before mandatory IFRS

adoption, after mandatory IFRS adoption, or both. Of these companies, 93% use fair value

accounting for property. Only 3% use fair value for plant, and only 4% use fair value for several

asset classes in PPE. The distributions of fair value use in the UK and Germany are similar.

Hence, the application of fair value accounting is, in practice, not only limited in terms of the

number of companies using it, but also in terms of the assets to which it is applied, namely

property. While this evidence is at odds with hypothesis H1, it supports H2 as property is the

only non-financial asset class for which a relatively liquid market is often present.


21/47

20

4.3 Analysis of the decision to use fair value

In this section, we examine the revealed preferences for valuation practices following

IFRS adoption in a cross-section of firms using logistic regression analysis. Our analysis draws

on two different subsamples and controls for common company-specific characteristics. First, we

analyse the sample of all companies that hold investment property. Second, we restrict our

analysis to the sample of companies that use fair value for PPE matched with a historical cost

control group. The summary statistics for variables used in this analysis are reported in Table 6.

All variables are defined in Appendix B. Because the number of observations and the set of

explanatory variables vary across the two subsamples, we report two separate sets of summary

statistics in Panels A and B.

4.3.1 Investment property

While IFRS provides UK companies with the first opportunity to switch to historical cost

for investment property, in Germany the opposite is the case. Thus, IFRS gives both German and

UK company managers an option to move to the asset valuation method not previously available

under local GAAP in these countries. Such a setting is particularly convenient to study revealed

preferences because significant switching from historical cost to fair value in Germany and, in

contrast, from fair value to historical cost in the UK is difficult to attribute to factors other than

the presence of considerable firm-specific benefits associated with the alternative accounting

treatment. We begin by exploring whether firm-specific cost and firms-specific benefits of fair

value accounting differ across industries. Note that in this regard the real estate industry is

unique, as its main assets (investment property) exhibit relatively liquid markets. Consequently,

under H2, we predict that German real estate firms are more likely to switch to fair value than


22/47

21

German firms in other industries, while UK real estate firms are less likely to switch to historical

cost than UK firms in other industries.

Our sample comprises the 275 companies (124 UK companies; 151 German companies)

that hold investment property. Depending on the specification, additional data requirements limit

the sample further. We begin with a basic regression that examines whether accounting methods

vary based on country of domicile and industry type,

1 2 3 4* 65 * 65

IFRSFair UK UK Sic Germany Germany Sic = + + + + , (1)

where UK (Germany) is an indicator variable that takes the value of one for UK (German)

companies and zero otherwise, and Sic65 is an indicator that takes the value of one when a

company has SIC code 65 (real estate) among its first five SIC codes and zero otherwise.

Equation 1 examines how consistency of valuation practices varies, conditional on whether real

estate is among a company's primary business activities. As discussed in Section 2, the UK and

Germany have very different institutional features. If institutions, as opposed to accounting

standardsper se, determine the accounting practice, then we expect to find the country variables

to be significant. Specifically, the coefficients 1 and 3 capture the consistency of reporting

methods in the UK and Germany in general, while 2and4measure increments in consistency

when real estate is among a company's primary business activities.

Table 7 Column 1 presents the regression estimates of Equation 1. The pseudo R-squared

suggests that Equation 1 explains a substantial portion (i.e., 34%) of the variance in the decision

to use fair value. The estimates indicate that companies domiciled in Germany are significantly

more likely to use historical cost after IFRS adoption (3). This effect, however, is significantly

smaller for companies whose primary industries include real estate (3 + 4). Companies

domiciled in the UK are generally more likely to use fair value under IFRS, although the effect is


23/47

22

small. This effect, however, is much stronger (and more significant) for companies in the real

estate business (1+ 2). The evidence is in line with H2. The switches from historical cost to

fair value among German real-estate-companies and from fair value to historical cost among UK

non-real-estate-companies, when companies are no longer constrained by accounting regulation,

is strong evidence that the real estate industry has net firm-specific benefits of fair value

accounting for investment property. This finding is consistent with the capital market benefits of

fair value accounting for real estate firms documented in Muller et al. (2008). Real estate

businesses' greater propensity to switch to fair value (or continue its use) is consistent with fair

value being a superior measure of economic performance in the real estate industry. However,

the evidence also indicates that the valuation practices remain somewhat persistent across IFRS

adoption (historical cost in Germany and fair value in the UK). This finding is policy-relevant

because it implies that the application of GAAP need not be uniform under one set of standards,

and is specific to institutional setting, in line with arguments in Ball (2006).

In Table 7, Columns 2 through 6, we augment Equation 1 with log of market

capitalization and an IFRS early adoption dummy (Muller et al. 2008) and test whether fair value

is positively associated with reliance on debt as predicted by hypothesis H3.

The key finding in Column 2 is that companies that rely more heavily on debt financing

are more likely to use fair value accounting for investment property. This is consistent with the

incremental costs of obtaining reliable fair value for financial reporting purposes being low when

they are already produced for financing purposes (Holthausen and Watts 2001). An alternative

explanation is that companies may choose fair value accounting because it allows them to avoid

covenant breaches (Cotter and Zimmer 1999).


24/47

23

To shed more light on this issue, in Model 3 we decompose leverage into its long- and

short-term components, as well as proxy for reliance on convertible debt. We find that short-term

leverage is more important than long-term debt in explaining fair value use. The coefficient on

convertible debt is also significantly positive. As accounting-based covenants are less important

and less common from short-term and convertible-debt perspectives, the results are inconsistent

with the conclusion that companies use fair value opportunistically to avoid covenant violations.

Rather this finding suggests that firms accessing debt markets frequently are more likely to use

fair value. This is intuitive as these companies need to produce reliable estimates for financing

purposes more frequently and therefore have a low incremental cost of committing to fair value.

Models 4 through 6 of Table 7 replace leverage with other leverage proxies frequently

used in contracts.15

We find that the ratio of total debt to operating income is positively related to

the use of fair value, while the coverage of interest and the current ratios are negatively related to

fair value use. These results are consistent with hypothesis H3 and confirm the effect of leverage

by showing that companies that rely more on debt are more likely to use fair value.

To provide further evidence on H3, we examine whether fair value companies access

debt markets more frequently than historical cost companies following IFRS adoption in 2005. If

indeed the availability of reliable fair value estimates relates to debt financing activities, fair

value choices should predict more frequent debt market access. Based on Worldscope data for

2006 and 2007, we construct several proxies for access to debt and equity financing.

Specifically, we proxy for future debt financing with the following variables: DebtIss1

(DebtIss2) indicates whether by 2007 total debt (long-term debt) had increased by more than

10% of the current market value of assets; FtrLev1 (FtrLev2) proxies for the level of future total

15We exclude leverage because these variables are highly correlated with leverage and therefore capture aspects of

the same construct.


25/47

24

debt (long-term debt) in 2007 while controlling for the level of current debt in the regression; and

DbtGrow1 (DbtGrow2) indicates growth in total debt (long-term debt). To ensure that our debt

financing variables are not capturing financing activity in general we measured equity market

access as a benchmark. Proxies for equity issuance over 2006 and 2007 are as follows: EqIss1

indicates whether combined net proceeds of equity issuance less proceeds from stock options

exceed 10% of market value of current assets; andEqIss2 is the ratio of net proceeds to current

market value of assets. We regress these proxies on both the fair value indicator variable and

controls for company characteristics that include country, size, leverage, and an SIC code 65

indicator.

We present our findings in Table 8. Columns (1) through (6) present regressions with the

six proxies for debt issuance used as the dependent variables, while columns (7) and (8) are

based on the two equity issuance variables. All proxies for debt issuance are statistically

significant and indicate a relation between fair value use and future debt financing. Proxies for

equity financing are insignificant at the conventional levels. While we have no strong prior for

why equity market access should relate to fair value use, the relation between fair value use and

future debt issuance supports the explanation that the costs of recognizing reliable fair value

estimates are lower when firms regularly enter debt markets, as predicted by H3.

4.3.2 Property, plant, and equipment (PPE)

We conduct a similar analysis of a company's decision, post-IFRS, to apply fair value to

PPE. A few distinctions, however, bear mentioning. First, we hand-collect the fair value

revaluation reserve data from companies annual reports. This enables us to compute book values

of equity, PPE, and total assets as ifcompanies used historical cost. Thus we can include book-

to-market and book leverage as additional explanatory variables. In addition, we include the ratio


26/47

25

of PPE to total assets to examine weather PPE-heavy companies are more likely to use fair value.

Second, the percentage of fair value companies in the population is low for this asset group;

therefore, to improve our ability to draw inferences, we match each fair value company to a

historical cost company.16

We perform this match according to country of domicile, two-digit

industry code, and the log of market value of equity and use the closest match. This procedure,

which requires non-missing market value of equity, yields 90 observations. Data availability

restrictions further reduce the sample to 87 (86 in Table 8 column 7) observations.

Table 9 presents the results from our logistic regression analysis. Because we match

according to country, industry, and size, we omit these as explanatory variables. In line with the

evidence for investment property and hypothesis H3, we find a positive and significant

association between market leverage (book leverage) and the use of fair value accounting.

Further analysis in column (3) reveals that, once again, short-term debt is at least as important as

long-term debt in this association. The portion of convertible debt is now significantly negatively

related to the use of fair value, a finding for which we currently have no explanation.

The effect of book-to-market is somewhat stronger in some specifications and indicates

that after IFRS adoption, high growth companies are less likely to use fair value. One

interpretation of this result is that companies with fewer growth opportunities use fair value

accounting as a means of avoiding overinvestment in fixed assets. In particular, common

accounting metricsfor example, return on assets or return on investmentare less likely to

reflect actual performance under historical cost accounting because the depreciated cost is

usually lower than market value, that is, the value in alternative use (see Appendix C). A

commitment to fair value accounting dilutes the return on assets, makes it more costly for

16We do not include all historical cost companies in the logit regressions because this would result in an unbalanced

sample. While our matched sample is sufficient to perform statistical tests, we acknowledge that a larger sample

would be preferable for making robust inferences.


27/47

26

management to hold unproductive assets, and when fair value estimates are reliable, improves

performance measurement. In other words, a commitment to fair value effectively forces

managers to incur rent on their investments' current values, regardless of the time of purchase

and their historical cost.

We further find a positive coefficient on PPE, but it does not attain statistical significance

in most specifications. The positive coefficient on FairInvPrin Table 9 Column 5 suggests that

companies that apply fair value to investment property are more likely to also apply fair value to

PPE. Controlling for this effect, however, does not alter our findings with respect to leverage or

book-to-market.

5. Summary

We investigate the choice between fair value and historical cost accounting for non-

financial assets when markets, rather than regulators, determine the outcome. In light of the

significant debate over fair value, understanding this choice is useful for regulators as it informs

about relative firm-specific costs and benefits of fair value accounting. The IFRS setting is

different from the settings in prior studies as it allows companies to choose between historical

cost and fair value accounting for non-financial assets, but requires companies to pre-commit to

their choice over time. We examined the accounting policies for intangible assets, investment

property, and PPE of 1,539 companies. With very few exceptions, we find that fair value is used

exclusively for property. We find that 3% of companies use fair value for owner-occupied

property, compared with 47% for investment property. The lack of companies that use fair value

for all other non-financial assets is inconsistent with fair value accounting yielding net firm-

specific benefits for those assets. The use of fair value for property alone is likely explained by

lower costs to reliably measure fair values in the presence of relatively liquid property markets.


28/47

27

The main cross-sectional determinant of fair value for both investment property and PPE

is reliance on debt financing. When fair value estimates are constructed for financing purposes,

they are likely to be relatively reliable, and the incremental costs of also recognizing them in

financial reports are low.

Overall, our evidence suggests that most managers do not perceive the net benefits of fair

value accounting to exceed those of historical cost accounting for non-financial assets. However,

the cross-sectional variation in the choice reveal that fair value is chosen over historical costs

when the cost of establishing reliable fair value estimates are low. This suggests that managers

resistance to fair value use is likely to be driven by the costs of establishing reliable fair value

estimates rather than a disagreement with standard setters on the conceptual merits of fair value

accounting. These results have policy implications, as they suggest that fair value accounting for

non-financial assets is costly for most firms. Thus, mandatory fair value accounting for illiquid

non-financial assets may only be socially optimal if fair value accounting is associated with

positive externalities that exceed the net costs imposed on those firms forced to use fair value

accounting. Our setting does not allow us to explore externalities associated with fair value

accounting. More research is needed to understand these issues.


29/47

28

References

Aboody, D., M. E. Barth, and R. Kasznik. 1999. Revaluations of fixed assets and future firm

performance: Evidence from the UK. Journal of Accounting and Economics 26 (1-

3):149-178.Ball, R. 1972. Changes in accounting techniques and stock prices. Journal of Accounting

Research10(2) (1):1-38.. 2006. International Financial Reporting Standards (IFRS): Pros and Cons for Investors.

Accounting and Business Research36 (Special Issue):5-27.

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 EconomicsVol. 31 (Issue 1-3):77-104.

Barth, M. E., and G. Clinch. 1998. Revalued financial, tangible, and intangible assets:

Association with share prices and non-market-based value estimates. Journal ofAccounting ResearchVol. 36:199-233.

Brown, P., H. Y. Izan, and A. L. Loh. 1992. Fixed asset revaluations and managerial incentives.

Abacus28 (1):36-57.

Cairns, D. 2006. The Use of Fair Value in IFRS.Accounting in Europe3 (1):5 - 22.Cairns, D., D. R. Massoudi, and A. Tarca. 2008. IFRS Fair Value Measurement and Accounting

Policy Choice in the United Kingdom and Australia: London School of Economics

working paper.Cotter, J., and I. Zimmer. 1995. Asset revaluations and assessment of borrowing capacity. In

Abacus: Blackwell Publishing Limited, 136-151.

. 1999. Disclosure Versus Recognition: The Case of Asset Revaluations. SSRN eLibrary.

Danbolt, J., and W. Rees. 2008. An experiment in fair value accounting: UK investmentvehicles.European Accounting Review17 (2):271 - 303.

Dietrich, R. J., M. S. Harris, and K. A. Muller. 2001. The reliability of investment property fair

value estimates.Journal of Accounting and Economics30 (2):125-158.

Easton, P. D., P. H. Eddey, and T. S. Harris. 1993. An Investigation of Revaluations of TangibleLong-Lived Assets.Journal of Accounting Research31 (3):1-38.

Fabricant, S. 1936. Revaluations of fixed assets, 1925 - 1934. National Bureau of Economic

ResearchBulletin 62.

Hail, L., C. Leuz, and P. D. Wysocki. 2009. Global accounting convergence and the potential

adoption of IFRS by the United States: An analysis of economic and policy factors: TheWharton School, University of Chicago, and MIT.

Herrmann, D., S. M. Saudagaran, and W. B. Thomas. 2006. The quality of fair value measures

for property, plant, and equipment.Accounting Forum30 (1):43-59.

Holthausen, R. W., and R. L. Watts. 2001. The relevance of the value-relevance literature forfinancial accounting standard setting. Journal of Accounting and Economics31 (1-3):3-

75.IASB. 2006. Preliminary views on an improved conceptual framework for financial reporting:

The objective of financial reporting and qualitative characteristics of decision-useful

financial reporting.Discussion paper, IASB July 2006.

Jensen, M. C., and W. H. Meckling. 1976. Theory of the firm: Managerial behavior, agency costsand ownership structure.Journal of Financial Economics3 (4):305-360.

Johnson, L. T. 2005. Relevance and reliability. The FASB Report.


30/47

29

Kothari, S. P., K. Ramanna, and D. Skinner. 2009. What should GAAP look like? A survey and

economic analysis: Chicago Booth and MIT.

Laux, C., and C. Leuz. 2009. The crisis of fair-value accounting: Making sense of the recentdebate.Accounting, Organizations and Society34 (6-7):826-834.

Leuz, C., and J. Wstemann. 2004. The Role of Accounting in the German Financial System. In

The German Financial System, edited by J. Krahnen and R. H. Schmidt: OxfordUniversity Press.

Muller, K. A., III. 1999. An examination of the voluntary recognition of acquired brand names in

the United Kingdom.Journal of Accounting and Economics26 (1-3):179-191.

Muller, K. A., III, and E. J. Riedl. 2002. External Monitoring of Property Appraisal Estimatesand Information Asymmetry.Journal of Accounting Research40 (3):865-881.

Muller, K. A., III, E. J. Riedl, and T. Sellhorn. 2008. Causes and consequences of choosing

historical cost versus fair value: Pennsylvania State University, Harvard Business School,Ruhr-Universitat Bochum.

Paton, W. A., ed. 1932. Accountants' Handbook. Vol. Second. New York: The Ronald Press

Company.

Ramanna, K., and R. L. Watts. 2009. Evidence from goodwill non-impairments on the effects ofusing unverifiable estimates in financial reporting. Harvard Business School working

paper 09-106.

Schipper, K. 2005a. Fair values in financial reporting. http://fars.org/2005AAAFairValue

KSchipper.pdf.

. 2005b. The introduction of International Accounting Standards in Europe: Implications

for international convergence. In European Accounting Review: The EuropeanAccounting Association, 101-126.

Sharpe, I. G., and R. G. Walker. 1975. Asset revaluations and stock market prices. Journal ofAccounting Research13 (2):293-310.

Shleifer, A., and R. W. Vishny. 1992. Liquidation Values and Debt Capacity: A Market

Equilibrium Approach. The Journal of Finance47 (4):1343-1366.

Sloan, R. G. 1999. Evaluating the reliability of current value estimates. Journal of Accountingand Economics26 (1-3):193-200.

Standish, P. E. M., and S.-I. Ung. 1982. Corporate signaling, asset revaluations and the stock

prices of British companies. In Accounting Review: American Accounting Association,

701.Wahlen, J. M., J. R. Boatsman, R. H. Herz, G. J. Jonas, K. G. Palepu, S. G. Ryan, K. Schipper,

C. M. Schrand, and D. J. Skinner. 2000. Response to the FASB Preliminary Views:

Reporting Financial Instruments and Certain Related Assets and Liabilities at Fair Value.Accounting Horizons14 (4):501-508.

Watts, R. L. 2006. What has the Invisible Hand Achieved?: MIT Sloan School of Management

working paper.Watts, R. L., and J. L. Zimmerman. 1979. The demand for and supply of accounting theories:

The market for excuses. The Accounting Review54 (2):273-305.

. 1986. Positive accounting theory: Prentice-Hall contemporary topics in accounting

series.Whittred, G., and Y. K. Chan. 1992. Asset revaluations and the mitigation of underinvestment.

Abacus28 (1):58-74.
http://fars.org/2005AAAFairValuehttp://fars.org/2005AAAFairValuehttp://fars.org/2005AAAFairValue


31/47

30

Appendix A: Examples of accounting practice

This appendix presents examples of fair value and historical cost accounting from the accounting policysection of annual reports of companies in our samples. Panel A presents an example of a switch from fair

value under UK-GAAP to historical cost under IFRS. Panel B presents an example of fair valueaccounting under both UK-GAAP and IFRS. Panel C presents an example of a German company that

uses fair value accounting under IFRS.

Panel A: Switch from fair value to historical cost

Annual report according to UK-GAAP for 2004

AMEC plc annual report and accounts 2004 (page 67)

12 Tangible assets (continued)

All significant freehold and long leasehold properties were externally valued as at 31 December 2004 byCB Richard Ellis Limited in accordance with the Appraisal and Valuation Manual of the Royal Institute

of Chartered Surveyors.

For the United Kingdom, the basis of revaluation was the existing use value for properties occupied by

group companies and the market value for those properties without group occupancy. For propertiesoutside the United Kingdom, appropriate country valuation standards were adopted that generally reflectmarket value.

No provision has been made for the tax liability that may arise in the event that certain properties aredisposed of at their revalued amounts.

The amount of land and buildings included at valuation, determined according to the historical costconvention, was as follows:

Group Group Company Company

2004 2003 2004 2003 million million million million

Cost 39.2 46.4 9.3 8.6Depreciation (10.61) (13.9) (2.5) (1.7)

Net book value 26.6 32.5 6.8 6.9

Annual report according to IFRS for 2005

AMEC plc annual report and accounts 2005 (page 102)

IAS 16 Property, plant and equipment

Under UK-GAAP, AMECs policy was to revalue freehold and long leasehold property on a regularbasis. Under IAS 16, AMEC has opted to carry property, plant, and equipment at cost less accumulateddepreciation and impairment losses. As permitted by IFRS 1, AMEC has frozen the UK-GAAP land and

buildings revaluations as at 1 January 2004 by ascribing the carrying value as deemed cost. The impact of

this change in policy is as follows:

the revaluation reserve is reclassified into retained earnings as at the date of transition;

the results of the external revaluation as at 31 December 2004 are reversed, reducing the value ofproperty, plant and equipment as at 31 December 2004 by 9.6 million; and

as part of the 2004 external revaluation, certain properties were revalued downwards. Under UK-GAAP, these deficits were charged against previous revaluations held in the revaluation reserve.

Under IFRS, these downward revaluations have been taken as indicators that the value of therelevant properties is impaired and as such, they have been charged to the income statement as


32/47

31

impairment charges in 2004. This reduces the profit for the year ended 31 December 2004 and thevalue of property, plant and equipment as at 31 December 2004 by 1.8 million.

Panel B: Fair value under both UK-GAAP and IFRS

Annual report according to UK-GAAP for 2004

The Wolverhampton & Dudley Breweries, PLC Annual report 2005 (page 44).

(e) Tangible fixed assets

Freehold and leasehold properties are stated at valuation or at cost. Plant, furnishings, equipment,and other similar items are stated at cost.

Freehold buildings are depreciated to their residual value on a straight line basis over 50 years. Other tangible fixed assets are depreciated to their residual value on a straight line basis at rates

calculated to provide for the cost of the assets over their anticipated useful lives. Leaseholdproperties are depreciated over the lower of the lease period and 50 years and other tangibleassets over periods ranging from three to 15 years.

Own labor directly attributable to capital projects is capitalized.

Valuation of properties:Trading properties are revalued professionally by independent valuers on a five-

year rolling basis. When a valuation or expected proceeds are below current carrying value, the assetconcerned is reviewed for impairment. Impairment losses are charged directly to the revaluation reserveuntil the carrying amount reaches historical cost. Deficits below historical cost are charged to the profit

and loss account except to the extent that the value in use exceeds the valuation, in which case this istaken to the revaluation reserve. Surpluses on revaluation are recognized in the revaluation reserve, except

to the extent that they reverse previously charged impairment losses, in which case they are recorded inthe profit and loss account. Any negative valuations are accounted for as onerous leases and includedwithin provisions (see note 20).


The Wolverhampton & Dudley Breweries, PLC Annual report 2006 (page 46).

Property, plant and equipment

Freehold and leasehold properties are stated at valuation or at cost. Plant, furnishings, equipment,

and other similar items are stated at cost. Depreciation is charged to the income statement on a straight-line basis to provide for the cost of

the assets less residual value over their useful lives. Freehold and long leasehold buildings are depreciated to residual value over 50 years. Short leasehold properties are depreciated over the life of the lease.

Other plant and equipment is depreciated over periods ranging from 3 to 15 years. Own labor directly attributable to capital projects is capitalized.

Land is not depreciated.

Valuation of properties - Properties are revalued by qualified valuers on a regular basis using open market

value so that the carrying value of an asset does not differ significantly from its fair value at the balancesheet date. When a valuation is below current carrying value, the asset concerned is reviewed forimpairment. Impairment losses are charged to the revaluation reserve to the extent that a previous gainhas been recorded, and thereafter to the income statement. Surpluses on revaluation are recognized in therevaluation reserve, except where they reverse previously charged impairment losses, in which case they

are recorded in the income statement.


33/47

32

Panel C: Fair value accounting by German company


Hypo Real Estate Group, Annual report 2006 (page 96)

12 Property, plant, and equipment

Property, plant, and equipment is normally shown at cost of purchase or cost of production. As anexception to this rule, land and buildings are shown with their fair value in accordance with IAS 16. Thecarrying amounts if the assets are subject to wear and tear are diminished by depreciation inaccordance with the expected service life of the assets. In the case of fittings in rented buildings, thecontract duration taking account of extension options is used as the basis of this contract duration is

shorter than the economic life.

Appendix B: Variable definitions

Fair_IFRS = one if the company uses fair value after adoption of IFRS, and zero otherwise.

UK = one if a company is domiciled in the UK, and zero otherwise.

UkSic65= one if a company has SIC 65 (real estate) among its first five SIC codes and is domiciled in theUK, and zero otherwise.

Germany= one if a company is domiciled in Germany, and zero otherwise.

GermanySic65= one if a company has SIC 65 (real estate) among its first five SIC codes and is domiciledin Germany, and zero otherwise.

Early= one if the company adopted IFRS before 2005, and zero otherwise.

Size= log of market value of equity.

PPEA = property, plant, and equipment less revaluation reserve divided by total assets less revaluationreserve.

MktLev = total liabilities divided by market value of assets (defined as book value of liabilities plusmarket value of equity) as of December 2005.

MktLevLong= long-term debt divided by market value of assets (liabilities plus market value of equity)

as of December 2005.

MktLevShort = short-term liabilities defined as total liabilities less long-term debt divided by market

value of assets (liabilities plus market value of equity) as of December 2005.

LevBook = book leverage defined as total liabilities divided by total assets net of fair value revaluationreserve.

LevBookLong = long-term debt divided by total assets net of fair value revaluation reserve.

LevBookShort = ratio of total liabilities minus long-term debt to total assets net of fair value revaluation

reserve.

Convertible= ratio of convertible debt to long-term debt.

DebtToOi= total liabilities divided by operating income.

Coverage= operating income divided by interest expense.

Current= current assets divided by current liabilities.

Dividend= one if company pays dividends, and zero otherwise.


34/47

33

RE = retained earnings scaled by the market value of equity plus total liabilities.

D(RE


35/47

34

value and historical cost subgroups. We find that, on average, the ratio of book value of total assets tomarket value of total assets is 16% higher for companies that recognize investment property at fair value;

the ratio of book value of equity to market value of equity is 27% higher. Among companies that applyfair value to property, plant, and equipment, we find that the ratio of book value of total assets to market

value of total assets and the ratio of book value of equity to market value of equity are, respectively, 31%and 87% higher than those of matched companies that use only historical cost. The differences in the

book values of assets and equity in both the investment property and property, plant, and equipmentsamples are all significant at the 1% level. We also examine how return on assets (ROA) differs betweenfair value and historical cost companies. We find a lower ROA in the property, plant, and equipment

sample among companies that recognize assets at fair value. In the investment property sample, we alsofind a lower ROA among companies that use fair value accounting; this difference, however, isstatistically insignificant.19

[Insert Table 1C here]

The evidence in Table 6 indicates that the decision to use the fair value method is associated with

economically significant differences in companies balance sheets, which makes companies that use fairvalue accounting appear less conservative in terms of their book-to-market ratios.

19 It is not surprising that fair value accounting for property decreases ROA because while, on average, fair value

accounting increases the book value of assets, upward revaluations do not affect the net income. For investment

property this effect is smaller because upward revaluations increase both net income and total assets (see Section 2).


36/47

Table 1: Sample selection processTable 1 presents the sample selection process and industry distribution. Panel A presents the selection process for

the UK samples. The cross-sectional sample consists of companies for which we can identify an annual report

according to IFRS. Our switch sample further requires that an annual report (according to UK-GAAP) be available

prior to mandatory IFRS adoption. Panel B presents the selection process for the German sample. To be included in

the German sample, companies must have available an annual report according to IFRS. Percentages are rounded

and thus may not exactly sum to 100%.

Panel A: UK samples

Active companies (FBRIT, March 2008) 2,312

Inactive companies (DEADUK, March 2008) + 5,597

UK listed companies in Worldscope 7,909

Companies that Worldscope does not classify as complying with

IFRS in 2005 or 2006 (mainly inactive companies) 6,464

IFRS companies 1,445

Companies not domiciled in the UK 270

UK companies subject to mandatory IFRS 1,175

Companies for which we cannot identify an IFRS annual report

or companies with more than one security listed on LSE 241UK Cross-Sectional Sample 934

Companies for which we cannot identify a UK-GAAP annual report 231

UK Switch Sample 703

Industry distribution in the UK samples

UK UK UK UK

Worldscope IFRS Cross-sectional Switch

No. Industry name All All Excl. N/A

obs. % % Obs. % Obs. % Obs. %

N/A No industry classification 3,139 40%

13 Aerospace 19 0% 0% 11 1% 5 1% 5 1%

16 Apparel 42 1% 1% 7 1% 7 1% 6 1%

19 Automotive 44 1% 1% 6 1% 6 1% 5 1%22 Beverages 50 1% 1% 11 1% 7 1% 5 1%

25 Chemicals 106 1% 2% 24 2% 20 2% 18 3%

28 Construction 234 3% 5% 65 5% 54 6% 49 7%

31 Diversified 48 1% 1% 8 1% 6 1% 6 1%

34 Drugs, Cosmetics, & Health care 178 2% 4% 52 4% 44 5% 31 4%

37 Electrical 55 1% 1% 13 1% 10 1% 6 1%

40 Electronics 539 7% 11% 136 11% 109 12% 90 13%

43 Financial 674 9% 14% 187 16% 140 15% 105 16%

46 Food 107 1% 2% 27 2% 23 2% 21 3%

49 Machinery & equipment 141 2% 3% 22 2% 22 2% 16 2%

52 Metal producers 213 3% 4% 59 6% 49 5% 23 3%

55 Metal product manufacturers 63 1% 1% 15 1% 12 1% 11 2%

58 Oil, gas, coal, & related services 251 3% 5% 64 6% 52 6% 29 4%

61 Paper 44 1% 1% 9 1% 6 1% 5 1%64 Printing & publishing 92 1% 2% 25 2% 19 2% 16 2%

67 Recreation 294 4% 6% 51 4% 41 4% 32 5%

70 Retail 211 3% 4% 56 4% 51 5% 43 6%

73 Textile 53 1% 1% 10 1% 8 1% 6 1%

76 Tobacco 10 0% 0% 6 0% 3 0% 3 0%

79 Transportation 73 1% 2% 19 2% 17 2% 14 2%

82 Utilities 232 3% 5% 62 6% 31 3% 23 3%

85 Other industries 997 13% 21% 230 18% 192 21% 135 19%

Total 7,909 100% 100% 1,175 100% 934 100% 703 100%


37/47

(Table 1 continued)

Panel B: The German sample

Active companies (March 2008) 1,437

Inactive companies (March 2008) + 10,126

German listed companies in Worldscope 11,563

Companies that Worldscope does not classify as complying withIFRS in 2005 or 2006 (mainly inactive companies) 10,117

IFRS companies 1,446

Companies not domiciled in Germany 635

German companies subject to mandatory IFRS 811

Companies for which we cannot identify an IFRS annual report 206

German sample 605

Industry distribution in the German sample

German German German

Worldscope IFRS Sample

No. Industry name All All Excl. N/Aobs. % % Obs. % Obs. %

N/A No industry classification 2,192 19% N/A N/A N/A N/A N/A

13 Aerospace 25 0% 0% 2 0% 1 0%

16 Apparel 72 1% 1% 19 2% 11 2%

19 Automotive 99 1% 1% 18 2% 10 2%

22 Beverages 93 1% 1% 11 1% 9 1%

25 Chemicals 207 2% 2% 34 4% 16 3%

28 Construction 235 2% 3% 34 4% 22 4%

31 Diversified 74 1% 1% 12 1% 8 1%

34 Drugs, Cosmetics, & Health care 610 5% 7% 36 4% 25 4%

37 Electrical 143 1% 2% 21 3% 13 2%

40 Electronics 1,832 16% 20% 95 12% 78 13%

43 Financial 1,430 12% 15% 131 16% 96 16%46 Food 154 1% 2% 11 1% 8 1%

49 Machinery & equipment 324 3% 3% 69 9% 54 9%

52 Metal producers 251 2% 3% 1 0% 1 0%

55 Metal product manufacturers 108 1% 1% 10 1% 6 1%

58 Oil, gas, coal, & related services 323 3% 3% 9 1% 5 1%

61 Paper 69 1% 1% 9 1% 6 1%

Documents

Christensen and Nikolaev 2011 Does Fair Value Accounting for Non-financial Assets Pass the Market Test