value relevance of international accounting standards harmonization: evidence from a- and b-share...

Journal of International Accounting, Auditing and Taxation14 (2005) 79–103

Value relevance of international accounting standardsharmonization: Evidence from A- and

B-share markets in China

Z. Jun Lina,∗, Feng Chenb

a Department of Accountancy and Law, Hong Kong Baptist University, Kowloon Tong, Kowloon, Hong Kongb Division of Market Surveillance, Shenzhen Stock Exchange, PR China

Abstract

Applying both the price-levels model and the lagged-price-deflated returns model, we investigated theincremental value relevance of the reconciliation of accounts from the Chinese Accounting Standards (CAS)to the International Accounting Standards (IAS) by those Chinese listed companies that have simultaneouslyissued A-shares and B-shares. In addition, we examined the usefulness of accounting numbers (earnings andbook values) and their value relevance to the A- and B-share markets in China. The study finds that earningsand book values of owners’ equity determined under CAS are more relevant accounting information forthe purpose of determining the prices of A- and B-shares. The CAS-based earnings changes were reflectedin stock returns in the B-share market, while the CAS-based earnings were closely associated with stockreturns in the A-share market. However, the study found that the reconciliation of earnings and book valuesfrom CAS to IAS basis is partially value-relevant, mainly to stock prices in the B-share market, while theearnings reconciliation is generally not value-added to stock returns in either the A- or the B-share market.The study results suggest that accounting numbers based on domestic accounting standards, in contrast toIAS, are more value-relevant in the Chinese stock market at present.© 2005 Elsevier Inc. All rights reserved.

Keywords: Value relevance; International accounting harmonization; Chinese stock market; Chinese accounting

1. Introduction

Stock companies reappeared in the Chinese economy in the late 1980s. Two stock exchanges,the Shanghai Stock Exchange (SHSE) and the Shenzhen Stock Exchange (SZSE), were estab-lished in late 1990 and early 1991, respectively, followed by a rapid expansion of capital market

∗ Corresponding author. Tel.: +852 3411 7537; fax: +852 3411 5581.E-mail address: [email protected] (Z. Jun Lin).

1061-9518/$ – see front matter © 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.intaccaudtax.2005.08.001

80 Z. Jun Lin, F. Chen / Journal of International Accounting, Auditing and Taxation 14 (2005) 79–103

in the country since then (Mookerjee & Yu, 1999; Fung, Lee, & Leung, 2000; Chen & Thomas,2003). Nonetheless, the Chinese stock market is still far from standardized with respect to its regu-latory framework. In particular, financial reporting and information disclosure is underdevelopedcompared with the standards in most industrialized countries. A major problem lies in the lack oftransparent and reliable accounting information to assist investors and other market participantsto effectively make decisions (Chan & Rotenberg, 1999; Aharony, Lee, & Wong. 2000). In theabsence of a sound regulatory system, some Chinese listed companies have released unreliable oreven fraudulent information to the market (Chen, Lee, Rui, Firth, & Gao, 2002). Concerns havebeen raised on the quality of accounting and reporting practices in China, owing to the increasingexposures of accounting scandals in the listed companies in recent years (Chen & Thomas, 2003).1

As a result of the increasing public demand for improving the relevance and reliability ofcorporate disclosure, it has become widely accepted that the Chinese accounting system, whichwas developed to serve the government’s centralized planning and control over the economy,should be overhauled and internationally accepted accounting standards and practices should beadopted (Ge & Lin, 1993; Mills & Cao, 1996; Lin & Chen, 1999). The internationalization ofChinese accounting is driven by a rapid progress of economic restructuring towards a market-oriented economy and the government’s intention of attracting more foreign capital. In particular,following China’s official entry to the World Trade Organization (WTO) in late 2001, the Chineseeconomy has become increasingly integrated with world markets. This development, again, callsfor the harmonization of Chinese accounting standards (CAS) and practices with the internation-ally accepted norms since accounting serves as the language of business.

The International Accounting Standards (IAS) promulgated by the International AccountingStandards Board (IASB)2 are widely regarded as the internationally accepted rule of accountingpractice, and the need to adopt it in China has been gradually recognized (Winkle, Huss, & Chen,1994; Xiao & Pan, 1997; Lin, Chen, & Tang, 2001). In fact, since the mid 1980s, the Chineseaccounting system has been undergoing continuing reforms aimed at creating a set of accountingstandards to replace the original rule-based accounting regulations (Davidson, Gelardi, & Li, 1996;Lin & Chen, 2000). At the same time as theAccounting Standards for Business Enterprises (ASBE)was introduced in 1993, the Chinese government also formulated theAccounting Regulationfor Companies Adopting Share-Capital Systems (amended in 1998), which applies exclusivelyto listed companies. Since 1997 the government has further promulgated several transaction-basedPractical Accounting Standards (PASs). These new accounting standards or regulationsare modeled on the IAS, although a substantial gap remains between the Chinese accountingstandards and IAS (Xiao & Pan, 1997; Chen, Gul, & Su, 1999; Lin et al., 2001).3

1 Among them, “Ying-guan-xia,” “Oriental Electronic Co.,” and “Zhen Bai Weng” are particularly notorious cases,which have caused substantial losses to investors and attracted widespread criticisms of the existing accounting andreporting practices of listed companies in China.

2 IASB was called the International Accounting Standards Committee (IASC) before 2002. Thereafter, the ‘InternationalAccounting Standards’ have been retitled to ‘International Financial Reporting Standards (IFRSs).

3 For instance,Chen et al. (1999)reported, on average, earnings determined by CAS were 20–30% higher than the IASamounts. After restating to IAS-based earnings, about 15% of the B-share companies changed from reporting profits tolosses, suggesting the difference between CAS and IAS was substantial. They identified 12 major items contributing to thedifferences in the reported incomes, which were attributable mainly to three areas; 1) varied standard requirements in CASand IAS, 2) effects of specific government policy such as mandatory consolidation of the original ‘dual rate’ system offoreign exchange and subsidies for employee housing, and 3) difference in the judgement (application of the ‘conservatism’convention in particular) made by domestic auditors and the ‘Big 5’ who applied CAS and IAS, respectively, in auditingChinese listed companies.

Z. Jun Lin, F. Chen / Journal of International Accounting, Auditing and Taxation 14 (2005) 79–103 81

The stock market in China is segmented. There is one market for A-shares, which are tradedamong domestic investors, and another for B-shares, which are issued exclusively to overseasinvestors.4 The segregation of the two markets reflects the government’s motivation of havinga more standardized market, although at a small scale, to directly attract foreign investors. TheB-share market is also treated as a reform experimentation in order to provide references for thedevelopment of the immature A-share market.5 The government regulatory authorities have setout different accounting rules and information disclosure requirements for the A- and B-sharemarkets. Companies issuing A-shares must prepare their financial statements based on CAS. B-share companies should prepare their financial statements following IAS and be audited by the Big5 international auditing firms. In practice, those companies issuing both A-shares and B-sharesare allowed to release their primary financial statements based on CAS and, simultaneously, thereconciled key accounting figures from CAS to IAS (Bao & Chow, 1999; Abdel-Khalik, Wong;& Wu, 1999; Jiang, 2001).

The segregation of A- and B-share markets and the different accounting and disclosure require-ments provide an opportunity to test the value relevance of accounting numbers prepared underthe two sets of accounting standards and, further, to evaluate empirically the effectiveness ofharmonizing CAS with IAS.6 Such kind of studies should be able to examine the necessity andfeasibility of adopting IAS in China. We thus, investigated the value relevance of accountingfigures determined by CAS and IAS in the emerging Chinese stock market. Regressions were runto test the association of accounting numbers (earnings and book values of owners’ equity) andmarket variables (price levels and stock returns) for those Chinese listed companies that issuedboth A-shares and B-shares during 1995–2000, in the unique context of the Chinese stock market.

By employing price-levels and price-deflated returns models, we tested empirically whetherCAS or IAS is more closely associated with stock prices and returns in the A-share and B-sharemarkets in China, and whether the IAS reconciliation has incremental information content. Wefound that, at the aggregate level, earnings and book values determined under CAS are relevantaccounting information to both A- and B-share markets. Although the IAS reconciliation ofbook values is incrementally value-relevant to stock valuation, the reconciliation of earningsis generally not value-added for determining stock returns in the Chinese stock market. In thesubsample comparison of companies listed in the two stock exchanges (i.e., SHSE and SZSE), weobserved that SHSE is more sensitive to the IAS reconciliation than SZSE. In addition, our studyresults indicate that the association of book values with equity valuation is relatively stable inthe B-share market, while the usefulness of earnings has significantly fluctuated and deterioratedover time in both A- and B-share markets. It is argued that the empirical findings of this study

4 B-shares have been issued to overseas institutional and individual investors since 1992 as a means to attract foreigncapital into the Chinese securities market. Domestic institutions and individual investors were not permitted to participatein the B-share market before early 2001 when the Chinese government decided to liberalize it.

5 Contrasting to the A-share market, the trading volume and market capitalization of B-share companies were small butthe operation of B-share market which follow the internationally accepted standards, has generated positive impact onthe development of A-share market in terms of the improvement of market regulation and information disclosure quality.In fact, the Chinese authorities have transferred the regulations of B-share market to A-share market since the late 1990s(including the introduction of new or amended accounting rules) before allowing domestic investors to participate in theB-share market in mid 2001and opening the A-share market for foreign investors in 2004.

6 Most value-relevance studies focus mainly on the association of accounting numbers and market variables in differentcountries. However, differences in accounting numbers may be compounded by other factors stemmed from variedenvironments. Thus, a comparison of the effects of different accounting standards in a single country with differentmarket settings should provide more convincing evidence of value relevance of accounting numbers.


are determined by the unique features of the emerging Chinese stock market and the specificcorporate structures of the Chinese listed companies.

The rest of this paper is organized as follows. A brief description of the development of thecapital market and the financial reporting environment in China is presented in the next section,and some relevant prior studies are reviewed as the theoretical framework of our study design.The research method, model specification, and sample selection are then described, followed bya presentation of the empirical results and the analysis. Finally there is a short conclusion.

2. Study background

2.1. The capital market and financial reporting environment in China

Stock companies were suspended in 1949, following the founding of the People’s Republic,when the new government introduced public (state) ownership of the means of production anda highly-centralized planned economy (Karmel, 1994). Stock companies reappeared in China inthe mid 1980s. In October 1990, the first stock exchange, the SHSE, was opened in Shanghai,followed by the SZSE in early 1991, to allow public listing and trading of stocks and othersecurities in the country (Ge & Lin, 1993; Burke, 1999). The two stock exchanges have expandeddramatically since then in line with the rapid growth of the Chinese economy. By the end of 2000,559 A-share and 55 B-share companies were listed on the SHSE, with a trading volume of RMB¥3,137.39 billion; 449 A-share and 58 B-share companies were listed on the SZSE, with a tradingvolume of RMB ¥2,945.28 billion; and the total amount of market capitalization had reachedRMB ¥4,809.08 billion (US$ 1 = RMB ¥8.11).Table 1presents a summary of the growth of thetwo stock exchanges in China in the last decade. The Chinese stock market is now one of themost rapidly growing capital markets in the world (Friedmann & Sanddorf-Kohle, 2002; Chen &Thomas, 2003).

Substantial differences exist in the structures of Chinese stock companies and securities trans-actions by comparison to those in developed countries (Friedmann & Sanddorf-Kohle, 2002). Forinstance, a majority of Chinese stock companies were restructured from the former State-ownedenterprises (SOEs) and there are three categories of share for most Chinese listed companies:state-owned shares (representing the state’s interest in the stock companies), legal-entity-owned(institutional) shares (representing the stocks owned by other SOEs or economic entities), andpublicly-held shares (representing shares held by either institutional or individual investors). Afurther distinction exists in relation to the publicly-traded shares: A-shares are issued exclusivelyto domestic investors and are denominated in RMB ¥ (Chinese currency); B-shares, denomi-nated in foreign currencies (US$ and HK$ in the SHSE and SZSE, respectively), are issued toand traded exclusively among overseas investors. Companies wishing to issue B-shares must getspecial approval from the government regulatory authorities, in particular the China SecuritiesRegulatory Commission (CSRC). Much more stringent threshold requirements have been set forthe listing or issuing of B-shares. Thus, only a very small portion of the listed companies havebeen permitted to issue B-shares. Before March 2001, A- and B-share markets were completelysegregated, although both the A- and B-shares of the same company entitle the holder to equalvoting rights and claims against the company’s earnings and net assets. Due to the segmentationof A- and B-share markets, there are substantial differences between the two in terms of tradingvolumes and the liquidity of stocks, along with a big gap in the prices of A- and B-shares (Chen,Lee, & Rui. 2001).


Table 1Growth of Chinese stock markets 1992–2000

Year 1992 1993 1994 1995 1996 1997 1998 1999 2000

No. of listed firmsSHSE

A-share 30 101 169 184 287 372 425 471 559B-share 9 22 34 36 42 50 52 54 55

SZSESZSEA-share 24 76 118 127 227 348 400 450 499B-share 9 19 24 34 43 51 54 54 58

Market capitalization (billion RMB)SHSE

A-share 52.06 206.77 248.35 243.37 531.61 903.25 1052.54 1444.07 2659.63B-share 3.79 12.80 11.66 9.20 16.17 18.56 10.05 13.98 33.45

SZSEA-share 45.75 125.10 103.25 87.69 413.24 812.17 877.39 1172.69 2085.94B-share 3.22 8.43 5.80 7.18 23.21 18.94 10.58 16.38 30.06

Trading volume (billion RMB)SHSE

A-share 23.27 230.15 562.67 304.26 902.02 1355.02 1230.35 1685.69 3102.97B-share 1.45 7.87 10.84 6.09 9.46 21.29 8.19 13.97 34.42

SZSEA-share 41.74 126.09 237.64 91.59 1203.21 1674.50 1111.35 1422.33 2924.90B-share 1.66 2.58 1.62 1.70 18.53 21.37 4.46 13.05 20.38

Source: China Securities and Futures Statistical Yearbook 2001.

The regulatory mechanism for the stock market is still under construction in China. Only inrecent years have the CSRC and the two stock exchanges begun paying serious attention to stan-dardizing financial reporting and information disclosure by listed companies (Chen & Thomas,2003). Currently, A-share companies are required to present their financial statements in line withCAS before the end of the fourth month after the year-end (normally the calendar year). Theymust submit their audited financial statements to the regulatory authorities and publish them ina few national newspapers designated by the CSRC. B-share companies are required to presenttheir financial statements in line with IAS to be audited by international accounting firms (e.g., the‘Big 5′) since the participants in the B-share market are overseas investors (CSRC, 1996). Thosecompanies which have issued both A-shares and B-shares usually publish their CAS-based finan-cial statements and their IAS reconciliation of key accounting numbers simultaneously (includinga statement of the differences caused by the different accounting standards and their impact onreported earnings and net assets).7 Following the demands for increasing internationalization ofChinese accounting in recent years, the CSRC has further required that the listed banking and

7 Under the existing regulations, companies issuing both A- and B-shares are allowed to publish CAS-based financialstatements along with IAS reconciliation of main accounting numbers such as operating income, net assets, net income,book value per share and EPS, instead of two sets of financial statements, in the designated national newspapers, for thesake of simplification of public disclosure and convenience of comparing accounting numbers prepared under CAS andIAS.


financial companies, even those issuing only A-shares, must disclose their accounting numbersunder both CAS and IAS after 2001 (CSRC, 2000).8

Government regulatory authorities have beefed up the regulation of information disclosure oflisted companies, especially after the implementation ofThe Securities Law of PRC in 1999.9

While the public has demanded improvements in the relevance and credibility of financial report-ing, a consensus has emerged among Chinese accounting academics and professionals that theaccounting and reporting systems should be thoroughly reformed through expanded use of IAS(Friedmann & Sanddorf-Kohle, 2002). Nonetheless many advocators argued that the interna-tionalization of Chinese accounting should be in an evolutional process with IAS being adoptedgradually in China.10 This idea has won the support of the government regulatory authorities.11 Itis generally believed that the increasing adoption of internationally accepted accounting standardswill not only improve the quality of Chinese accounting but also promote a smooth growth of thestock market in China (Chen & Thomas, 2003).

2.2. Relevant prior studies

Value-relevance of accounting information is one of the main topics in capital market account-ing research (Collins, Maydew, & Weiss, 1997; Holthausen & Watts, 2001; Barth, Beaver, &Landsman, 2001; Kothari, 2001; Frank, 2002). A common feature of this line of research, as sum-marized byKothari (2001), assumes that accounting information is considered value relevant if itis significantly associated with equity market values. The value-relevance studies have generallybeen conducted from two different perspectives, i.e., the relative and the incremental informa-tion content of accounting numbers based on their statistical associations with stock prices orreturns (Biddle, Seow, & Siegel. 1995). The incremental approach examines whether one set ofaccounting measures has information content beyond that provided by the other set of accountinginformation. It applies when one set of accounting measures is viewed as given and an assess-ment is made with respect to the incremental contribution of another set of accounting measures.The relative comparison approach examines which accounting measure has greater informationcontent. It applies when mutually exclusive choices are to be made among various accountingalternatives or when a ranking of the information content of varied accounting measures is made(Harris & Muller, 1999; Holthausen & Watts, 2001).

Several studies investigated the relative informativeness of accounting earnings by comparingthe value-relevance and timeliness of the accounting earnings reported under various sets of

8 On December 30, 2001, CSRC issued theTentative Rules on the Supplement Audits for Companies Issuing A-shareof Stocks, which requires all A-share companies that intend to issue new shares to hire one of the “Big 5” internationalaccounting firms to perform supplementary audits in accordance with IAS. However, this regulation caused seriouscontroversy and has been resisted by Chinese auditors. Its implementation has been delayed.

9 TheSecurities Law was enacted by the Standing Committee of the National People’s Congress (NPC) in December1998 after more than 6 years of preparation. This Law contains some provisions on basic principles and requirements oninformation disclosure for public listing and trading of securities in China.10 Therefore, the introduction of IAS to B-share companies has been treated as an experiment before its extension to

A-share companies later on.11 As pointed out by the Assistant Minister of the Ministry of Finance (MOF) at the central government, MOF has been

working continuously in recent years to promote the internationalization of Chinese accounting (Shanghai SecuritiesDaily, December 11, 2001). The Chairman of the China Securities Regulatory Commission (CSRC) also stated recentlythat the internationalization of Chinese accounting would be a necessary condition for the development of the securitiesmarket in China (Hong Kong Economics Journal, November 16, 2002).


accounting standards (Alford, Jones, Leftwich, & Zmijewski, 1993; Joos & Lang, 1994; Barth,Clinch, & Shibano, 1999). Generally the study results documented that the reported earnings basedon the national (domestic) GAAP had greater information content than the foreign GAAP in termsof their association with stock returns (Bodnar & Weintrop, 1997; Holthausen & Watts, 2001). Itis argued that the phenomenon may be explained by factors such as institutional differences in taxlaw, corporate governance structure, intercorporate securities holdings, that the unique aspects ofbusiness environment in the home country are better reflected in the accounting numbers derivedfrom the national GAAP (Chan & Seow, 1996; Barth et al., 2001).

Some researchers have applied the reconciliation measurements between two sets of account-ing standards as additional variables to analyze the incremental value-relevance of accountinginformation and have reported mixed results. Some studies found the reconciliation informa-tion to be value-relevant in terms of owners’ equity and earnings reconciliation (Amir, Harris,& Venuti, 1993; Harris & Muller, 1999) but others reported that the reconciliation from foreignGAAP to the US GAAP or IAS was not value-relevant. Investors were indifferent to the GAAPdiscrepancies even if the magnitude of these differences relative to firms’ earnings or book valueswas material (Bandyopadhyay, Hanna, & Richardson, 1994; Chan & Seow, 1996; Rees & Elgers,1997; Niskanen, Kinnunen, & Kasanen, 2000; Frank, 2002).

The stability of the value-relevance of earnings and book values over time also has beeninvestigated.Collins et al. (1997)argued that the incremental value-relevance of earnings haddeclined in the US market between 1953 and 1993 while the value-relevance of book values ofowner’s equity increased over that time. The value-relevance of combined earnings and bookvalues did not decline and even increased slightly over the same period.Ely and Waymire, (1999)andFrancis and Schipper (1999)obtained similar evidence for the periods of 1927–1993 and1952–1994, respectively. However,Lev and Zarowin (1999)documented that the usefulness ofearnings, cash flows and book values of equity, all deteriorated over the 1977–1996 period inthe US capital market. In addition,Wallace (2000)found that the value relevance to evalua-tion of firms’ safety, financial strength, price stability and technical ranking had all deterioratedsignificantly.

2.3. Relevant studies on accounting information and the Chinese capital market

Because the re-emergence of China’s capital market has occurred relatively recently, fewempirical studies have been conducted on the association between accounting numbers and marketperformance in China.Abdel-Khalik et al. (1999)examined the varied market structures and therelated information disclosure environment for A- and B- shares in the Chinese stock market.Through a general comparison, they concluded that the disclosure requirements for the A-sharemarket (e.g., the reporting standards and the extent of information disclosed) were relatively low.However, the information environment for the B-share market was relatively standardized. Theycontend that the tangible outcomes of such different information environments for A- and B-share markets include a significant price gap between A- and B- shares. In addition, they appliedSharpe’s market model and Beaver’s U-statistic to investigate the association between accountinginformation and market returns in 1994–1995 at SHSE and SZSE. They found that IAS-basedaccounting earnings reflected the changes in B-share prices while CAS-based accounting earningshad no observable effect on changes in A-share prices.

Bao and Chow (1999)examined the relative value relevance of IAS-based accounting infor-mation to equity valuation in the B-share market during 1993–1996 in China. They investigatedwhether CAS-based numbers or IAS adjustment amounts are more closely associated with B-


share prices. They found that the reported earnings based on IAS were significantly related tostock prices while book values were not. Their study reveals that the IAS-based reporting modelhas additional explanatory power beyond the CAS-based reporting model. They argue that earn-ings and book values based on IAS jointly have greater information content to B-shares than doCAS-based accounting numbers.

More recently,Chen et al. (2001)investigated whether domestic investors in the Chinese stockmarket perceived accounting information based on Chinese GAAP to be value-relevant in theA-share market in the period of 1991–1998. They reported that accounting information is value-relevant despite the young age of the market and the perception of inadequate accounting andfinancial reporting in China. They determined that the price model provides stronger evidenceof the value relevance than the returns model. Chen et al. partitioned their sample into A- shareand AB-share groups to test whether Chinese GAAP-based accounting numbers have variedinformation content to the two groups of companies that are subject to different reporting anddisclosure requirements. They found the value relevance to be higher for companies issuing onlyA-shares than companies issuing AB-share. They attributed this finding to AB-share companiesthat have alternative sources of competing information other than financial reporting, althoughthis contention is short of convincing empirical evidence.

3. Study design and data

It should be pointed out that the studies ofAbdel-Khalik et al. (1999)andBao and Chow(1999) have limitations since their sample sizes are small and the test periods are relativelyshort. In particular, the financial reporting practices were far from satisfactory in the periodin which they were testing since the accounting and information disclosures of Chinese listedcompanies diverged substantially from the international norms, and the Chinese stock marketwas less standardized at its initial stage (Chan & Rotenberg, 1999). For instance,Abdel-Khaliket al. (1999)indicated that public information could be found for less than 25% of the listedA-share companies in 1993 and for only 2% of the listed B-shares companies. In 1994 thesepercentages had increased to about 50% and 12% respectively. The test results for those earlyyears of the Chinese stock market may not be sufficiently representative. In addition,Abdel-Khalik et al., (1999)used only a market returns model whileBao and Chow (1999)applied aprice-levels model to test the relative value relevance of accounting numbers. AlthoughChen etal. (2001)employed both the price-levels and returns models, their regression analysis focusedonly on the A-share market. For their A-share and AB-share comparison, only the Chinese-GAAP based information was used in the partitioned regressions, merely an indirect test of therelative value relevance of accounting information in the A-share market. In fact their studyhas not tested the incremental value relevance of accounting information prepared using CASand IAS.

We, therefore, incorporated some modifications in our study design. We set out to test the incre-mental value relevance of the accounting information prepared under the two sets of accountingstandards for A- and B-shares, and we chose 1995–2000 as our test period, longer and morerecent than that used in the prior studies. Further, we applied both a price-level model and a returnmodel in order to mitigate the scale effect problem associated with the price-levels model (Easton,Harris, & Ohlson, 1992; Barth & Kallapur, 1996). We tested the effects of the difference caused byCAS and IAS on stock prices and stock returns in both the A- and B-share markets, an extensionof the only B-share or A-share market approach applied by similar studies in the literature. Inaddition, we wanted to examine whether there is a different degree of information value relevance


in the SHSE and SZSE since the composition of B-share investors is significantly varied in thetwo stock exchanges (Li, 1999; Fung et al., 2000).12

3.1. Value relevance to equity valuation and stock returns

We employed the price-per-share regression model to test the value relevance of the accountingnumbers to equity valuation in the Chinese stock market. This model was developed byOhlson(1995) to determine what is reflected in a firm’s value (Barth et al., 2001). It is assumed thatstock prices reflect not only the information content of earnings and book values but also otherinformation being released to the market (Ohlson & Shroff, 1992; Ohlson, 1995). Thus:

Pjt = β0 + β1BVPSCASjt + β2EPSCAS

jt + β3Ujt + ηjt (1)

wherePjt is stock price of firmj at 4 months after the fiscal year end of periodt; BVPSCASjt is book

value per share of owners’ equity determined under CAS for firmj at timet; EPSCASjt is earnings

per share determined under CAS for firmj at timet; Ujt is the nonrandom “other information”apart from book value and earnings per share determined under CAS.

If CAS earnings and book values are significant to equity valuation,Ujt could be omitted. Butif the reconciliation of accounting numbers from CAS to IAS provides additional information, the“other information” should include the IAS reconciliation amounts of earnings and book values.Our testing model is restated as:

Pjt = β0 + β1BVPSCASJt + β2EPSCAS

jt + β3BVPSIAS−CASjt + β4EPSIAS−CAS

jt + ηjt (2)

where BVPSIAS−CASjt is difference in book value per share reported under IAS and CAS for firm

j at timet; EPSIAS−CASjt is difference in earnings per share reported under IAS and CAS for firm

j at timet.Under the existing regulations, Chinese listed companies must publish their annual reports

within 4 months after the end of each calendar year. Therefore, the stock prices were measuredon the date 4 months after the fiscal year end. All variables were measured in Chinese currency(RMB ¥), so the foreign currency amounts for B-shares are converted into RMB ¥ based on theexchange rates on the reporting date in each year. We expected the coefficients on EPSCAS

jt and/or

BVPSCASjt to be positive, which would indicate that CAS-based earnings and/or book values have

relevant information content. The coefficients on BVPSIAS−CASjt and EPSIAS−CAS

jt provide a testof whether the IAS reconciliation contains incremental value to the market after controlling forthe reported numbers based on CAS.

In order to overcome the scale effects in price level regression, we also adopted returns modelto test the value relevance of accounting numbers.13 The return model is deflated by lagged stock

12 Although only overseas investors were permitted to participate in the B-share market before early 2001, the B-sharemarket participants are mainly overseas institutional investors in the SHSE that is located in the long-history financialcenter of China, while about 90% of B-share participants in SZSE, located in a newly emerged Special Economic Zone,are in fact the individual residents living in the neighboring Hong Kong and Macao or even the domestic citizens whocould trade B-shares with the accounts set up by holding travel permit to, or through their close relatives, in Hong Kong orMacao. Apparently the overseas institutional investors and the individuals with Chinese background or connection mayreact differently to accounting information prepared under CAS or IAS.13 Some researchers argue that the price model has certain drawbacks, mainly arising from the scale effect in levels

regression, which increases R2 and the coefficient of variance (Kothari & Zimmerman, 1995). Other researchers opt for


price as suggested byBiddle et al. (1995)to mitigate the potential heteroscedasticity problem(Harris & Muller, 1999; Kothari, 2001). Since both earnings and earnings change are asso-ciated with stock returns (Easton et al., 1992; Easton, 1999), the variables EPSIAS−CAS

jt and

�EPSIAS−CASjt in Eq.(3) are the focus of our regression tests. We expectφ3 andφ4 to be nonzero

if the earnings reconciliation from CAS to IAS contains incrementally value relevant information;

Rjt = φ0 + φ1EPSCAS/Pjt−1jt + φ2�EPSIAS−CAS/Pjt−1

jt + φ3EPSIAS−CAS/Pjt−1jt

+ φ4�EPSIAS−CAS/Pjt−1jt + πjt (3)

whereRjt is annual stock return for firmj at timet; EPSCAS/Pjt−1jt is earnings per share determined

by CAS and deflated by beginning stock price for firmj at timet; EPSIAS−CAS/Pjt−1jt is difference

in earnings per share determined under IAS and CAS, deflated by beginning stock price for firmj at time t; �EPSCAS/Pjt−1

jt is change in CAS-based earnings per share deflated by beginning

stock price for firmj at timet; �EPSIAS−CASjt is change in the difference in earnings per share

determined under IAS and CAS, deflated by beginning stock price for firmj at timet; πjt is theerror term.

Eqs.(2) and (3)were applied not only to test the value relevance of CAS-based accountingnumbers to equity valuation and stock returns in the Chinese stock market, but also to test whetherthe IAS reconciliation is incrementally value-added to the B-share and A-share markets, respec-tively. We intended to examine whether the two sets of accounting information, released to theinvestors simultaneously, elicited different reactions in the two markets.

3.2. Sample selection and data collection

In order to test the information content in A-share and B-share markets, listed firms that issuedboth A- and B-shares on the SHSE and the SZSE during 1995–2000 were selected. A few firmsissuing only B-shares or with incomplete data were excluded. Thus, each firm is matched withitself in the sample, so each firm-year observation is used as its own control to avoid the potentialmatching procedure problem (Chan & Seow, 1996). Since the numbers of B-share companieslisted on the SHSE and the SZSE changed (e.g., new listing or delisting) continuously, the samplesize varies for each year of observation. Each firm-year observation satisfied four data-availabilitycriteria: (1) the CAS-based earnings, book values of owners’ equity, and IAS-based earnings andbook value numbers were available from the annual reports; (2) a firm’s A- and B-shares wereboth been listed for more than one year to permit the calculation of stock returns, (3) the stockprices were available in the database of the Taiwan Economic Journal (TEJ), which is widelyused for studies on Chinese stock market performance; and (4) there was a nonzero reporteddifference between the earnings and book value determined under CAS and IAS for the samefirm. The test period begins at 1995 since a large portion of the Chinese listed companies didnot provide appropriate financial disclosures to the public before that year. The sample ends withthe data for the year 2000 for two reasons: (1) it is the last year of strict market segmentationbefore the Chinese government liberalized the B-share market to permit domestic participants

the use of returns model (Barth & Kallapur,1996). However,Amir et al. (1993)argued that the response coefficients maybe biased in the returns model as the reconciliation items can sometimes be anticipated (p. 240). Therefore,Kothari andZimmerman (1995)contended that a combined use of stock returns and price levels models may yield more convincingevidence (Beaver & Ryan, 2000).


Table 2Sample distribution for value-relevance analysis

Year Firm-year observations

All B-share firms Entire sample SHSE subsample SZSE subsample

Number T.A.a

(RMB¥million)

Number %b T.A.a

(RMB¥million)

Number %c T.A.a

(RMB¥million)

Number %c T.A.a

(RMB¥million)

1995 70 1656.2 53 75.7 1868.3 32 88.9 1894.8 21 61.8 1829.51996 85 1869.5 58 68.2 2156.9 32 76.2 2093.2 26 60.5 2235.31997 101 2064.5 71 70.3 2272.6 36 72.0 2194.9 35 68.6 2352.51998 106 2200.7 79 74.5 2359.4 39 75.0 2280.3 40 74.1 2436.41999 108 2456.8 78 72.2 2563.2 38 70.4 2440.3 40 74.1 2680.02000 113 2827.0 76 67.3 2950.4 38 69.1 2957.8 38 65.5 2943.0

Total 583 415 71.2 215 200

a This is the mean value of total assets (T.A.) for the B-share companies, expressed in Chinese currency unit of RMB¥million.

b The percentages indicate the proportion of entire sample to all B-share companies in each year under observation.c The percentages indicate the proportion of sample companies to all B-share companies listed in the SHSE and SZSE,

respectively.

in early 2001 and (2) we wanted to test whether the introduction of theSecurities Law, theamended accounting regulations, or the practical accounting standards (PASs) introduced forlisted companies in 1998, caused any difference in the value relevance of accounting numbers. Atotal of 415 firm-year observations were obtained. The sample distribution is described inTable 2.The sample includes more than 70% of all B-share companies during the testing period, coveringcompanies from manufacturing, commercial, telecommunication, banking, transportation, publicutilities, and other industrial sectors. Based on the average total assets as indicated inTable 2, thesample companies are slightly larger in size than all B-share companies combined.

As a data accuracy check, we randomly selected the data for a few firms from the TEJ databaseand compared them to the corresponding figures in the financial statements of the same companiespublished in the designated national newspapers. No substantial discrepancy was found in the datacollected from these separate sources.

4. Empirical results

Table 3reports the descriptive statistics for the variables included in the regression models. Asindicated in Panel (A), using the price-levels model, the mean stock price for the entire sampleof B-shares is ¥3.74314 while the mean stock price for the entire sample of A-shares is ¥11.846,a difference of –¥8.103, which indicates that the price of A-shares, on average, is more than 3times that of B-shares (at the significance level of 0.01 based on two-tailedt-test). The book valueper share and earnings per share reported under CAS are ¥2.549 and ¥0.150, respectively for theentire sample of B-shares across time. The mean differences of book value per share and earningsper share determined under the two sets of accounting standards (IAS versus CAS) are – ¥0.116

14 All variables are measured in Chinese currency (RMB¥), so the original foreign currency amounts reported for B-shareshave been converted into RMB¥ based on official exchange rates on the reporting date for each year under observation.


Table 3Descriptive statistics on variables used in regression models

Variable Mean S.D. Minimun Maximum N

Panel (A): descriptive statistics on price levels model:Pjt = β0 + β1BVPSCASjt + β2EPSCAS

jt + β3BVPSIAS−CASjt +

β4EPSIAS−CASjt + ηjt

Pjt—B-share 3.743 3.248 0.431 18.470 415Pjt—A-share 11.846 5.460 3.800 35.980DPjt −8.103 3.858 −24.725 −1.223BVPSCAS

jt 2.549 1.077 −2.013 6.350EPSCAS

jt 0.150 0.351 −2.295 1.358BVPSIAS–CAS

jt −0.116a 0.248 −1.474 0.993EPSIAS–CAS

jt −0.041b 0.104 −0.483 0.910

Panel (B): descriptive statistics on stock returns model: B-shares model:Rjt = φ0 + φ1EPSCAS/Pjt−1jt +

φ2�EPSCAS/pjt−1jt + φ3EPSIAS−CAS/pjt−1

jt + φ4�EPSIAS−CAS/pjt−1jt + πjt

Rjt 0.252 1.644 −0.784 9.460 415EPSCAS/Pjt−1

jt 0.010 0.250 −2.339 0.888�EPSCAS/Pjt−1

jt −0.009 0.311 −1.171 3.788EPSIAS–CAS/Pjt−1

jt −0.026c 0.088 −0.754 0.432�EPSIAS–CAS/Pjt−1

jt −0.002 0.125 −1.314 0.645

Panel (C): descriptive statistics on stock returns model: A-shares model:Rjt = φ0 + φ1EPSCAS/Pjt−1jt +

φ2�EPSCAS/pjt−1jt + φ3EPSIAS−CAS/Pjt−1

jt + φ4�EPSIAS−CAS/Pjt−1jt + πjt

Rjt 0.301 0.648 −0.465 3.697 415EPSCAS/Pjt−1

jt 0.012 0.043 −0.342 0.168

�EPSCAS/Pjt−1jt −0.006 0.043 −0.189 0.381

EPSIAS–CAS/Pjt−1jt −0.005c 0.014 −0.113 0.099

�EPSIAS–CAS/Pjt−1jt −0.001 0.019 −0.177 0.122

Note: t-test was run to compare the mean differences of book value per share (BVPS) and earnings per share (EPS)prepared under CAS and IAS, respectively.

a For BVPS, the amount based on IAS is lower than that based on CAS at a significance level of 0.10.b While the difference for the two mean amounts of EPS is significant at 0.05 level.c Based ont-test result, the difference between the earnings amounts under IAS and CAS deflated by prices in prior

period is significantly different from zero at 0.0 l level.

(significance level of 0.10) and – ¥0.041 (significance level of 0.05), respectively. Apparentlyboth earnings and book values restated based on IAS are lower than the CAS amounts.

Panel (B) inTable 3reports the descriptive statistics of the variables included in the stockreturn model for B-shares. For the entire sample, the mean annual stock return is 25.2% for theperiod 1995–2000. The mean of earnings per share under CAS deflated by lagged stock pricesof B-shares is 0.010. The mean of earnings change reported under CAS deflated by lagged stockprices of B-shares is−0.009. The difference in earnings under IAS and CAS deflated by laggedstock prices of B-shares is−0.026 (significantly different from zero at 0.01 level). The differencein the earnings change under IAS and CAS deflated by lagged stock prices of B-shares is−0.002.The last two numbers indicate that, on average, IAS-based earnings are lower compared with theCAS amounts.

Panel (C) inTable 3reports the descriptive statistics of the variables included in the returnmodel for A-shares. For the entire sample, the mean annual stock return is 30.1% for the period1995–2000. The mean return for A-shares is about 20% greater than that for B-shares (25.2%).The mean of CAS earnings deflated by lagged stock prices of A-shares is 0.012. the CAS-based


Table 4Regression results of CAS and IAS measures on price-levels model: B-shares modelPjt = β0 + β1BVPSCAS

jt +β2EPSCAS


jt + ηjt

β0 β1 β2 β3 β4 R2adj

N

Entire2.251 (5.11*** ) 0.414 (2.25** ) 2.912 (5.16*** ) 0.173 4152.545 (5.79*** ) 0.444 (2.45** ) 2.571 (4.45*** ) 1.312 (1.96* ) 3.471 (2.44** ) 0.201 415

SHSE1.956 (2.88*** ) 0.353 (1.21) 3.944 (3.85*** ) 0.140 2152.319 (3.41*** ) 0.394 (1.37) 3.373 (3.18*** ) 1.008 (1.43) 4.317 (2.14** ) 0.166 215

SZSE2.609 (4.43*** ) 0.427 (1.78* ) 2.543 (3.65*** ) 0.196 2002.896 (4.88*** ) 0.469 (1.97** ) 2.085 (2.85*** ) 2.788 (2.14** ) 1.197 (0.56) 0.216 200

Pjt: stock price of firmj 4 months after the fiscal year end of periodt; BVPSCASjt : book value per share of owners’ equity

calculated under Chinese accounting standards for firmj at timet; EPSCASjt : earning per share calculated under Chinese

accounting standards for firmj at time t; BVPSIAS–CASjt : difference in book value per share under IAS and Chinese

accounting standards for firmj at timet; EPSIAS–CASjt : difference in earning per share under IAS and Chinese accounting

standards for firmj at timet; Values in brackets aret-statistics.** Statistical significance at 0.05 level.

*** Significance at 0.01 level.

earnings change deflated by lagged stock prices of A-shares is−0.006. The difference in theearnings determined under IAS and CAS deflated by lagged stock prices of A-shares is−0.005.The difference in earnings changes under IAS and CAS deflated by lagged stock prices of A-shareis −0.001. These statistics confirm that earnings per share reported for A-shares is, on average,higher than that for B-shares.

4.1. Value relevance to B-share market

Table 4reports the results of the estimated price-level model by pooling all observations ofB-shares across time. When only CAS amounts are included in the price-levels regression for theentire sample of B-shares, the adjustedR2 is 17.3%. The estimated coefficients on book valueand earnings are both positive at the significance levels of 5% and 1%, respectively, indicatingthat the CAS-based earnings and book value per share are value-relevant to stock prices in theB-share market. When the amounts of reconciliation between CAS and IAS are included, theadjustedR2 increases to 20.1%. Furthermore, the estimated coefficients on IAS adjusted bookvalue and earnings per share are both significantly different from zero, which suggests that the IASreconciliation adjustment provides incremental information for equity valuation in the B-sharesmarket.

Table 5reports the results of the estimated stock returns model for B-shares by pooling allobservations across time. The adjustedR2 of the returns model is 1.4% for all B-share compa-nies in the sample when only CAS-based variables are included in the regression. The estimatedcoefficient on CAS earnings deflated by lagged stock prices does not significantly differ fromzero, but the coefficient on earnings change is positive at the significance level of 5%, indicat-ing that CAS-based earnings change is value-relevant. This is consistent with the contentionthat earnings change is an explanatory variable frequently used as a measure of the unexpected


Table 5Regression results of CAS and IAS measures on stock returns model: B-shares modelRjt = φ0 + φ1EPSCAS/Pjt−1

jt +φ2�EPSCAS/Pjt−1

jt + φ3EPSIAS−CAS/Pjt−1jt + φ4�EPSIAS−CAS/Pjt−1

jt + πjt

φ0 φ1 φ2 φ3 φ4 R2adj

N

Entire0.807 (9.99*** ) 0.250 (0.65) 0.604 (1.97** ) 0.014 4150.796 (9.23*** ) 0.268 (0.69) 0.864 (2.60*** ) −0.661 (−0.54) 1.743 (1.92* ) 0.021 415

SHSE0.954 (7.58*** ) −1.584 (−2.16** ) 3.492 (3.21*** ) 0.038 2150.892 (6.38*** ) −1.389 (−1.90) 4.222 (3.77*** ) −2.021 (−1.14) 3.985 (2.64** ) 0.065 215

SZSE0.776 (7.24*** ) 0.894 (1.83* ) 0.264 (0.87) 0.029 2000.773 (7.07*** ) 0.893 (1.80* ) 0.300 (0.87) −0.341 (−0.19) 0.275 (0.24) 0.019 200

Rjt: annual stock return for firmj for period ending 4 months after the fiscal year end of periodt; Pjt-1: stock price forfirm j 4 months after the fiscal year end of periodt − 1; EPSCAS

jt : earning per share for B-shares calculated under Chinese

accounting standards for firmj at timet; �EPSCASjt : change of earning per share for B-shares under Chinese accounting

standards for firmj from periodt − 1 to periodt; EPSIAS–CASjt : difference in earning per share for B-shares under IAS and

Chinese standards for firmj at timet; �EPSIAS–CASjt : change of earning per share difference for B-share under IAS and

Chinese standards for firmj from periodt − 1 to t; Values in brackets aret-statistics.** Statistical significance at 0.05 level.

*** Significance at 0.01 level.

earnings (Easton et al., 1992; Easton, 1999). When IAS reconciliation amounts are included, theadjustedR2 increased to 2.1%, in other words, additional explanatory power is gained. The esti-mated coefficient on CAS-based earnings change increases from 0.604 to 0.864 at the significancelevel of 1% (t = 2.60,p < 0.01). However, the estimated coefficient of IAS-based earnings adjust-ment,EPSIAS−CAS

jt , is negative with no statistical significance (coefficient =−0.661,t =−0.54),suggesting that the IAS earnings reconciliation amount is not incrementally value-added forstock returns after controlling for the CAS amount. But the estimated coefficient of earningschange adjustment is positive at the significant level of 10% (φ4 = 1.743, t = 1.92, p < 0.10),suggesting that the IAS reconciliation of earnings change is marginally value-added to stockreturns in the B-share market. Compared with the price-levels model (R2 = 0.201), however,the adjustedR2 of the returns model declines dramatically. This finding is consistent with theresults of other value relevance studies which indicate that the adjustedR2 is usually lower inthe returns-based model than in the price-levels model (Harris & Muller, 1999; Beaver &Ryan,2000).

Based on the results of the two regression models, we conclude that the book value and earningsobtained under CAS are value-relevant to equity valuation in the B-share market. This result isconsistent with prior studies in industrialized countries which found that the national (domestic)GAAP measures are more closely associated with market performance than foreign GAAPs (Chan& Seow, 1996; Rees & Elgers, 1997; Harris & Muller, 1999; Niskanen et al., 2000). In other words,the domestic GAAP is a better measure of valuation as reflected in stock prices. Considering theIAS reconciliation, both book value and earnings adjustments are incrementally value-added tostock valuation but not to the determination of stock returns. The results demonstrate that, for theentire sample of B-share companies, the IAS reconciliation of book value and earnings is partiallyvalue- relevant.


Table 6Regression results of CAS and IAS measures on price-levels model: A-shares modelPjt = β0 + β1BVPSCAS

jt +β2EPSCAS


jt + ηjt


N

Entire10.266 (13.80*** ) 0.279 (0.90) 5.807 (6.09*** ) 0.166 41510.564 (14.08*** ) 0.319 (1.03) 5.336 (5.41*** ) 1.997 (1.86* ) 2.402 (0.99) 0.174 415

SHSE11.160 (9.34*** ) −0.072 (0.14) 6.921 (3.83*** ) 0.099 21511.597 (9.58*** ) 0.126 (0.25) 6.191 (3.28*** ) 1.313 (1.50) 5.034 (1.40) 0.107 215

SZSE9.346 (9.89*** ) 0.404 (1.04) 5.418 (4.80*** ) 0.230 2009.801 (10.13*** ) 0.457 (1.18) 4.849 (4.06*** ) 3.499 (1.65) 1.766 (0.50) 0.239 200

4.2. Value relevance to A-share market

Table 6reports the results of the estimated price-levels model, pooling all observations ofA-shares across time. When only CAS amounts are included in the price-levels model for theentire sample, the adjustedR2 is 16.6%. The estimated coefficient on book value is positivebut not at a statistically acceptable level, while that for earnings is positive at the significancelevel of 1%. These two coefficients indicate that earnings reported under CAS are reflected instock valuation, but book value per share is not value relevant in the A-share market. When IASreconciliation amounts are included, the adjustedR2 is only slightly changed (17.4%). Althoughthe estimated coefficient on IAS-adjusted book value,BVPSIAS−CAS

jt is positive, it is only at themarginal significance level of 10% (coefficient = 1.997,t = 1.86). The reconciliation of earningsdoes not have an effect at the acceptable significance level. These statistics suggest that the IASreconciliation is generally not value-added to the determination of stock prices in the A-sharemarket.

Table 7reports the results of the estimated stock returns model for A-shares, pooling allobservations across time. The regression results indicate that the adjustedR2 is 7.5% for the entiresample when only CAS-based variables are included. The estimated coefficient on earnings is

Table 7Regression results of CAS and IAS measures on stock returns model: A-shares modelRjt = φ0 + φ1EPSCAS/Pjt−1

jt +φ2�EPSCAS/Pjt−1


jt + πjt


N

Entire0.261 (7.85*** ) 3.651(4.50*** ) 0.996 (1.24) 0.075 4150.236 (6.52*** ) 3.709 (4.57*** ) 1.411 (1.59) −5.275 (−1.55) 3.117 (1.26) 0.076 415

SHSE0.248 (6.52*** ) 0.332 (0.23) 3.888 (1.92* ) 0.021 2150.207 (4.64** ) 0.822 (0.56) 4.270 (2.09** ) −5.911 (−1.66* ) 5.383 (1.78* ) 0.028 215

SZSE0.321 (5.53*** ) 4.147 (3.75*** ) 0.781 (0.75) 0.092 2000.301 (5.05*** ) 4.152 (3.75*** ) 1.068 (0.88) −6.703 (−1.35) 2.552 (0.67) 0.092 200


significant at 1% (coefficient = 3.651,t = 4.50), indicating that CAS earnings are value-relevant.But CAS-based earnings changes are not associated with stock returns. When IAS reconciliationamounts are included, the adjustedR2 is nearly the same (7.6%). The estimated coefficient on CASearnings remains significantly positive at the 1% level (coefficient = 3.709,t = 4.57), while the CASearnings change has no significant effect (coefficient = 1.441,t = 1.59) when IAS reconciliationamounts are included in the regression. None of the estimated coefficients on IAS reconciliationamounts of earnings (coefficient =−5.275,t =−1.55) and earnings change (coefficient = 3.117,t = 1.26) is significantly different from zero at the acceptable level.

The test results for these two models indicate that the earnings reported under CAS are impor-tant information for determining stock prices and returns in the A-share market, but the CAS-basedbook value is not value-relevant in the A-share market. This result differs from the finding reportedby Chen et al. (2001). In the IAS-based reconciliation, neither book value nor earnings is incre-mentally value-added either in the price-levels model or in the stock returns model. Thus, it canbe concluded that the reconciliation of earnings and shareholders’ equity numbers prepared underCAS to IAS-based amounts is not incrementally value-added for the A-share market.

Comparing the regression results for A- and B-share markets, stock returns in B-share marketare mainly determined by earnings change (�2 is significantly positive at the 1% level), while thestock returns on A-shares are mainly associated with current year earnings (�1 is significantlypositive at the 5% or 1% level). This difference may be due to the different types of investors andthe segmentation of the A- and B-share markets. In order words, domestic investors (A-shares)are more concerned about current year earnings, while overseas investors (B-shares) give greaterweight to earnings changes as an indication of future earnings.

4.3. Subsample comparison

In addition to value relevance testing at the aggregate level (e.g., for the entire sample), wehave also run regressions for two subsamples of companies that were listed in the SHSE andSZSE, respectively.Tables 4 and 5also report the regression results of the price-levels and stockreturns models in the B-share market for the two subsamples, pooling all observations acrosstime. Mixed results were obtained. InTable 4, it can be seen that only CAS earnings are reflectedin stock prices in the SHSE, while both CAS book value and earnings are value- relevant to deter-mining stock prices in the SZSE. When IAS reconciliation amounts are included, the earningsadjustment amount has an incremental effect on B-share prices in the SHSE (coefficient = 4.317,t = 2.14,p < 0.05), but it is the book value reconciliation amount that has an effect on B-sharesin the SZSE (coefficient = 2.788,t = 2.14,p < 0.05). In the returns model, as indicated inTable 5,the CAS-based earnings change is value-relevant mainly in the SZSE while CAS earnings havea marginal effect in both the SHSE and SZSE. When IAS reconciliation amounts are includedin the regression, the adjustment of earnings changes is value-relevant only in the SHSE (coeffi-cient = 3.985,t = 2.64,p < 0.05) but not value-added in the SZSE. The IAS adjustment amount ofearnings is not incrementally value-added to stock returns in both the SHSE and SZSE since thecoefficients do not significantly differ from zero.

Table 6also demonstrates that CAS-based earnings are an important factor to determine A-share prices in both the SHSE and SZSE, as the coefficients of�2 are both significantly positive atthe 1% level. The CAS-based book value is not value-relevant for A-share price levels. The IASreconciliation is not value-added to A-share prices in either of the two stock exchanges since noneof the coefficients on the adjustment amounts differs significantly from zero. The stock returnsregression results indicated inTable 7show that the CAS-based earnings deflated by lagged

Z.Jun

Lin,F.C

hen/JournalofInternationalA

ccounting,Auditing

andTaxation

14(2005)

79–10395

Table 8Descriptive statistics on variables used in yearly regression modelsa

Variable Year

1995 1996 1997 1998 1999 2000

Panel (A): descriptive statistics on price levels modela: Pjt = β0 + β1BVPSCASjt + β2EPSCAS


jt + ηjt

Pjt–B-share 2.757 (1.475) 4.942 (3.171) 2.524 (2.019) 1.390 (0.993) 2.207 (1.120) 8.676 (2.573)Pjt–A-share 8.240 (3.809) 13.670 (6.663) 11.329 (5.188) 8.432 (2.790) 13.316 (4.487) 8.677 (2.572)DPjt −5.483 (3.134) −8.728 (3.931) −8.806 (3.498) −7.042 (2.212) −11.042 (3.885) −6.813 (3.703)BVPSCAS

jt 2.659 (0.713) 2.663 (0.761) 2.637 (0.850) 2.394 (1.104) 2.395 (1.251) 2.622 (1.406)EPSCAS

jt 0.284 (0.247) 0.206 (0.278) 0.162 (0.363) 0.022 (0.490) 0.125 (0.317) 0.150 (0.263)BVPSIAS–CAS

jt −0.085 (0.326) −0.141 (0.283) −0.184 (0.290) −0.138 (0.219) −0.100 (0.152) −0.045 (0.206)EPSIAS–CAS

jt −0.060 (0.074) −0.056 (0.097) −0.052 (0.099) −0.023 (0.146) −0.052 (0.113) −0.016 (0.056)

Panel (B): descriptive statistics on stock returns model: B-shares model:Rjt = φ0 + φ1EPSCAS/Pjt−1jt + φ2�EPSCAS/Pjt−1

jt + φ3EPSIAS−CAS/Pjt−1jt +

φ4�EPSIAS−CAS/Pjt−1jt + πjt

Rjt 0.071 (0.277) 1.070 (1.098) −0.487 (0.169) −0.396 (0.150) 0.862 (0.604) 3.512 (1.592)EPSCAS/Pjt−1

jt 0.086 (0.057) 0.058 (0.108) 0.186 (0.074) −0.073 (0.352) 0.020 (0.407) 0.051 (0.122)�EPSCAS/Pjt−1

jt −0.043 (0.054) −0.031 (0.095) −0.017 (0.079) −0.103 (0.270) −0.003 (0.133) −0.010 (0.311)EPSIAS–CAS/Pjt−1

jt −0.030 (0.046) −0.028 (0.056) −0.017 (0.034) −0.019 (0.110) −0.056 (0.148) −0.006 (0.035)�EPSIAS–CAS/Pjt−1

jt −0.008 (0.040) 0.001 (0.052) 0.001 (0.030) 0.016 (0.132) −0.044 (0.234) −.0002 (0.125)

Panel (C): descriptive statistics on stock returns model: A-shares model:Rjt = φ0 + φ1EPSCAS/Pjt−1jt + φ2�EPSCAS/Pjt−1

jt + φ3EPSIAS−CAS/Pjt−1jt +

φ4�EPSIAS−CAS/Pjt−1jt + πjt

Rjt 0.134 (0.280) 1.092 (0.977) −0.076 (0.253) −0.158 (0.206) 0.702 (0.553) 0.229 (0.301)EPSCAS/Pjt−1

jt 0.039 (0.037) 0.024 (0.050) 0.007 (0.027) −0.007 (0.059) 0.012 (0.040) 0.011 (0.020)�EPSCAS/Pjt−1

jt −0.017 (0.026) −0.013 (0.038) −0.006 (0.029) −0.018 (0.045) 0.0138 (0.069) −0.001 (0.018)EPSIAS–CAS/Pjt−1

jt −0.009 (0.012) −0.008 (0.016) −0.005 (0.010) −0.002 (0.017) −0.007 (0.018) −0.001 (0.005)�EPSIAS–CAS/Pjt−1

jt −0.001 (0.011) −0.000 (0.016) −0.000 (0.010) 0.003 (0.022) −0.006 (0.029) 0.003 (0. 010)

a The numbers represent the means of related variables in each year while the numbers in brackets are the corresponding standard deviations.

96Z

.JunL

in,F.Chen

/JournalofInternationalAccounting,A

uditingand

Taxation14

(2005)79–103

Table 9Results of yearly regression—price levels model


N

Panel (A): B-shares model:Pjt = β0 + β1BVPSCASjt + β2EPSCAS


jt + ηjt

1995 0.368 (0.87) 0.379 (2.40** ) 4.692 (9.89*** ) 0.155 (0.47) −1.041 (−0.73) 0.754 531996 1.984 (1.79* ) 0.604 (1.29) 7.471 (5.44*** ) −1.537 (−1.37) 7.368 (2.32** ) 0.575 581997 −0.009 (−0.01) 0.917 (3.41*** ) 2.135 (3.13*** ) 0.652 (0.91) 2.413 (1.13) 0.551 711998 0.175 (0.58) 0.568 (4.56*** ) 0.197 (0.67) 1.003 (2.77*** ) 0.472 (0.83) 0.548 791999 1.002 (4.18*** ) 0.425 (3.94*** ) 1.444 (3.06*** ) 0.335 (0.49) −0.786 (−0.95) 0.565 782000 7.948 (13.81*** ) −0.150 (−0.61) 6.945 (5.20*** ) −0.219 (−0.18) −0.368 (−0.09) 0.486 76

Total 415

Panel (B): A-shares1995 4.586 (2.37** ) 0.597 (0.80) 5.750 (2.65** ) 2.735 (1.81* ) −11.897 (−1.83* ) 0.230 531996 7.592 (3.08*** ) 1.235 (1.19) 14.978 (4.88*** ) 0.797 (0.32) 3.409 (0.48) 0.520 581997 6.416 (3.62*** ) 1.580 (2.16** ) 7.106 (3.85*** ) 0.128 (0.07) 7.328 (1.42) 0.500 711998 6.212 (5.90*** ) 0.922 (2.12** ) 1.339 (1.31) 0.210 (0.17) −0.474 (−0.24) 0.302 791999 10.923 (8.41*** ) 0.862 (1.47) 2.215 (0.87) 0.214 (0.06) −1.430 (−0.32) 0.087 782000 15.898 (11.00*** ) −0.734 (−1.18) 8.780 (2.62** ) −2.003 (−0.64) −1.681 (−0.16) 0.047 76

Total 415

The values in brackets aret-statistics;* statistical significance at 0.10 level;** significance at 0.05 level;*** significance at 0.01 level.


price levels are value-relevant information in the SZSE (coefficient = 4.147,t = 3.75,p < 0.01) butnot value-relevant in the SHSE. However, the IAS adjustment amounts of earnings and earningschange are marginally relevant in the SHSE at the 10% significance level, though they are notincrementally value-added to stock returns for A-shares in the SZSE.

Overall, the subsample comparison reveals that CAS-based earnings are relevant to investorsin both the SHSE and the SZSE, but the IAS reconciliation is only marginally value-added toinvestors in the SHSE. This result could be explained by the fact that the market participants inthe SHSE, where there is a greater proportion of foreign institutional investors than in the SZSE,may have a better understanding of the differences between CAS and IAS, and responded moresensitively to IAS reconciliation adjustments. In other words, investors in the SHSE would havegiven a relatively greater weight to IAS measures than their counterparts in the SZSE.

4.4. Yearly comparison

We further tested the two regression models for yearly comparison. Descriptive statistics onyearly variables are presented inTable 8. As indicated by the results of the price-levels model inPanel (A) ofTable 9, in the B-share market, the coefficients of CAS-based earnings are positiveat the significance level of 1% for all years under observation except for 1998. The CAS-basedbook value is significantly positive at the 1% level in 1995 and 1997–1999. When IAS recon-ciliation amounts are included, the book value adjustment amount is only value-added in 1998(coefficient 1.003,t = 2.77,p < 0.01) while the earnings adjustment amount is value-added only in1996 (coefficient = 7.368t = 2.32,p < 0.05). The yearly regression results confirm that CAS-basedearnings and book values have been reflected in B-share prices while the IAS reconciliation hasonly a marginal influence. In addition, theR2 of the price-level model in the yearly regressionis relatively stable with a slight decline over time. Thus, no convincing evidence was obtainedto indicate whether the value relevance of CAS-based accounting numbers to the price levels ofB-shares has increased or decreased in the period 1995–2000. This result differs from the findingsof Bao and Chow (1999), who reported an increasing trend of the explanatory power of earningsand book value during the testing period of 1993–1996. However, as indicated in Panel (B) ofTable 9, CAS-based earnings and book value have an effect on A-share prices to varied extentsin different years and the IAS reconciliation is not incrementally value added in any year except1995. TheR2 of yearly price-level models for A-shares fluctuated quite substantially during thetest period, suggesting that there was a greater volatility in A-share prices in contrast to B-shares.

In the yearly stock returns models, CAS-based earnings and earnings change have varyingdegrees of explanatory power in each year during the testing period. As indicated in Panel (A)of Table 10, in the B-share market, CAS-based earnings change is not value-relevant in mostyears except for 1999–2000, while CAS earnings are value-relevant mainly in 1995 and 1997.Nonetheless the IAS reconciliation of earnings and earnings change is generally not incrementallyvalue-added in most of the years under observation. Furthermore, theR2 in the yearly returnsmodels demonstrates a declining trend. Similar results are obtained for the A-share model inPanel (B) ofTable 10. This finding is consistent with some studies in the US which found that thevalue relevance of earnings to stock returns deteriorates over time (Lev & Zarowin, 1999).

5. Analysis and discussion

There are two types of accounting and reporting requirements for companies issuing A- and B-shares in China. Investors may intuitively rely more on accounting information prepared under IAS

98Z

.JunL

in,F.Chen

/JournalofInternationalAccounting,A

uditingand

Taxation14

(2005)79–103

Table 10Results of yearly regressions—stock returns model


N

Panel (A): B-shares model:Rjt = φ0 + φ1EPSCAS/Pjt−1jt + φ2�EPSCAS/Pjt−1


jt + πjt

1995 −0.147 (−1.44) 2.457 (3.38*** ) 1.279 (1.77* ) −1.919 (−1.70* ) 0.367 (0.31) 0.296 531996 1.046 (2.87*** ) 2.462 (0.82) 0.866 (0.26) 3.255 (0.85) −0.408 (−0.10) 0.074 581997 −0.525 (−20.78*** ) 1.253 (3.12*** ) −0.151 (−0.40) −0.649 (−1.12) 1.354 (2.10** ) 0.234 711998 −0.396 (−21.90*** 0.087 (1.16) 0.107 (1.00) −0.634 (−2.41** ) 0.403 (1.65) 0.095 791999 0.805 (11.04*** ) −0.256 (−1.34) 0.402 (2.99*** ) −0.334 (−0.58) 0.115 (0.30) 0.077 782000 3.510 (15.07*** ) −2.331 (−1.21) 4.818 (2.96*** ) −3.130 (−0.51) 2.266 (1.14) 0.056 76

Total 415

Panel (B): A-shares1995 −0.082 (−1.26) 4.559 (4.72*** ) 0.727 (0.50) −6.690 (−1.67) 6.775 (1.82* ) 0.305 531996 0.269 (1.31) 21.342 (6.21*** ) −17.269 (−4.17*** ) −11.506 (−1.05) 12.322 (1.20) 0.407 581997 −0.084 (−1.96* ) 1.940 (1.12) 1.313 (0.81) −0.525 (−0.14) 4.783 (1.27) 0.076 711998 −0.148 (−5.60*** ) −0.618 (−1.05) 1.939 (2.09** ) −4.100 (−1.64) 3.305 (1.40) 0.029 791999 0.660 (8.88*** ) 0.139 (0.08) 2.340 (2.05** ) −3.718 (−0.76) 3.088 (0.98) 0.011 782000 0.258 (5.55*** ) −2.914 (−1.44) 7.070 (3.21*** ) −0.394 (−0.05) 1.397 (0.97) 0.009 76

Total 415

The values in brackets aret-statistics;* statistical significance at 0.10 level;** significance at 0.05 level;*** significance at 0.01 level.


since CAS-based accounting information has been perceived generally as less easy to understandand not comparable (Abdel-Khalik, Wong, & Wu, 1999; Chen, Lee, & Rui, 2001; Tang, 2000).The results of our study do not strongly support this perception. In fact, we observed that theCAS-based earnings were the most important variable to explain stock prices and stock returnsin the Chinese stock market. The finding, although counter-intuitive, is in line with the resultsof several studies in developed countries (Chan & Seow, 1996; Harris & Muller, 1999), whichconfirm that earnings based on national (domestic) GAAPs are more value-relevant. CAS-basedbook value is value-relevant to stock valuation in the B-share market although it is not value-addedin the A-share market. With respect to the value relevance of IAS reconciliation adjustment, weobtained mixed results. The IAS reconciliation of book value and earnings is value-added inthe price-levels model for the entire sample of B-shares, and the adjustment of earnings changefrom CAS to IAS is marginally value-added in the price deflated returns model. However, theexplanatory power of earnings to stock returns has generally declined over years, while the valuerelevance of book value was relatively stable during the testing period in the B-share market.Various explanations are offered below.

First, the greater attention paid by investors in both A- and B-share markets to the CAS-basedaccounting numbers may be due to their greater concerns about the specific economic, legal andsocial factors in the domestic market. In particular, investors are more familiar with the domesticmarket’s regulatory and information disclosure systems, thus the CAS-based accounting numbersare more value-relevant to the market. This is especially so in China, a country where accountingand financial reporting practices are taxation-oriented. Currently, CAS-based accounting numbers(particularly earnings) are in conformity with the taxable income that must be reported to, andapproved by, government tax authorities. Since dividend payout is generally based on after-taxincome, CAS-based earnings and/or book value are more naturally reflected in the price levelsand stock returns in the Chinese stock market.

Second, the stock market in China is currently not a mature market. The “pricing mechanism” isoften distorted since the government’s policy has a fairly strong influence on the stock market. TheChinese government has frequently intervened in market activities by taking administrative mea-sures to “calm down” or “bring in order” to the stock market (Anderson, 2000). The government’sad hoc interventions not only increase the volatility of stock prices, frustrating the government’sinitial intention, but also influence stock market performance (Su & Fleisher, 1998; Mookerjee &Yu, 1999; Chen et al., 2002).15 Consistent with the contention that value relevance may be affectedby market inefficiency (Aboody, Hughes, & Liu, 2002), the association of accounting numberswith stock prices and market returns is inevitably distorted in the newly established Chinese stockmarket. In other words, the government’s use of a “visible hand” instead of an “invisible hand” tocontrol the developing stock market may impair the usefulness of accounting numbers to investorsand other market participants in China. In fact, the value relevance of accounting information,IAS-reconciliation in particular, to stock prices and market returns is discounted in the immaturemarket dominated by frequent government interventions.

Third, the “thin-trade” nature of B-shares may be another factor undermining the usefulness ofIAS reconciliation information. From 1993 to 2000, the trading volumes of B-shares in the SHSE

15 For example, the stock market in China in late 1996 and early 2001 was buoyant, with a significant amount of insidertrading, price manipulation and speculation. Then the government warned through public media of taking action to “calmdown” the market. Shortly after, the government introduced some administrative measures to control the trading volatility,such as setting a permitted range of daily price fluctuation and raising the transaction levy and stamp duty. This led tosubstantial declines in stock prices and trading volumes.


were about ¥1.45, ¥7.87, ¥10.84, ¥6.09, ¥9.46, ¥21.29, ¥8.19 and ¥34.42 billions respectively.The trading volumes were ¥1.66, ¥2.58, ¥1.62, ¥1.70, ¥18.53, ¥21.37, ¥4.46 and ¥20.38 billionsin the SZSE, respectively. These trading volumes are fairly small compared with the tradingvolumes of A-shares. For instance, the ratios of annual trading volumes of B-shares to A-sharesare 6.23, 3.42, 1.92, 2.00, 1.05, 1.57, 0.66, and 1.11% over 1993–2000, respectively, in the SHSE,and 3.97, 2.05, 0.68, 1.85, 1.54, 1.27%, 0.40, and 0.69%, respectively, in the SZSE over the sameperiod. In addition, the daily trading volume was only, on average, HK$78.2 millions in 2000for B-shares in the SZSE.Aboody et al. (2002)contended that ‘thin trade’ may lead to pricingineffectiveness in an equity market. In the B-share market, given such thin trading volumes, theeffectiveness (usefulness) of accounting information, even the IAS reconciliation adjustment,could have been impaired.

Fourth, a unique feature of the securities market in China is the segmentation of owners’equity. There may be as many as four different categories of shares in a Chinese listed company,e.g., state owned shares, legal-entity owned shares, shares held by domestic investors (A-shares)and shares held by overseas investors (B-shares). Under the existing governmental regulations,the first two categories of shares are not allowed to be traded on the stock market, althoughthey account for, on average, the largest percentage (about 60–70%) of the total shares of listedcompanies. Non-circulation of such a large portion of equity shares may have an adverse impacton market performance. One obvious problem is the substantial price gap between A-shares andB-shares since the large numbers of domestic investors have no investment outlet other thanA-shares (Chen, Lee et al., 2001). The huge demand for limited numbers of A-share stocks andrestricted participation in B-share trading may have led to a distorted equity valuation in the market(Chen, Chen, & Su, 2001). As a result, the restriction of share circulation and segmented marketparticipation may have prevented an effective use of accounting information, IAS reconciliationadjustments in particular, in the Chinese stock market.

Finally, one possible explanation for the trend of decliningR2 over time in the yearly returnsmodels and for the non-incremental information content of IAS reconciliation of earnings may bethe gradual standardization of the stock market and the improvement in accounting and financialreporting practices in China. In particular, substantial changes are already under way in the settingand implementation of new rules for corporate reporting and disclosure. The introduction of theSecurities Law and new accounting standards, such as the amended accounting regulation forlisted companies and the PASs, in force since 1998, may have dramatically narrowed the gapbetween Chinese accounting and the internationally accepted norms (IAS in particular). Thegradual decrease of the differences between CAS and IAS may have led to an ongoing decline inthe information content of IAS reconciliation.

6. Conclusion

Through empirical testing of the value relevance of accounting information to the securityvaluation and stock returns in the Chinese stock market, we find that earnings prepared underCAS were the most important variable to explain stock prices and returns in both B- and A-sharemarkets. Book value of owners’ equity reported under CAS is partially value-relevant. However,the reconciliation of earnings and shareholders’ equity from CAS to IAS, at the aggregate level,are value-added only for stock prices in the B-share market. The IAS reconciliation of earningshas no significant information content for stock returns in either the A- or the B-share market.

In a newly developed capital market with stringent government controls and frequent policyintervention, the use of accounting information is secondary to market participants, since account-


ing numbers might have not been fully reflected in equity valuation or market returns. Although itis widely accepted that IAS is superior to CAS and should be increasingly adopted in the financialreporting of Chinese companies, the IAS reconciliation adjustment cannot generate incrementalinformation content in China’s existing immature market environment. This phenomenon may bedue to such factors as investors’ general familiarity with/or favor of the domestic taxation/marketregulation and reporting standards, thin-trades in the B-share market, segmentation of owners’equity, and the narrowing gap between CAS and IAS in recent years. Nonetheless, none of thesefactors should be regarded as an excuse for rejecting the harmonization of Chinese accountingstandards and practices to the internationally accepted accounting norms. As the stock marketmoves toward maturity, in pace with an increased integration of the Chinese economy with worldmarkets after China’s entry to the WTO, domestic and overseas investors will increasingly demandinternationally comparable and relevant accounting information. The harmonization of CAS withIAS should be beneficial in accelerating accounting development and economic growth in China.

The study attempted to empirically test the value relevance of accounting numbers in the newly-developed Chinese stock market. We obtained some evidence of market reactions to accountingnumbers (earnings and book values) prepared under the Chinese Accounting Standards and theInternational Accounting Standards and have shown that the IAS reconciliation of accountingnumbers is partially value-added, mainly for equity valuation in the B-share market. It is expectedthat our results will help the business community, inside and outside of China, gain an understand-ing of the recent development of the Chinese stock market and Chinese accounting practices. Itwill also provide some input to Chinese policy makers to establish, amend, and implement newaccounting standards.

This study tested a sample of Chinese companies issuing both A- and B-shares simultaneously.Subject to data availability, the sample size is relatively small. Thus, caution is necessary for theinference of the study results. Future study may be conducted to examine the differences amongsubsample companies divided by industry to enrich the study findings. In addition, a test of time-specific effects in the yearly regressions may also be performed in order to obtain further insightinto the value relevance of accounting numbers in the Chinese stock market.

Acknowledgements

The authors are grateful to two anonymous reviewers and Professor Kathleen Sinning, theEditor, for their insightful comments and editorial assistance to the prior versions of this paper.Comments received from the participants at Asian Pacific Conference on International AccountingIssues in Los Angeles, USA, 2002, and The Asian Academic Accounting Association Conferencein Seoul, South Korea, 2003, are highly appreciated.

References

Abdel-Khalik, A. R., Wong, K. A., & Wu, A. (1999). The information environment of China’s A and B shares: can wemake sense of the numbers?The International Journal of Accounting, 34(4), 467–489.

Aboody, Hughes, J., & Liu, J. (2002). Measuring value relevance in a (possible) inefficient market.Journal of AccountingResearch, 40(4), 965–986.

Aharony, Lee, J. C. J., & Wong., T. J. (2000). Financial packaging of IPO firms in China.Journal of Accounting Research,38, 103–126.

Alford, A., Jones, J., Leftwich, R., & Zmijewski, M. (1993). The relative informativeness of accounting disclosures indifferent countries.Journal of Accounting Research, 31(Supplement), 183–223.


Amir, E., Harris, T. S., & Venuti, E. K. (1993). A comparison of the value-relevance of US versus non-U.S. GAAPaccounting measures using Form 20-F reconciliations.Journal of Accounting Research, 31(Supplement), 230–264.

Anderson, D. M. (2000). Taking stock in China: company disclosure and information in China’s stock markets.GeorgetownLaw Journal, 88, 1919–1952.

Bandyopadhyay, S. P., Hanna, J. D., & Richardson, G. (1994). Capital market effects of U.S.-Canada GAAP differences.Journal of Accounting Research, 32, 262–277.

Bao, B. H., & Chow, L. (1999). The usefulness of earnings and book value for equity valuation in emerging capital markets:evidence from listed companies in the People’s Republic of China.Journal of International Financial Managementand Accounting, 10(2), 85–104.

Barth, M. E., & Kallapur, S. (1996). The effects of cross-sectional scale differences on regression results in empiricalaccounting research.Contemporary Accounting Research, 13(2), 527–567.

Barth, M. E., Beaver, W. H., & Landsman, W. R. (2001). The relevance of the value relevance literature for financialaccounting standard setting: another view.Journal of Accounting and Economics, 31, 77–104.

Barth, M. E., Clinch, G., & Shibano, T. (1999). International accounting harmonization and global equity markets.Journalof Accounting and Economics, 26, 201–235.

Beaver, W. H., & Ryan, S. G. (2000). Bias and lags in book value and their effects on the ability of the book-to-marketratio to predict book return on equity.Journal of Accounting Research, 38, 127–148.

Biddle, G. C., Seow, G. S., & Siegel, A. F. (1995). Relative versus incremental information content.ContemporaryAccounting Research, 12(1), 1–23.

Bodnar, G. M., & Weintrop, J. (1997). The valuation of the foreign income of US multinational firms: a growth opportunitiesperspective.Journal of Accounting Research, 24, 69–97.

Burke, M. E. (1999). China’s stock markets and the world trade organization.Law and Policy in International Business,30, 321–368.

Chan, M. W. L., & Rotenberg, W. (1999). Accounting, accounting education and economic reform in the People’s Republicof China.International Studies of Management & Organization, 29, 37–53.

Chan, K. C., & Seow, G. S. (1996). The association between stock returns and foreign GAAP earnings versus earningsadjusted to U.S. GAAP.Journal of Accounting and Economics, 21, 139–158.

Chen, G. M., Lee, B. S., & Rui, O. (2001). Foreign ownership restrictions and market segmentation in China’s stockmarkets.The Journal of Financial Research, 24, 133–155.

Chen, G. M., Lee, B. S., Rui, O., Firth, M., & Gao, N. (2002). The information content of concurrently announced earnings,cash dividends, and stock dividends: an investigation of the Chinese stock market.Journal of International FinancialManagement & Accounting, 13(2), 101–124.

Chen, J., & Thomas, S. C. (2003). The ups and downs of the PRC securities market.The China Business Review, 30(1),26–45.

Chen, J., Gul, F. A., & Su, X. (1999). A comparison of reported earnings under Chinese GAAP vs IAS: evidence fromthe Shanghai stock exchange.Accounting Horizons, 13(2), 91–111.

Chen, J., Chen, S., & Su, X. (2001). Is accounting information value-relevant in the emerging Chinese stock market?Journal of International Accounting Auditing and Taxation, 10(1), 1–22.

China Securities Regulatory Commission (CSRC), 1996. Rules for Information Disclosure by Companies Listing andIssuing B-share Stocks (in Chinese). Beijing.

China Securities Regulatory Commission (CSRC), 2000. Requirements on Information Disclosures by Banking and OtherFinancial Institutions Seeking Publicly Listing and Issuing of Stocks (in Chinese). Beijing.

Collins, D. W., Maydew, E. L., & Weiss, I. S. (1997). Changes in the value-relevance over the past forty years.Journalof Accounting and Economics, 24, 39–67.

Davidson, R. A., Gelardi, A. M. G., & Li, F. (1996). Analysis of the conceptual framework of China’s new accountingsystem.Accounting Horizons, 10(1), 58–74.

Easton, P. D., Harris, T. S., & Ohlson, J. A. (1992). Aggregate accounting earnings can explain most of security returns.Journal of Accounting and Economics, 15, 119–142.

Easton, P. D. (1999). Security returns and the value relevance of accounting data.Accounting Horizons, 13(4), 399–412.Ely, K., & Waymire, G. (1999). Accounting standards-setting organizations and earnings relevance: longitudinal evidence

from NYSE common stocks 1927–93.Journal of Accounting Research, 37, 293–317.Francis, J., & Schipper, K. (1999). Have financial statements lost their relevance?Journal of Accounting Research, 37,

319–352.Frank, K. (2002). The effect of growth on value relevance of accounting data.Journal of Business Research, 55(1), 69–78.Friedmann, R., & Sanddorf-Kohle, W. G. (2002). Volatility clustering and nontrading days in Chinese stock markets.

Journal of Economics and Business, 54(2), 193–217.


Fung, H., Lee, W., & Leung, W. K. (2000). Segmentation of A- and B-share Chinese equity markets.The Journal ofFinancial Research, 23(2), 179–195.

Ge, J. S., & Lin, Z. J. (1993). Economic reforms and accounting internationalization in China.International Journal ofAccounting. Auditing and Taxation, 2(2), 129–143.

Harris, M. S., & Muller, K. A., III. (1999). The market valuation of IAS versus US-GAAP accounting measures usingForm 20-F reconciliations.Journal of Accounting and Economics, 26:, 285–312.

Holmes, W. C. (1997). China’s financial reforms in the global market.Law and Policy in International Business, 28,715–777.

Holthausen, R. W., & Watts, R. L. (2001). The relevance of the value-relevance literature for financial accounting standardsetting.Journal of Accounting Research, 31, 3–75.

Jiang, Y. H. (2001). A comparison of the differences in earnings reported under two sets of accounting standards byB-share companies listed in SHSE and SZSE.Securities Market Journal (in Chinese), 2, 36–39.

Joos, P., & Lang, M. (1994). The effects of accounting diversity: evidence from the European Union.Journal of AccountingResearch, 32(Supplement), 141–168.

Karmel, S. M. (1994). Emerging securities markets in China; capitalism with Chinese characteristics.The China Quarterly,140, 1105–1116.

Kothari, S. P., & Zimmerman, J. L. (1995). Price and return models.Journal of Accounting and Economics, 20, 155–192.Kothari, S. P. (2001). Capital market research in accounting.Journal of Accounting and Economics, 31, 105–231.Lev, B., & Zarowin, P. (1999). The boundaries of financial reporting and how to extend them.Journal of Accounting

Research, 37, 353–385.Li, X. X. (1999). The information disclosure of Chinese listed companies: its status quo and prospect. InIn Contemporary

Issues in China Accounting and Finance. Hong Kong: Oxford University Press (China) Ltd, 334–338.Lin, Z. J., & Chen, F. (1999). Applicability of the conservatism accounting convention in China: empirical evidence.The

International Journal of Accounting, 34, 517–537.Lin, Z. J., & Chen, F. (2000). Toward internationally accepted accounting practices: improvement of Chinese accounting

regulations.Research in Accounting in Emerging Economies, 4, 3–15.Lin, Z. J., Chen, F., & Tang, Q. L. (2001). An empirical evaluation of the new system of business accounting in China,

Journal of International Accounting. Auditing & Taxation, 10, 23–49.Mills, R., & Cao, Y. (1996). Accounting and financial reporting in China.Management Accounting, 74, 26–30.Mookerjee, R., & Yu., Q. (1999). An empirical analysis of the equity market in China.Review of Financial Economics,

8, 41–60.Niskanen, Kinnunen, J. J., & Kasanen, E. (2000). The value relevance of IAS reconciliation: empirical evidence from

Finland.Journal of Accounting and Public Policy, 19(2), 119–137.Ohlson, J. A., & Shroff, P. K. (1992). Changes versus levels in earnings as explanatory variables for returns: some

theoretical considerations.Journal of Accounting Research, 30, 210–226.Ohlson, J. A. (1995). Earnings, book value, and dividends in security valuation.Contemporary Accounting Research,

11(2), 661–687.Rees, L., & Elgers, P. (1997). The market’s valuation of nonreported accounting measures: retrospective reconciliations

of non-U.S. and U.S. GAAP.Journal of Accounting Research, 35, 115–127.Su, D., & Fleisher, B. M. (1998). Risk, return and regulation in Chinese stock markets.Journal of Economics and Business,

50, 239–256.Tang, Y. W. (2000). Bumpy road leading to internationalization: a review of accounting development in China.Accounting

Horizons, 14, 93–102.Wallace, W. (2000). The value relevance of accounting: the rest of the story.European Management Journal, 18(6),

675–682.Winkle, G. M., Huss, H. F., & Chen, X. Z. (1994). Accounting standards in the People’s Republic of China: responding

to economic reforms.Accounting Horizon, 8(3), 48–57.Xiao, Z., & Pan, A. (1997). Developing accounting standards on the basis of a conceptual framework by the Chinese

government.The International Journal of Accounting, 32(3), 279–299.Zhang, T. X. (1998). The accounting environment and the unreliability of financial accounting information.Accounting

Research (in Chinese) No., 6.

value relevance of international accounting standards harmonization: evidence from a- and b-share...

Documents