This article was downloaded by: [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi]

On: 29 September 2013, At: 20:11

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Accounting and Business Research

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rabr20

Value‐relevance of presenting changes in fair value

of investment properties in the income statement:
Evidence from Hong Kong
a b
Stella So & Malcolm Smith
a
Department of Accountancy and Law, Hong Kong Baptist University
b
Malcolm Smith, School of Accounting, Finance & Economics, Edith Cowan University,
Western Australia, 270 Joondalup Drive, Joondalup, WA, 6027 Phone: (08) 6304 5263 E-mail:
Published online: 04 Jan 2011.

To cite this article: Stella So & Malcolm Smith (2009) Value‐relevance of presenting changes in fair value of investment
properties in the income statement: Evidence from Hong Kong, Accounting and Business Research, 39:2, 103-118, DOI:
10.1080/00014788.2009.9663352

To link to this article: http://dx.doi.org/10.1080/00014788.2009.9663352

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Accounting and Business Research, Vol. 39. No. 2. pp. 103-118. 2009 103

Value-relevance of presenting changes in fair

value of investment properties in the income
statement: evidence from Hong Kong
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

Stella So and Malcolm Smith*

Abstract—This study investigates the value-relevance of the revision introduced in HKAS 40 (2004) ‘Investment
Property’ on the presentation of changes in the fair value of investment properties. The revision follows that introduced
in IAS 40 (2000) as Hong Kong adopted the International Financial Reporting Standards in 2005. As introduced in
IAS 40 (2000), HKAS 40 (2004) requires that companies choosing to adopt the fair value model have to present
changes in the fair value of investment properties in the income statement. Previously under the Hong Kong account-
ing standard SSAP 13 (2000), such changes were presented primarily in the revaluation reserve. Using both a three-
day short window centred around the earnings announcement date and a 12-month long window, this study provides
evidence that investors value the HKAS 40 (2004) revision in the presentation of the changes in fair value of invest-
ment properties. Based on a sample of listed property companies in Hong Kong during 2004–2006, the results of this
study show a significantly higher market price reaction and returns association when changes in fair value of invest-
ment properties are presented in the income statement. The results of this study are of interest not only to academic
researchers, but to practitioners and standard setters as they assess the decision usefulness of the revised presentation.
Keywords: financial reporting; Hong Kong; investment; property; fair value

1. Introduction Accounting Practice No. 13 ‘Accounting for

This study examines the value-relevance of the ac- Investment Properties’ (HKSA, 2000), hereafter
counting information reported as a consequence of SSAP 13 (2000), such changes (labelled, however,
Hong Kong Accounting Standard No. 40 as open market value2 changes) were presented
‘Investment Property’ (HKICPA, 2004), hereafter
HKAS 40 (2004), on the presentation of changes 1 The fair value model discussed in this paper is introduced
in fair value of investment properties. The date in IAS 40 (2000). Although it had been revised in 2003 and su-
1 January 2005 marked the beginning of a new era perseded by IAS 40 (2003) (IASCF, 2008a), the revision in
when accounting standards in Hong Kong became 2003 is related to clarification of aspects of property interest
fully converged with International Financial that are held under an operating lease and is therefore of no
Reporting Standards (IFRS). As part of the final relevance to this study. Of relevance to this study are the fair
value model and the required presentation of fair value
phase of the convergence, a number of new ac- changes of investment properties in the income statement
counting standards which are word-for-word under the fair value model, which were introduced in IAS 40
equivalents of the IFRSs were issued to replace the (2000). Throughout this paper, reference is therefore made to
existing pronouncements. HKAS 40 (2004) is one IAS 40 (2000) and not to IAS 40 (2003).
2
SSAP 13 (2000) adopted the definition of open market
of them. Following International Accounting value from the Hong Kong Institute of Surveyors which de-
Standards No. 40 ‘Investment Properties’ (2000) fined it as the best price at which an interest in a property
(IASC, 2000), hereafter IAS 40 (2000),1 HKAS 40 might reasonably be expected to be sold at the date of valua-
(2004) requires that companies choosing to adopt tion assuming a willing seller, a reasonable period in which to
the fair value model must present changes in the negotiate the sale taking into account the nature of the proper-
ty and the state of market, that values will remain static during
fair value of investment properties in the income that period, that the property will be freely exposed to the open
statement. Previously under Statement of Standard market, and that no account will be taken of any additional bid
by a purchaser with a special interest (HKSA, 2000). The
IASB defines fair value as the amount for which an asset could
be exchanged between knowledgeable, willing parties in an
*Stella So is in the Department of Accountancy and Law at arm’s length transaction (IASCF, 2008a). The former IASC, in
Hong Kong Baptist University and Malcolm Smith is at the its Basis for Conclusions on IAS 40 (2000), states that its con-
School of Accounting, Finance & Economics at Edith Cowan cept of fair value for investment properties is similar to the
University, Western Australia. International Valuation Standards Committee (IVSC) concept
Correspondence should be addressed to: Professor Malcolm of market value. IVSC defines market value as the estimated
Smith, School of Accounting, Finance & Economics, amount for which an asset should exchange on the date of val-
Edith Cowan University, 270 Joondalup Drive, Joondalup WA uation between a willing buyer and a willing seller in an arm’s
6027, Western Australia. Tel: (08) 6304 5263. E-mail: length transaction after proper marketing wherein the parties
[email protected]. had each acted knowledgeably, prudently and without compul-
This paper was accepted for publication in November 2008. sion (IASCF, 2008c).
 104 ACCOUNTING AND BUSINESS RESEARCH

primarily3 in a different location, the revaluation these prior findings and associated theory, it is sur-
reserve. prising that Owusu-Ansah and Yeoh (2006) find
Changes in the fair value of investment proper- no difference whether unrealised gains (losses
ties, whether increases or decreases, represent were not studied) on investment properties in New
gains or losses which are unrealised. There have Zealand are reported in the income statement or
been concerns that the inclusion and presentation revaluation reserve. Further research is therefore
of such unrealised gains and losses in the income needed to provide more evidence about the impact
statements might lead to undue increases in earn- of presentation location generally, and of changes
ings volatility and investor confusion (HKSI, in fair value of investment properties in particular.
2006). The former International Accounting Second, existing research (see Landsman (2007)
Standards Committee (IASC) argues, however, in for a detailed summary) focuses on financial assets
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

its Basis for Conclusions on IAS 40 (2000) that and liabilities (Barth, 1994; Barth et al., 1995,
such presentation provides the most relevant and 1996, 2006); Eccher et al., 1996; Nelson, 1996,
transparent view of the financial performance of Carroll et al., 2003; Hirst et al., 2004; Hodder et
investment properties (IASCF, 2008c). After all, al., 2006; Danbolt and Rees, 2008) and employee
the objective of financial statements is not to share options (Espahbodi et al., 2002; Robinson
smooth profit figures (McBride, 2006), but to re- and Burton, 2004). While there are studies on non-
flect economic reality for economic decision mak- financial assets, for intangible assets and tangible
ing (IASCF, 2008b). long-lived assets (e.g. Barth and Clinch, 1998;
The replacement of SSAP 13 (2000) by HKAS Aboody et al., 1999; Muller and Riedl, 2002) and
40 (2004) in Hong Kong thus provides a unique on investment properties (i.e. Dietrich et al., 2001;
opportunity for this study to examine the issues re- Owusu-Ansah and Yeoh, 2006), the empirical evi-
lating to the impact of the presentation location, dence is largely based on fair value disclosures in
whether in the income statement or revaluation re- the notes to the accounts from the 1990s. More
serve, for changes in fair value of investment prop- studies should be conducted using data after the
erties. The results of this study will have implementation of the fair value accounting
implications for companies around the world standards (e.g. IAS 39 ‘Financial Instruments:
that prepare their financial statements using Recognition and Measurement’, and HKAS 40
International Financial Reporting Standards. ‘Investment Property’ studied in this study).
Like all other value-relevance studies, this study
use share prices and share returns to infer whether 3 Under SSAP 13 (2000), investment properties were to be
investors consider accounting information to be suf- included in the balance sheet at their open market value and
ficiently relevant and reliable to be useful in making changes in the open market value were to be recognised pri-
investment decisions (Maines and Wahlen, 2006; marily in the revaluation reserve, i.e. in the revaluation reserve
Landsman, 2007). Value-relevance tests are gener- unless (1) its balance was insufficient to cover a deficit, in
which case the amount by which the deficit exceeded the
ally joint tests of relevance and reliability4 (Barth et revaluation reserve balance was to be charged to the income
al., 2001), where reliability is more than agreement statement; and (2) a revaluation surplus subsequently arises,
about a measure (measurement verifiability), but this surplus was to be credited to the income statement to the
also involves the correspondence between descrip- extent of the deficit previously charged (HKSA, 2000).
4 This paper assumes the application of the Framework for
tion, classification and presentation (representation- the Preparation and Presentation of Financial Statements pub-
al faithfulness), of the phenomenon it purports to lished by the former International Accounting Standards
represent (IASCF, 2008b; Maines and Wahlen, Committee (IASC) in 1989 and re-adopted by the International
2006). This study examines the relevance and relia- Accounting Standards Board (IASB) in 2001 (hereafter
bility of the HKAS 40 (2004) revision on the pres- Framework (1989), IASCF, 2008b). In Framework (1989), rel-
evance and reliability are two of the four principal qualitative
entation of changes in fair value of investment characteristics of decision-useful financial statements (IASCF,
properties, by evaluating the revision’s ability to im- 2008b). Information is relevant if it influences the economic
pact upon the abnormal share returns to investors. decisions of users by helping them evaluate past, present or fu-
There are three distinct issues which motivate ture events or confirming, or correcting, their past evaluations.
Information is reliable when it is free from material error and
this study. bias and can be depended upon by users to represent faithful-
First, while the efficient market hypothesis sug- ly that which it purports to represent or could reasonably be
gests that the presentation location of accounting expected to represent. Of note is that the term reliability in
information is not valued by investors, results Framework (1989) has been proposed to be replaced by ‘faith-
ful representation’ in the Exposure Draft of an Improved
from prior research have shown otherwise (e.g. Conceptual Framework for Financial Reporting, jointly pub-
Hirst and Hopkins, 1998; Maines and McDaniel, lished in May 2008 by the IASB and the US Financial
2000; Barth et al., 2003; Hirst et al., 2004; Accounting Standards Board (FASB). The boards considered
Chambers et al., 2006; and Lee et al., 2006). The that faithful representation encompasses all the key qualities
that Framework (1989) includes as aspects of reliability and
presentation location of an accounting amount af- therefore proposed the replacement. No attempt is however
fects not only its relevance but also its reliability made in this paper to anticipate the outcome of the exposure
(IASCF, 2008b; Maines and Wahlen, 2006). Given draft, which is expected to be finalised in the first half of 2009.
 Vol. 39 No. 2. 2009 105

Third, Hong Kong is a world-recognised centre changes in fair value are inextricably linked as in-
for property construction, development and invest- tegral components of the financial performance of
ment (KPMG China and Hong Kong, 2004) and an investment property and are therefore presented
the total market capitalisation of its property in- in the income statements (IASCF, 2008c).
dustry (conglomerates/consolidated enterprises Although IAS 40 (2000) permits entities to choose
excluded) in 2006 was HK$13,249bn, represent- between a fair value model or a cost model, the
ing 11% of the total market capitalisation of all Basis for Conclusions on IAS 40 (2000) states
Hong Kong Stock Exchange Main Board equities clearly that it is highly unlikely that a subsequent
(HKSE, 2005, 2006). Investment property is a sig- change from the fair value model to the cost model
nificant component of many company balance can be made on the grounds of more appropriate
sheets in Hong Kong and the way it is accounted presentation (IASCF, 2008c).
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

for has become an issue of prominent interest in However, Penman (2007) does not entirely
Hong Kong in 2005–2006 (McBride, 2006). agree; he evaluates historical cost and fair value
This study employs a sample of listed companies accounting from two perspectives – equity valua-
from the Hong Kong property industry and exam- tion and stewardship and concludes that while fair
ines whether the relevance and reliability of value accounting is a plus at a conceptual level, the
disclosures are enhanced through the adoption of minuses add up with fair value implemented as
HKAS 40 (2004) and the reporting of changes in exit price (whether estimated or observed in active
fair value of investment properties in the income markets) and the problems with historical cost ac-
statement. Only companies from the property indus- counting remains unresolved.
try are included because these companies are ex- Singleton-Green (2007) summarises the prob-
pected to hold substantial levels of investment lems of fair value accounting as: (1) the lack of ac-
properties on their balance sheets and their earnings tive markets for most assets and liabilities, which
are thus likely to be more sensitive to the adoption means that most fair value measurements are esti-
of HKAS 40 (2004). Depending on the sample mates and are highly subjective and potentially
company’s accounting year-end, and the incidence unreliable; (2) costly information, especially for
of early adoption of HKAS 40 (2004), the sample smaller companies; and (3) the recognition of prof-
period in this study extends from 2004 to 2006. its based on fair values, which mean that unre-
Using models adapted from Easton and Harris alised profits or losses from changes in fair value
(1991), Amir et al. (1993) and Barth (1994), re- are recognised, and result in greater volatility and
sults from this study provide evidence on the high- unpredictability. This study focuses on the third
er value-relevance of presenting fair value gains or issue, the presentation of changes in fair value of
loss in the income statement versus presenting investment properties, in the income statement
them in the revaluation reserve. The higher value- versus the revaluation reserve.
relevance is found to be evident in both the short- Empirical studies assessing the relevance and re-
window market reaction to the release of annual liability of fair value accounting versus historical
earnings announcements and the long-window cost-based accounting focus on financial instru-
market-adjusted annual returns. Taken together, ments, and the results from these studies are gen-
our results give support to the HKAS 40 (2004) erally mixed. Barth (1994) finds that, for a sample
presentation which investors appear to value more of US banks with data from 1971–1990, disclosed
than previously under SSAP 13 (2000). fair value estimates of investment securities pro-
The paper proceeds as follows. Section 2 de- vide significant incremental5 explanatory power
scribes related prior research and Section 3 details for bank share prices beyond that provided by his-
the research method employed. Section 4 discuss- toric costs. Fair value gains and losses of invest-
es the empirical results, and the paper concludes in ment securities (constructed from two annually
Section 5 with a summary and discussion of re- disclosed fair value estimates) are, however, found
search opportunities. to have no significant incremental explanatory
power for annual returns (changes in share price),
due to the increased measurement errors (Barth,
2. Related prior studies 1994). Similar results are obtained in Barth et al.
IAS 40 (2000) represents the first time that the (1995), Barth et al. (1996), Eccher et al. (1996)
IASB permits a fair value model for non-financial and Nelson (1996), all using bank data. Results
assets (IASCF, 2008c). Under the fair value model, from Carroll et al. (2003) differ; instead of using
investment properties are carried at fair values and bank data, they sample closed-end mutual funds
changes in fair value, whether up or down, are in- which typically have investment securities (report-
cluded in the profit or loss for the period and pre-
sented in the income statements. Supporters of the
fair value model believe that fair values give users 5
According to Biddle et al. (1995), an incremental compar-
of financial statements more useful information ison determines whether one accounting measure provides in-
than other measures, such as depreciated cost, and formation content beyond that provided by another.
 106 ACCOUNTING AND BUSINESS RESEARCH

ed at fair values) comprising virtually all their as- ures of selling price than respective historical
sets and with negligible liabilities and other assets. costs. Dietrich et al. (2001) also find that the relia-
This is an advantage because the potential problem bility of appraisal estimates increases when moni-
introduced by measuring some assets and liabili- tored by external appraisers and Big Six auditors.
ties at fair value but others at historical cost, is The New Zealand (hereafter NZ) SSAP No. 17
eliminated. Significant association between share ‘Accounting for Investment Properties and
prices and the fair value of investment securities, Properties Intended for Sale’ (NZSA, 1989) previ-
as well as between share returns and fair value se- ously allowed NZ companies the choice of recog-
curities gains and losses are found. To examine nising unrealised gains or losses either in the
whether differences in the reliability of the fair income statement, or as movements in an invest-
value of investment securities affect investors’ as- ment property revaluation reserve, unless the total
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

sessments of the usefulness of the information, of the reserve was insufficient to cover a deficit, in
Carroll et al. (2003) examine the association be- which case the amount of deficit was to be charged
tween share prices and fair values across different in the income statement as part of operating results.
fund types and find that in all cases, including those The NZ equivalent of IAS 40 came into effect on
traded in thin markets, there is a significant associ- 1 January 2005, resulting in the elimination of the
ation between the share prices and fair values. choice of recognising unrealised gains in the reval-
In contrast, Danbolt and Rees (2008), using UK uation reserve. Owusu-Ansah and Yeoh (2006) in-
data, report no support for full fair value account- vestigate the relative value-relevance of the two
ing. While fair value income is considerably more alternative accounting treatments for unrealised
value-relevant than historic cost income, the higher gains on investment properties, based on a sample
relevance disappears in the presence of changes in of NZ companies over the period 1990 to 1999,
fair value accounting balance sheet values. Danbolt when the choice was still available. Their results
and Rees (2008) interpret their results as evidence show that recognition of unrealised gains in the in-
of the absence of an obvious advantage from adopt- come statement is not superior to recognition of un-
ing fair value income accounting if fair value bal- realised gains in the revaluation reserve in terms of
ance sheet values are available to the user. their value-relevance. However, Owusu-Ansah and
Value-relevance research studies the association Yeoh (2006) include only companies with positive
between fair value estimates and share prices or re- changes in the value of their investment properties.
turns. Sloan (1999) comments that while this asso- Taken together, findings from prior studies of
ciation provides evidence that investors find fair firms in the US, UK and Australian capital markets
value estimates to be relevant, the inferences re- during the 1990s suggest that investors have been
garding reliability are indirect and limited by the provided with fair value information (whether
fact that share prices reflect many factors other recognised or disclosed) that is generally reliable
than the fair value estimates. Dietrich et al. (2001) and relevant (whether fair value estimated by man-
subsequently use a direct approach to investigate agement or independent valuer). More research
the reliability of mandatory annual fair value ap- should be undertaken to test empirically whether
praisal estimates by chartered surveyors for UK in- relevance and reliability improve after the imple-
vestment properties and find that appraisal mentation of the fair value standards on financial
estimates understate actual selling prices but are instruments (e.g. IAS 39) and with the extension
considerably less biased and more accurate meas- of fair value accounting to non-financial assets
(i.e. IAS 40).
6
Like Owusu-Ansah and Yeoh (2006), this study
Two approaches are developed and used in Barth (1994), examines the extension of fair value accounting to
the market value approach and the earnings capitalisation ap-
proach. The former examines balance sheet items, the latter investment properties and the presentation of their
examines income statement items. Owusu-Ansah and Yeoh fair value changes in the income statements (rather
(2006) use a modified version of the market value approach, than in the revaluation reserve) in particular.
which is however not considered in this study. The market Unlike Owusu-Ansah and Yeoh (2006), this study
value approach is based on a valuation model where the mar-
ket value of equity equals a firm’s assets minus its liabilities on employs data from accounting periods when the
its balance sheet (Barth, 1994). The market value approach is related fair value accounting standard HKAS 40 is
useful if the asset being studied is reported differently in differ- implemented. Comparison is then made with those
ent balance sheets, for example, at historical cost or fair value from the immediate pre-implementation account-
because of the choice permitted for investment securities in
Barth (1994). In this study, investment properties are reported
ing periods when SSAP 13 (2000) was in effect.
in the balance sheet either under SSAP 13 (2000) at open mar- Also, unlike Owusu-Ansah and Yeoh (2006), this
ket values or under HKAS 40 (2004) at fair values. Similar study includes companies with both increases and
amounts will be reported in the balance sheets under SSAP 13 decreases in fair values and uses a return model
(2000) and HKAS 40 (2004). On the other hand, the earnings
figures in the income statements are different, depending on
adapted from Easton and Harris (1991), Amir et al.
which accounting standard is followed, which explains why (1993) and the earnings capitalisation approach6
this study has chosen to use the earnings capitalisation model. from Barth (1994).
 Vol. 39 No. 2. 2009 107

3. Research method nancial accounting theory and research since Ball

Data for this study are collected from two and Brown (1968). The information perspective
sources. The financial statement data and results posits that investors, in response to the release of a
announcement dates are derived from the pub- company’s financial statements, will analyse the
lished financial statements downloaded from statements for new and unexpected information
http://www.hkex.com.hk. The Morgan Stanley and revise their beliefs about the company’s future
Capital International (hereafter MSCI) Equity performance, causing movements in the compa-
Hong Kong Index is used in this study as the proxy ny’s share prices and resulting in abnormal returns
for equity market return in Hong Kong. The equi- to investors (Ball and Brown, 1968). Thus finan-
ty market data including the MSCI Equity Hong cial statement information is considered decision-
Kong Index are obtained from Reuters 3000Xtra. useful if it is new and unexpected, and results in
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

Companies are eligible for selection if through- abnormal returns to the investors. Easton and
out the sample period 2003–2006: (1) they are in Harris (1991) confine financial information to re-
the property industry; (2) they are listed in the ported income or earnings from the income state-
Main Board of the Hong Kong Stock Exchange; ment and develop a model relating earnings level
(3) they report investment properties in their finan- and earnings change to abnormal return, as fol-
cial statements; and (4) there is no change of ac- lows:
counting year-end. These sampling criteria yield Rjt – E[Rjt] = a1 + a2 {Ajt / Pjt-1}
an initial sample of 92 companies.
Each company in the sample is studied twice, + a3{(Ajt – Ajt-1) / Pjt-1} + ejt
first in its final accounting period using SSAP 13 where:
(2000), and then in its first accounting period
using HKAS 40 (2004). Rjt is the return on security j at time t,
Two approaches are used in this study, the short- E[Rjt] is the expected return on security j at time
window event study approach and the longer win- t,
dow return-earnings associations approach.
In both approaches, three7 control variables are Ajt is the accounting earnings per share of firm j
added to control for their potentially confounding over the time period t-1 to t,
effects, including firm size and leverage which are
Ajt-1 is the accounting earnings per share of firm
commonly controlled for in value-relevance stud-
j over the prior time period t-2 to t-1, and
ies (e.g. Fan and Wong, 2002). Market-wide prop-
erty price level in Hong Kong as proxied by the Pjt-1 is the beginning-of-period security price per
Centa-City Index8 is also included as the analysis share at t-1
in this study involves comparison of data collected
To evaluate the value-relevance of the HKAS 40
across different time periods.
(2004)’s revised presentation of gains and losses in
3.1. Short-window event study fair value of investment properties in the income
The short-window event study approach is based statement, earnings are partitioned into two com-
on the information or signalling perspective ponents: (1) earnings before gains and losses in fair
(Beaver, 1981) on the decision usefulness of finan- value of investment properties; and (2) gains and
cial reporting, a perspective that has dominated fi- losses in fair value of investment properties, as in
Barth (1994). Both components are then scaled by
total market value. In this study, the gains and loss-
7
es in fair value of investment properties, as scaled
As HKAS 40 (2004) encourages, but does not mandate,
external independent valuation for investment properties, and by total market value, are assumed to have an ex-
SSAP 13 (2000) only requires such a valuation at least every pectation of zero as changes in fair value general-
three years, whether or not there is an independent valuation ly arise from random walk processes,9 should be
should therefore be controlled for in this study. However sub- transitory in nature (Barth, 1994; Chambers et al.,
sequent data analysis shows that all the companies sampled in
this study engage an external independent valuation appraiser
2006) and represent unexpected information.
for their investment properties during all the sample year-ends An abnormal return in this study is measured by
regardless of whether SSAP 13 (2000) or HKAS 40 (2004) is the difference between the company’s return dur-
adopted. There is therefore no such need to include a control ing the period and the return on the market portfo-
variable for independent valuation in this study. lio, i.e. Rjt – Rmt, where Rmt is the return based on
8
The Centa-City Index is a monthly index based on all
transaction records as registered with the Land Registry in the market portfolio at time t, also known as the
Hong Kong, to reflect market-wide property price levels dur- market-adjusted rate of return (Amir et al., 1993).
ing the month in Hong Kong. Market-wide return is removed and the abnormal
9
Changes in fair value of investment properties may behave return obtained thus represents company-specific
as a random walk with drift, resulting in a non-zero expected
value. The drift may be reflecting the change in the market-
returns. Brown and Warner (1980) find the market-
wide property price levels, which is therefore controlled for in adjusted rate of return to perform reasonably well
this study. under a wide variety of conditions when compared
 108 ACCOUNTING AND BUSINESS RESEARCH

with the more conventional market model used in ings source, the effects from other information are
Ball and Brown (1968) and other studies (e.g. Ball therefore narrowed in this short-window analysis.
and Kothari, 1991). The regression equation (1) below is used to
In order to evaluate the value-relevance of a evaluate the differential value-relevance of the
company’s reporting of gains and losses in fair HKAS 40 (2004)’s required reporting of gains and
value of investment properties in the income state- losses in fair value of investment properties in the
ment, this study assesses how investors respond to income statement, using a three-day short window
such information when it first becomes publicly approach as adapted from Easton and Harris
available, by observing the company’s share price (1991), Amir et al. (1993) and Barth (1994) (omit-
movements during a short window of three days ting company j and year t subscripts). A differen-
surrounding the company’s results announcement. tial intercept dummy variable AFTER is used to
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

In Hong Kong, the Hong Kong Stock Exchange indicate whether HKAS 40 (2004) or SSAP 13
Listing Rules require its Main Board listed compa- (2000) is adopted during the accounting year.
nies to publish preliminary results the next busi-
ness day after approval by the board of directors ARS = + α1 + α2 AFTER + α3 EARNB (1)
and in any event not later than four months after + α4 ΔEARNB + α5 IPVC
the date upon which the financial period ended. + α6 EARNB*AFTER
The preliminary results include announcements of
+ α7 ΔEARNB*AFTER + α8 IPVC*AFTER
balance sheets and income statements together
with most of the notes to the accounts which have + α9 FIRM SIZE + α10 LEVERAGE
been audited and agreed by the companies’ audi- + α11 CCINDEX + ε
tors and will be published in the annual reports.
The preliminary results announcement day is where:
therefore identified as the day when the informa- ARS is the three-day buy-and-hold10 abnormal
tion on the reporting of gains and losses in fair return (adjusted for dividends and share splits),
value of investment properties first becomes pub- centred around the preliminary results an-
licly available. nouncement day, calculated using the market-
Because during a three-day short window there adjusted return.
are relatively few company-specific events other
than the results announcement, a three-day short- AFTER is a dummy variable to indicate whether
window association between the abnormal returns HKAS 40 (2004) or SSAP 13 (2000) is in effect
and accounting information released in the results during the accounting year. AFTER is set equal
announcement suggests that the accounting infor- to one if HKAS 40 (2004) is adopted for the first
mation is responsible for the abnormal return and time during the accounting year, and zero other-
provides new decision-useful information to in- wise.
vestors. Other information in the preliminary re- EARNB is the earnings before gains and losses.
sults announcement might also be responsible for In the case of HKAS 40 (2004) this is measured
the abnormal return, for example, a change in com- as earnings before gains and losses in fair value
pany strategy. However, by including only compa- of investment properties. In the case of SSAP 13
nies from the property industry, where investment (2000) this is measured as the earnings before
properties comprise a large proportion of company deficits in open market value of investment
assets and fair value gains and losses, a large earn- properties in excess of the revaluation reserve,
or surpluses in open market value of investment
10 While there is disagreement in the literature regarding the
properties in excess of any deficits previously
method for calculating long-run abnormal returns because of charged to the income statement (see footnote 3).
the inherent right-skewed non-normal distribution problem EARNB is scaled by the total market value at the
(Lyon et al., 1999), the choice of method is not so important for first day of the fifth month after the beginning of
the measurement of short-run abnormal returns (Fama, 1998, the accounting year.
Jakobsen and Voetmann, 2003). Because a horizon of three to
five years is referred to in the literature as long-run, both the ΔEARNB is the difference between EARNB in
three-day short-window and 12-month long-window abnormal the current year and EARNB in the prior year,
returns in this study provide short-run returns. Commonly used
abnormal returns in event studies are buy-and-hold abnormal scaled by the total market value at the first day
returns (BHAR) (monthly abnormal returns compounded) and of the fifth month after the beginning of the
cumulative abnormal returns (CAR) (monthly abnormal re- accounting year.
turns summed and averaged). Lyon et al. (1999) suggests
BHAR is suitable in answering the question of whether sample IPVC is the gains and losses in fair value of
firms earned abnormal returns over a particular period of time, investment properties recognised in the income
while CAR is preferred where sample firms persistently earn statement as required by HKAS 40 (2004), or
abnormal monthly returns. Fewer data are collected for BHAR
because the monthly returns compounded are simply the annu- the investment properties’ open market value
al return; BHAR is therefore used in this study. increases and decreases disclosed in the notes
 Vol. 39 No. 2. 2009 109

to the accounts, scaled by the total market value clearly), while Haw et al. (1999) report significant
at the first day of the fifth month after the price reaction using three-day windows. Also, be-
beginning of the accounting year. cause EARNB and ΔEARNB in this study are
earnings and earnings change before investment
FIRM SIZE is the natural logarithm of the book
properties’ value change, they are less likely to be
value of the total assets at the beginning of the
accounting year. transitory (i.e. less likely to be ‘surprises’) and
may show a weaker relationship with the abnormal
LEVERAGE is the ratio of the book value of returns. In contrast, α5 for IPVC may show a
debt to the total assets at the beginning of the ac- stronger relationship. Results from Barth et al.
counting year. (1990 and 1994) give support to these predictions.
CCINDEX is the difference between the Centa-
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

3.2. Long-window abnormal return – unexpected

City Index at the end and the beginning of the earnings association
accounting year, divided by the beginning index This study also extends the returns window to
amount. a longer period of 12 months. While a three-day
The focus of this study is on the significance of short window association between abnormal re-
the differential slope coefficient α8 of the interac- turn and accounting information suggests that the
tion variable IPVC*AFTER which reflects the sig- latter is causing the former, this is not a valid claim
nificance of the differential causal effect of the when the association is evaluated over a long win-
change in the accounting standard requirement. A dow. A longer window opens the returns up to a
positive and significant β8 gives support for the host of other value-relevant events. Share prices
higher informativeness of the changes introduced reflect all available information, not just the ac-
by HKAS 40 (2004) in the presentation of changes counting information. As a result, it cannot be
in fair value of investment properties. claimed that the reported accounting information
The coefficients α6 and α7 of EARNB*AFTER causes the abnormal return during the 12-month-
and ΔEARNB*AFTER reflect the differential long window period. The most that can be claimed
causal effect of the HKAS 40 revision on earnings is that the reported accounting information and the
and earnings change amounts before the open mar- abnormal return are associated.
ket value/fair value change of investment proper- The regression equation (2) for the long window
ties (i.e. EARNB and ΔEARNB). Because the is basically the same as equation (1) for the short
HKAS 40 revision is not concerned with such window, except abnormal return is measured over
earnings and earnings change, coefficients α6 and 12 months rather than three days.
α7 should be of no significance. However, as a re-
sult of the full convergence of HKFRS with IFRS ARL = + β1 + β2 AFTER + β3 EARNB (2)
in Hong Kong in 2005, there are other financial + β4 ΔEARNB + β5 IPVC
reporting requirement changes taking place con- + β6 EARNB*AFTER
currently with HKAS 40 (2004). These other re- + β7 ΔEARNB*AFTER + β8 IPVC*AFTER
porting changes may interact with EARNB and
ΔEARNB to cause an effect on the abnormal re- + β9 FIRM SIZE + β10 LEVERAGE
turns. The sign and significance of a6 and a7 re- + β11 CCINDEX + ε
flect the differential causal value relevance of where:
these other reporting changes. They are, however,
outside the scope of this study and no prediction is ARL is the buy-and-hold11 abnormal return
made about them. (adjusted for dividends and share splits), for the
The coefficients α3 and α4 represent the effects 12 months beginning the first day of the fifth
of EARNB and ΔEARNB in three-day short win- month after the beginning of the accounting
dows when SSAP 13 (2000) is adopted in the fi- year, calculated using market adjusted return.
nancial statements (i.e. when AFTER equals zero).
Although most related prior studies find earnings The Hong Kong Stock Exchange Listing Rules
and/or earnings change to be significant in ex- allow the Hong Kong Main Board listers to pub-
plaining abnormal returns, they do not measure the lish their annual financial statements (as part of the
effects over short windows. When short windows preliminary results announcements) at the latest by
are used, the results are mixed. For example, Amir the last day of the fourth month after the account-
et al. (1993) finds no market reaction to either ing year-end. The 12-month window in this study
earnings or earnings change around five-day an- therefore corresponds to the 12 months ending on
nouncement windows (and interprets this as re- this day (i.e. 12 months beginning eight months
flecting an inability to differentiate the news prior to and ending four months after the account-
ing year).
As with the short window regression, the higher
11 See footnote 10. value-relevance in the longer run of the HKAS 40
 110 ACCOUNTING AND BUSINESS RESEARCH

Table 1
First time adoption of HKAS 40 (2004)

No. of companies
adopting HKAS 40 (2004)
Financial statements year-end for the first time
31 December 2004 3e
31 March 2005 4e
30 April 2005 1e
30 June 2005 7e
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

31 July 2005 2e
31 December 2005 45
31 March 2006 22
30 April 2006 1
30 June 2006 5
31 July 2006 2
Total 92
e Early adoption of HKAS 40 (2004)

(2004) presentation of gains and losses in fair the first time compared to the year(s) when SSAP
value of investment properties in the income state- 13 (2000) is adopted for the last time.
ment is assessed by a positive and significant value On the whole, when companies apply HKAS 40
of β8. (2004) for the first time, they are experiencing
Predictions similar to those for the earlier three- higher earnings and higher market values and of-
day short window (or no prediction) are made for fering their investors higher abnormal returns; this
the other variables, except stronger relationships may be attributable to the strong economy in Hong
are expected because association rather than cau- Kong in 2005 and 2006. The Centa-City Index has
sation is studied in 12-month windows. indeed been increasing during the sample period,
although at a significantly lower rate when HKAS
40 (2004) is applied for the first time. Firm size
4. Empirical results
and Centa-City index changes are both controlled
Although HKAS 40 (2004) allows a free choice
for in this study. Also all independent variables in
between cost and fair value models, all 92 compa-
this study are scaled by the company’s beginning
nies in the initial sample chose to adopt the fair
market value. Results show a significant increase
value model.
in the proportion of investment properties relative
All the 92 companies are retained for data analy-
to total assets, from 0.345 when SSAP 13(2000) is
sis, with extreme variable values verified against
applied to 0.403 when HKAS 40 (2004) is applied.
their sources. Since no procedural errors or ex-
There is also an indication of higher earnings
traordinary events are identified, all the data col-
volatility14 as a result of applying HKAS 40
lected for the 92 companies are retained for the
(2004). The mean gains and losses in fair value of
subsequent analysis.12
investment properties are HK$827m which is al-
Each company is evaluated twice, in two consec-
most equal to the earnings before such gains and
utive accounting years before and after the adop-
losses of HK$848m. In contrast, the mean invest-
tion of HKAS 40 (2004).
ment properties open market value excess deficits
Table 1 describes the distribution of accounting
or surpluses of HK$24m amounts to only 3% of
year-ends, years of last-time following of SSAP 13
(2000) and years of first-time adoption of HKAS
40 (2004) for the 92 companies in this study. 12 Unreported findings show that similar results are ob-
Appendix A details their identities. Most compa- tained if all extreme variable values are excluded.
13 Unreported t-test results show early adopters to have sig-
nies have March 31 or December 31 accounting
year-ends, and adopt HKAS 40 (2004) for the first nificantly higher mean amounts of total assets, investment
properties, total earnings and market values. Unreported re-
time in 2005 or 2006. While HKAS 40 (2004) gression results show similar results when a dummy variable
mandates adoption for annual periods beginning to indicate early adopters is included in the regression equa-
on or after 1 January 2005, 17 companies choose tions.
14 The increased volatility may be limited to our sample pe-
to adopt HKAS 40 (2004) early.13
riod (2004–2006). As volatility should be a function of year-
Tables 2A and 2B contain descriptive statistics over-year change, we need to examine more years in order to
for the 92 sample companies in the study during conclude whether or not an increase in volatility has resulted
the year(s) when HKAS 40 (2004) is adopted for from applying HKAS 40 (2004).
 Vol. 39 No. 2. 2009 111

Table 2A
Descriptive statistics for financial statement and market data (2004–2006)
92 companies

N Mean SD Max Min

Selected market and economic variables
When HKAS 40 (2004) is adopted for the first time
Company’s market value (HK$m) 92 9,511 25,642 178,887 41
Centa-City Index change 92 0.067 0.081 0.315 –0.038
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

When SSAP 13(2000) is adopted for the last time

Company’s market value (HK$m) 92 7,488 22,623 174,070 21
Centa-City Index change 92 0.278 0.096 0.459 0.072
Paired sample t-tests of mean differences (HKAS 40 (2004) over SSAP 13) (t-value in parenthesis)
Company’s market value (HK$m) 92 2,023 ***
(4.114)
Centa-City Index change 92 –0.211 ***
(–20.723)
Selected financial statement variables
When HKAS 40 (2004) is adopted for the first time
Total assets (HK$m) 92 15,996 34,459 196,720 95
Investment properties (HK$m) 92 6,069 15,068 116,733 0
Investment properties to total assets 92 0.403 0.348 1.547 0
Debt to total assets 92 0.182 0.149 0.790 0
Total earnings 92 1,675 3,478 20,038 –571
When SSAP 13(2000) is adopted for the last time
Total assets (HK$m) 92 14,837 32,684 198,860 43
Investment properties (HK$m) 92 5,022 12,654 100,775 0
Investment properties to total assets 92 0.345 0.308 1.289 0
Debt to total assets 92 0.162 0.156 0.536 0
Total earnings 92 1,055 2,569 18,180 –213
Paired sample t-tests of mean differences (HKAS 40 (2004) over SSAP 13) (t-value in parenthesis)
Total assets (HK$m) 92 1,159 ***
(3.662)
Investment properties (HK$m) 92 1,047 ***
(3.704)
Investment properties to total assets 92 0.058 **
(2.405)
Debt to total assets 92 0.020
(1.583)
Total earnings 92 620 ***
(4.049)
aIPFVGL:
Gains and losses in fair value of investment properties (reported in income statement under
HKAS 40 (2004))
bIPOMVEDS: Investment properties’ open market value excess deficits and surpluses (reported in income

statement under SSAP 13)

cIPOMVC: Investment properties’ open market value changes (disclosed in notes to the accounts under

SSAP 13)
* Significant at the 10% level
** Significant at the 5% level
*** Significant at the 1% level
 112 ACCOUNTING AND BUSINESS RESEARCH

Table 2B
Descriptive statistics for regression variables (2004–2006)
92 companies

N Mean SD Max Min

Regression variables
When HKAS 40 (2004) is adopted for the first time
Short-window abnormal return 92 –0.005 0.063 0.217 –0.191
Long-window abnormal return 92 0.156 0.543 1.911 –0.997
Earnings before IPFVGLa (HK$m) 92 848 2,125 11,794 –1,897
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

Earnings change before IPFVGLa (HK$m) 92 105 755 3,184 –2,851

IPFVGLa (HK$m) 92 827 2,052 12,982 –20
When SSAP 13 (2000) is adopted for the last time
Short-window abnormal return 92 –0.013 0.064 0.175 –0.372
Long-window abnormal return 92 0.152 0.563 2.649 –1.158
Earnings before IPOMVEDSb (HK$m) 92 757 1,773 10,601 –216
Earnings change before IPOMVEDSb 92 316 847 6,777 –459
(HK$m)
IPOMVEDSb (HK$m) 92 24 146 886 –784
IPOMVCc (HK$m) 92 312 1,110 8,139 –784
T-tests of mean differences (HKAS 40 (2004) over SSAP 13) (t-value in parenthesis)
Short-window abnormal return 92 0.008
(0.94)
Long-window abnormal return 92 0.004
(0.06)
Earnings before IPFVGLa / IPOMVEDSb 92 91
(HK$m) (1.02)
Earning change before IPFVGLa / 92 –211 *
IPOMVEDSb (HK$m) (–1.66)
IPFVGLa / IPOMVCc (HK$m) 92 515 ***
(3.793)
a
IPFVGL: Gains and losses in fair value of investment properties (reported in income statement under
HKAS 40 (2004))
b
IPOMVEDS: Investment properties’ open market value excess deficits and surpluses (reported in income
statement under SSAP 13)
c
IPOMVC: Investment properties’ open market value changes (disclosed in notes to the accounts under
SSAP 13)
* Significant at the 10% level
*** Significant at the 1% level

Table 3
Earnings volatility

N Mean SD Max Min

Applying HKAS 40 (2004)
Earnings volatility (number of standard 92 0.947 0.728 1.788 –0.167
deviations)
Applying SSAP 13 (2000)
Earnings volatility (number of standard 92 0.419 0.589 1.747 –0.190
deviations)
Paired-sample t-test results:
Paired differences: 0.528 number of standard deviations
t (91) = 4.679
p = 0.000
 Vol. 39 No. 2. 2009 113

the earnings before such excess deficits or surplus- consistent with the results from prior studies using
es of HK$757m. long windows where significance of earnings is
Further indication of higher earnings volatility15 found together with higher overall R2.
is available in Table 3 showing the results of a The coefficient β8 of the interaction variable
paired-sample t-test performed to compare the IPVC*AFTER in equation (2) is positive and sig-
earnings volatility before and after the application nificant at the 10% level (p = 0.069), consistent
of HKAS 40 (2004). Earnings volatility is ex- with the result from the short window regression in
pressed as the number of standard deviations from this study. This result conflicts with that from
a five-year mean (mean of the earnings of the five Owusu-Ansah and Yeoh (2006), who find that the
years ending on the year of HKAS 40 (2004) recognition of unrealised gains in the income state-
application). Earnings volatility is significantly ment is not superior to that in the revaluation re-
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

higher after the adoption of HKAS 40 (2004) serve in terms of value-relevance. However, the
(t = 4.678, p = 0.000). results from this present study are more persuasive
on a number of grounds: (1) Owusu-Ansah and
4.1. Short window event study Yeoh (2006) is based on data from the 1990s be-
Table 4 reports the regression results from the fore the fair value accounting requirement came
estimation of equation (1). Results for the short- into effect, while our study is based on the actual
window event study provide evidence that the reactions from investors to the implementation of
presentation of changes in fair value of investment the fair value model (i.e. HKAS 40 (2004)) com-
properties in the income statements as required by pared to their previous reaction under SSAP 13
HKAS 40 (2004) is more informative to investors (2000); (2) this study has a cleaner test setting as
than the presentation required by SSAP 13 (2000). we use only the earnings capitalisation approach
Investors respond to the information on changes in which focuses on income statement items, where-
fair value in the income statement, as released in as Owusu-Ansah and Yeoh (2006) adopt a market
the results announcement, causing abnormal re- value approach which examines balance sheet
turns. The coefficient α8 of the interaction variable items. This latter approach is not appropriate since
IPVC*AFTER in equation (1) is positive and sig- the values of investment properties (open market
nificant at the 5% level (p = 0.022). value or fair value) reported in balance sheets are
As expected, neither earnings (EARNB) nor similar, before and after the change in accounting
earnings change before investment properties’ open requirements (see footnote 6); (3) the sample com-
market value/changes in fair value (ΔEARNB) is panies in this present study belong to a single in-
significant in explaining the abnormal return with- dustry – property; as detailed earlier, the effects
in the short window when SSAP 13 (2000) is from other information are narrower if the sample
adopted in the financial statements. is restricted to those companies drawn only from
Although the overall R2 is only 1.0%, this is con- the property industry. Property companies are ex-
sistent with the results from prior short-window pected to have more precise value estimates for
studies. their investment properties, but the Owusu-Ansah
All the coefficients are positive except that of in- and Yeoh (2006) sample includes companies from
vestment properties’ value changes (i.e. IPVC), different industries; (4) the findings of this present
which is negative (but not statistically significant). study have greater generality, in that they embrace
Barth et al. (1990) and Barth (1994) also find sim- companies with both value increases and decreas-
ilar negative coefficients for securities market es, while Owusu-Ansah and Yeoh (2006) confine
price gains and losses and interpret them as evi- their sample to those companies with positive
dence of a market that perceives that securities changes in values only.
gains and losses are used to smooth earnings. The coefficient β6 of the interaction variable
EARNB*AFTER in equation (2) is also positive
4.2. Long-window abnormal return – unexpected
and significant at the 10% level (p = 0.092). This
earnings association
gives support for the value-relevance of other fi-
The regression results for the long window ab-
nancial reporting changes that are taking place
normal return and unexpected earnings association
are reported in Table 4. As the window opens concurrently with HKAS 40 (2004) as a result of
wider, the earnings before changes in open market the full convergence of HKFRS with IFRS in
value or fair value (i.e. EARNB) in equation (2) Hong Kong in 2005. These other reporting
also become significant at the 5% level and the changes interact with the earnings before changes
overall adjusted R2 increases to 17.7%. This is in the fair value of open market value or fair value
of investment properties (i.e. EARNB) to result in
a stronger association with abnormal returns.
15 We also test the change in volatility after controlling for
As with the results from equation (1) for the
the change in the market-wide property price levels (measured short window, all the coefficients are positive ex-
by the Centa-City Index). Similar results are obtained. cept those of IPVC and ΔEARNB*AFTER, which
 114 ACCOUNTING AND BUSINESS RESEARCH

Table 4
Regression results

ARS Equation (1) ARL Equation (2)

Intercept –0.101 * –1.561 ***
(–1.72) (–3.37)
AFTER 0.009 0.121
(0.57) (0.99)
EARNB 0.010 0.327 **
(0.51) (2.19)
Δ EARNB
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

0.000 0.168
(0.00) (1.29)
IPVC –0.024 –0.076
(–1.49) (–0.60)
EARNB *AFTER 0.025 0.439 *
(0.77) (1.70)
Δ EARNB *AFTER 0.001 –0.230
(0.05) (–1.34)
IPVC *AFTER 0.062 ** 0.392 *
(2.31) (1.83)
FIRM SIZE 0.003 0.066 ***
(1.28) (3.12)
LEVERAGE 0.005 0.124
(0.16) (0.49)
CCINDEX 0.047 0.693
(0.85) (1.58)
Adj. R2 0.010 0.177
N 184 184
ARS = three-day buy-and-hold abnormal return (adjusted for dividends and share splits), centred
around the preliminary results announcement day, calculated using the market-adjusted return
ARL = buy-and-hold abnormal return (adjusted for dividends and share splits), for the 12 months be-
ginning the first day of the fifth month after the beginning of the accounting year, calculated
using the market-adjusted return
AFTER = a dummy variable to indicate whether HKAS 40 (2004) or SSAP 13 is in effect during the ac-
counting year. AFTER is set equal to one if the new HKAS 40 (2004) is adopted for the first
time during the accounting year, and zero otherwise
EARNB = earnings before gains and losses in fair value of investment properties recognised under the
new HKAS 40 (2004), OR the earnings before investment properties’ open market deficits in
excess of the revaluation reserve balance or investment properties’ subsequent open market
value surpluses in excess of the deficits previously charged to the income statement, recognised
under the old superseded SSAP 13. EARNB is scaled by the total market value at the first day
of the fifth month after the beginning of the accounting year
ΔEARNB = difference between EARNB in the current year and EARNB in the prior year, scaled by the
total market value at the first day of the fifth month after the beginning of the accounting year
IPVC = gains and losses in fair value of investment properties recognised in the income statement as
required by the new HKAS 40 (2004), OR the investment properties’ open market value in-
creases and decreases disclosed in the notes to the accounts, scaled by the total market value
at the first day of the fifth month after the beginning of the accounting year
FIRM SIZE = natural logarithm of the book value of the total assets at the beginning of the accounting year
LEVERAGE = ratio of the book value of debt to the total assets at the beginning of the accounting year
CCINDEX = difference between the Centa-City Index at the end and at the beginning of the accounting year,
divided by the beginning index amount
* significant at the 10% level
** significant at the 5% level
*** significant at the 1% level
 Vol. 39 No. 2. 2009 115

are negative (but not statistically significant). in the income statement, compared to reporting
such changes in the revaluation reserve. This study
4.3. Sensitivity analysis informs this debate by providing evidence on the
If the earnings partitioning approach developed value-relevance of the presentation of changes in
in Barth (1994) is to be followed exactly in this fair value in the income statement for Hong Kong
study, the variable IPVC (changes in fair value of listed companies in the properties industry. Results
investment properties) should be equal to the ex- show a significant market price reaction to invest-
cess open market value deficits or surpluses recog- ment properties’ fair value change information as
nised in the income statement for the accounting included in companies’ annual results announce-
year when SSAP 13 (2000) is followed. This study ments. Results also show a significant association
has chosen to use the open market value increases between the market-adjusted annual share returns
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

or decreases (as disclosed in the notes to the ac- and the presentation of the investment properties’
counts), in order to be consistent with the fair fair value change in the income statement. These
value gains and losses that are used to measure the results strongly suggest that investors appear to
variable IPVC when HKAS 40 (2004) is adopted. place more value on HKAS 40 (2004)’s presenta-
A sensitivity analysis is conducted using the ex- tion of changes in fair value of investment proper-
cess open market value deficits or surpluses when ties in the income statement, when compared with
SSAP 13 (2000) is followed and similar results are the presentation in the revaluation reserve under
obtained. SSAP 13 (2000). The results also support the exist-
Another sensitivity analysis is conducted using ing literature on the value-relevance of presenta-
the Hong Kong Hang Seng Composite Index as the tion locations of accounting amounts in general.
proxy for equity market return in Hong Kong, and The results from this study also have implica-
again, similar results are found. The Hong Kong tions for companies around the world that prepare
Hang Seng Composite Index covers 90% of the their financial statements using International
market capitalisation of the shares listed on the Financial Reporting Standards.
Main Board of SEHK and there are currently 200 However, because all the companies in our sam-
constituent shares in this index. ple choose to adopt the fair value model, we do not
have a control group of companies that do not
5. Conclusions adopt the fair value model. We cannot therefore
Part of the debate about the adoption of fair value eliminate the possibility that our results are driven
accounting for investment properties is on the by other events happening at the same time as the
value-relevance of presenting changes in fair value adoption of HKAS 40 (2004).

Appendix A
92 companies in the sample (184 firm-standard)

First time
Last time adoption of
SEHK Accounting following HKAS 40
Code year-end SSAP 13 (2004)
1 1 Cheung Kong (Holdings) Ltd 31 December 2003 2004
2 154 Beijing Development (Hong Kong) Ltd 31 December 2003 2004
3 758 Junefield Department Store Group Ltd 31 December 2003 2004
4 35 Far East Consortium International Ltd 31 March 2004 2005
5 172 Goldbond Group Holdings Ltd 31 March 2004 2005
6 277 Tern Properties Co Ltd 31 March 2004 2005
7 412 Heritage International Holdings Ltd 31 March 2004 2005
8 735 Oriental Investment Corporation Ltd 30 April 2004 2005
9 10 Hung Lung Group Ltd 30 June 2004 2005
10 12 Henderson Land Development Company Ltd 30 June 2004 2005
11 83 Sino Land Company Ltd 30 June 2004 2005
12 97 Henderson Investment Ltd 30 June 2004 2005
13 101 Hang Lung Properties Ltd 30 June 2004 2005
14 131 Cheuk Nang (Holdings) Ltd 30 June 2004 2005
15 247 Tsim Sha Tsui Properties Ltd 30 June 2004 2005
16 488 Lai Sun Development Co Ltd 31 July 2004 2005
17 1125 Lai Fung Holdings Ltd 31 July 2004 2005
18 14 Hysan Development Company Ltd 31 December 2004 2005
 116 ACCOUNTING AND BUSINESS RESEARCH

Appendix A
92 companies in the sample (184 firm-standard) (continued)

First time
Last time adoption of
SEHK Accounting following HKAS 40
Code year-end SSAP 13 (2004)
19 24 Burwill Holdings Ltd 31 December 2004 2005
20 28 Tian An China Investment Company Ltd 31 December 2004 2005
21 34 Kowloon Development Company Ltd 31 December 2004 2005
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

22 41 Great Eagle Holdings Ltd 31 December 2004 2005

23 56 Allied Properties (HK) Ltd 31 December 2004 2005
24 66 MTR Corporation Ltd 31 December 2004 2005
25 68 Lee Hing Development Ltd 31 December 2004 2005
26 75 Y.T. Reality Group Ltd 31 December 2004 2005
27 89 Tai Sang Land Development Ltd 31 December 2004 2005
28 106 Shenzhen Hing-Tech Holdings Ltd 31 December 2004 2005
29 115 Grand Field Group Holdings Ltd 31 December 2004 2005
30 123 Guangzhou Investment Company Ltd 31 December 2004 2005
31 127 Chinese Estate Holdings Ltd 31 December 2004 2005
32 132 China Investment Holdings Ltd 31 December 2004 2005
33 141 Great China Holdings Ltd 31 December 2004 2005
34 156 Lippo China Resources Ltd 31 December 2004 2005
35 169 China Fair Land Holdings Ltd 31 December 2004 2005
36 171 Silver Grant International Industries Ltd 31 December 2004 2005
37 173 K. Wah International Holdings Ltd 31 December 2004 2005
38 184 Keck Seng Investments (Hong Kong) Ltd 31 December 2004 2005
39 201 Magnificent Estates Ltd 31 December 2004 2005
40 219 Shun Ho Technology Holdings Ltd 31 December 2004 2005
41 230 Onfem Holdings Ltd 31 December 2004 2005
42 242 Shun Tak Holdings Ltd 31 December 2004 2005
43 251 SEA Holdings Ltd 31 December 2004 2005
44 257 China Everbright International Ltd 31 December 2004 2005
45 258 Tomson Group Ltd 31 December 2004 2005
46 286 G-Prop (Holdings) Ltd 31 December 2004 2005
47 366 Luks Industrial (Group) Ltd 31 December 2004 2005
48 373 Allied Group Ltd 31 December 2004 2005
49 431 Greater China Holdings Ltd 31 December 2004 2005
50 588 Beijing North Star Co Ltd 31 December 2004 2005
51 604 Shenzhen Investment Ltd 31 December 2004 2005
52 617 Paliburg Holdings Ltd 31 December 2004 2005
53 635 Playmates Holdings Ltd 31 December 2004 2005
54 649 Shimao International Holdings Ltd 31 December 2004 2005
55 683 Kerry Properties Ltd 31 December 2004 2005
56 688 China Overseas Land & Investment Ltd 31 December 2004 2005
57 730 Shougang Concord Grand (Group) Ltd 31 December 2004 2005
58 878 Soundwill Holdings Ltd 31 December 2004 2005
59 898 Multifield International Holdings Ltd 31 December 2004 2005
60 1200 Midland Holdings Ltd 31 December 2004 2005
61 1207 Shanghai Real Estate Ltd 31 December 2004 2005
62 2355 Baoye Group Company Ltd 31 December 2004 2005
63 20 Wheelock and Company Ltd 31 March 2005 2006
64 22 Mexan Ltd 31 March 2005 2006
65 49 Wheelock Properties Ltd 31 March 2005 2006
66 88 Tai Cheung Holdings Ltd 31 March 2005 2006
67 129 Asia Standard International Group Ltd 31 March 2005 2006
68 160 Hon Kwok Land Investment Company Ltd 31 March 2005 2006
69 163 Emperor International Holdings Ltd 31 March 2005 2006
70 164 Premium Land Ltd 31 March 2005 2006
71 166 New Times Group Holdings Ltd 31 March 2005 2006
 Vol. 39 No. 2. 2009 117

Appendix A
92 companies in the sample (184 firm-standard) (continued)

First time
Last time adoption of
SEHK Accounting following HKAS 40
Code year-end SSAP 13 (2004)
72 202 Interchina Holdings Co Ltd 31 March 2005 2006
73 213 National Electronics Holdings Ltd 31 March 2005 2006
74 224 Pioneer Global Group Ltd 31 March 2005 2006
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

75 237 Safety Godown Co Ltd 31 March 2005 2006

76 278 Wah Ha Realty Company Ltd 31 March 2005 2006
77 287 Winfair Investment Company Ltd 31 March 2005 2006
78 298 Chuang’s China Investments Ltd 31 March 2005 2006
79 480 HKR International Ltd 31 March 2005 2006
80 499 Hycomm Wireless Ltd 31 March 2005 2006
81 711 Chun Wo Holdings Ltd 31 March 2005 2006
82 1036 Winsor Properties Holdings Ltd 31 March 2005 2006
83 1124 Costal Greenland Ltd 31 March 2005 2006
84 2340 Synergies Holdings Ltd 31 March 2005 2006
85 288 Berjaya Holding (HK) Ltd 30 April 2005 2006
86 16 Sun Hung Kai Property Ltd 30 June 2005 2006
87 17 New World Development Co Ltd 30 June 2005 2006
88 54 Hopewell Holdings Ltd 30 June 2005 2006
89 659 NWS Holdings Ltd 30 June 2005 2006
90 917 New World China Land Ltd 30 June 2005 2006
91 193 Capital Estate Ltd 31 July 2005 2006
92 1191 China Rich Holdings Ltd 31 July 2005 2006

References Accounting Research, 36: 199–233.

Aboody, D., Barth, M.E. and Kasznik, R. (1999). Barth, M.E., Clinch, G. and Shibano, T. (2003). ‘Market
‘Revaluations of fixed assets and future firm perform- effects of recognition and disclosure’. Journal of
ance’. Journal of Accounting Research, 36: 161–191. Accounting Research, 41(4): 581–609.
Amir, E., Harris, T.S. and Venuti, E.K. (1993). ‘A compar- Barth, M.E., Landsman, W.R. and Wahlen, J.M. (1995).
ison of the value-relevance of US versus non-US GAAP ‘Fair value accounting: effects on banks’ earnings volatil-
accounting measures using form 20-F reconciliations’. ity, regulatory capital, and value of contractual cash
Journal of Accounting Research, 31: 230–264. flows’. Journal of Banking and Finance, 19 (June):
Ball, R. and Brown, P. (1968). ‘An empirical evaluation of 577–605.
accounting income numbers’. Journal of Accounting Beaver, W.H. (1981). Financial Reporting: An Accounting
Research, 6: 159–178. Revolution. Prentice Hall.
Ball, R. and Kothari, S.P. (1991). ‘Security returns around Biddle, G.C., Seow, G.S. and Siegel, A.F. (1995). ‘Relative
versus incremental information content’. Contemporary
earnings announcements’. Accounting Review, 66(4):
Accounting Research, 12(1–1): 1–23.
718–738.
Brown, L. and Warner, J.B. (1980). ‘Measuring security
Barth, M.E. (1994). ‘Fair value accounting: evidence from price performance’. Journal of Financial Economics, 8:
investment securities and the market valuation of banks’. 205–258.
Accounting Review, 69(1): 1–25. Carroll, T.J., Linsmeier, T.J. and Petroni, K.R. (2003). ‘The
Barth, M.E. (2006). ‘Including estimates of the future in reliability of fair value versus historical cost information:
today’s financial statements’. Accounting Horizons, evidence from closed-end mutual funds’. Journal of
20(3): 271–286. Accounting, Auditing and Finance, (Winter): 1–24.
Barth, M.E., Beaver, W.H. and Landsman, W.R. (1996). Chambers, D., Linsmeier, T.J., Shakespeare, C. and
‘Value-relevance of banks’ fair value disclosures under Sougiannia, T. (2006). ‘An evaluation of SFAS No. 130
SFAS No. 107’. Accounting Review, (October): 513–537. comprehensive income disclosures’. Available at SSRN:
Barth, M.E., Beaver, W.H. and Landsman, W.R. (2001). http://ssrn.com/abstract=646622.
‘The relevance of the value relevance literature for finan- Danbolt, J. and Rees, W. (2008). ‘An experiment in fair
cial accounting standard setting: another view’. Journal of value accounting: UK investment vehicles’. European
Accounting and Economics, 31: 77–104. Accounting Review, 17(2): 271–303.
Barth, M.E., Beaver, W.H. and Wolfson, M.A. (1990). Dietrich, J.R., Harris, M.S. and Muller, K.A. (2001). ‘The
‘Components of earnings and the structure of bank share reliability of investment property fair value estimates’.
prices’. Financial Analysts Journal, 46(3): 53–60. Journal of Accounting and Economics, 30(2): 125–158.
Barth, M.E. and Clinch, G. (1998). ‘Revalued financial, Easton, P. and Harris, T. (1991). ‘Earnings as an explanato-
tangible, and intangible assets: associations with share ry variable for returns’. Journal of Accounting Research,
prices and non-market-based value estimates’. Journal of (Spring): 19–36.
 118 ACCOUNTING AND BUSINESS RESEARCH

Eccher, E.A., Ramesh, K. and Thiagarajan, S.R. (1996). ‘Fair Investment Property. International Accounting Standards.
value disclosures by bank holding companies’. Journal London: IASCF.
of Accounting and Economics, (August–December): Jakobsen, J. and Voetmann, T. (2003). ‘Post-acquisition
79–117. performance in the short and long run: evidence from the
Espahbodi, H., Espahbodi, P., Rezaee, Z. and Tehranian, H. Copenhagen Stock Exchange 1993–1997’. European
(2002). ‘Stock price reaction and value relevance of Journal of Finance, 9(4): 323–342.
recognition versus disclosure: the case of stock-based KPMG China and Hong Kong. (2004). Property and
compensation’. Journal of Accounting and Economics, Infrastructure. KPMG.
33: 343–373. Landsman, W.R. (2007). ‘Is fair value accounting informa-
Fama, E.F. (1998). ‘Market efficiency, long-term returns tion relevant and reliable? Evidence from capital market
and behavioral finance’. Journal of Financial Economics, research’. Accounting and Business Research Special
49: 283–306. Issue: International Accounting Policy Forum: 19–30.
Fan, J.P.H. and Wong, T.J. (2002). Corporate ownership Lee, Y., Petroni, K. and Shen, M. (2006). ‘Cherry picking,
Downloaded by [Universitas Dian Nuswantoro], [Ririh Dian Pratiwi SE Msi] at 20:11 29 September 2013

structure and the informativeness of accounting earnings financial reporting quality, and comprehensive income re-
in East Asia. Journal of Accounting and Economics, 33: porting choices: the case of property-liability insurers’.
401–425. Contemporary Accounting Research, 23(3): 665–700.
Haw, I.M., Qi, D.A. and Wu, W. (1999). ‘Value relevance Lyon, J.D., Barber, B.M. and Tsai, C.L. (1999). ‘Improved
of earnings in an emerging capital market: the case of methods for tests of long-run abnormal returns’. Journal
A-shares in China’. Pacific Economic Review, 4(3): of Finance, 54(1): 165–201.
337–347. Maines, L.A. and McDaniel, L.S. (2000). ‘Effects of com-
Hirst, D.E. and Hopkins, P.E. (1998). ‘Comprehensive in- prehensive-income characteristics on non-professional in-
come disclosures and analysts’ valuation judgments’. vestors’ judgments: the role of financial-statement
Journal of Accounting Research, 36 (Supplement): 47–75. presentation format’. Accounting Review, 75: 179–207.
Hirst, D.E., Hopkins, P.E. and Wahlen, J.M. (2004). ‘Fair Maines, L.A. and Wahlen, J.M. (2006). ‘The nature of ac-
values, income measurement, and bank analysts’ risk counting information reliability: inferences from archival
and valuation judgements’. Accounting Review, 79(2): and experimental research’. Accounting Horizons, 20(4):
453–472. 399–425.
Hodder, L.D., Hopkins, P.E. and Wahlen, J.M. (2006). McBride, P. (2006). ‘Making a statement: what are finan-
‘Risk-relevance of fair-value income measures for com- cial statements designed to do?’ A Plus, (December):
mercial banks’. Accounting Review, 81(2): 337–375. 42–44.
Hong Kong Institute of Certified Public Accountants Muller, K.A. and Riedl, E.J. (2002). ‘External monitoring
(HKICPA). (2004). HKAS 40 Investment Property. Hong of property appraisal estimates and information asymme-
Kong Financial Reporting Standards. Hong Kong: try’. Journal of Accounting Research, 40: 865–881.
HKICPA. (Available at http://www.hkicpa.org.hk/ Nelson, K.K. (1996). ‘Fair value accounting for commer-
hksaebk/HKSA_Members_ Handbook_Master/volumeII/ cial banks: an empirical analysis of SFAS No. 107’.
hkas40.pdf.) Accounting Review, 71 (April): 161–182.
Hong Kong Securities Institute (HKSI). (2006). HKSI reg- New Zealand Society of Accountants (NZSA, which in
ulatory update: HKAS 40 on trial. Hong Kong Securities, 1996 was renamed as Institute of Chartered Accountants
(September/October): 19–20. of New Zealand with an operating name of New Zealand
Hong Kong Society of Accountants (HKSA, which in Institute of Chartered Accountants, NZICA) (1989).
2004 was renamed as Hong Kong Institute of Certified SSAP 17 Accounting for Investment Properties and
Public Accountants, HKICPA) (2000). SSAP 13 Properties Intended for Sale. Statement of Standard
Accounting for Investment Properties. Statement of Accounting Practice. New Zealand: NZSA. (Available
Standard Accounting Practice. Hong Kong: HKSA. at http://www.nzica.com/ AM/Template.cfm?Section
(Available at http://www.hkicpa.org.hk/professional =Financial_Reporting_Standards_files&Template=/CM/
technical/accounting/rm/withdrawn/ssap13.pdf.) ContentDisplay.cfm&ContentID=13486.)
Hong Kong Stock Exchange (HKSE). (2006). HKEx Fact Owusu-Ansah, S. and Yeoh, J. (2006). Relative value rele-
Book 2006: 20–1. vance of alternative accounting treatment for unrealized
Hong Kong Stock Exchange (HKSE). (2005). HKEx Fact gains: implications for the IASB. Journal of International
Book 2005: 22–3. Financial Management and Accounting, 17(3): 232–255.
International Accounting Standards Committee (IASC, Penman, S.H. (2007). Financial reporting quality: is fair
which in 2001 was renamed as International Accounting value a plus or a minus? Accounting and Business
Standards Board, IASB). (2000). IAS 40 Investment Research Special Issue: International Accounting Policy
Property. International Accounting Standards. London: Forum: 33–44.
IASC. Robinson, D. and Burton, D. (2004). ‘Discretion in finan-
International Accounting Standards Committee Foundation. cial reporting: the voluntary adoption of fair value
(IASCF). (2008a). IAS 40 Investment Property. accounting for employee stock options’. Accounting
International Accounting Standards. London: IASCF. Horizons, 18(2): 97–108.
International Accounting Standards Committee Singleton-Green, B. (2007). ‘Fair Value Accounting’.
Foundation. (IASCF). (2008b). Framework for the Practice Society Alert, (January): Institute of Chartered
Preparation and Presentation of Financial Statements. Accountants in English and Wales (ICAEW).
London: IASCF. Sloan, R.G. (1999). ‘Evaluating the reliability of current
International Accounting Standards Committee Foundation. value estimates’. Journal of Accounting and Economics,
(IASCF). (2008c). IASC Basis for Conclusions on IAS 40 26: 193–200.