Convergence to IFRS, Accounting Quality and the Role of Regional Institutions: Evidence from China. Maggie Hao Minghe Sun Jennifer Yin*

Transcription

1 Convergence to IFRS, Accounting Quality and the Role of Regional Institutions: Evidence from China Maggie Hao Minghe Sun Jennifer Yin* School of Business University of Texas at San Antonio USA *Corresponding Author. Department of Accounting, University of Texas at San Antonio, One UTSA Circle, San Antonio, TX Phone: (210) Acknowledgement: We thank participants at the 2013 American Accounting Association Annual Conference for their comments and suggestions and the University of Texas at San Antonio College of Business-Office of International Engagement for their generous support of this research. 0

2 Convergence to IFRS, Accounting Quality and the Role of Regional Institutions: Evidence from China Abstract Through the transition from a regulated economy to a market-oriented economy, China has pursued a gradual approach in its institutional reform. This approach, however, leads to an uneven institutional development across regions. In this study, we examine how this unevenness in regional institution development affects firms accounting quality after China converges with the International Financial Reporting Standards. We find that, in general, the convergence to IFRS is associated with an increase in aggressive reporting of accruals for Chinese firms. Further analysis on the regional institution development level reveals that there is an incremental increase in accrual aggressiveness for firms located in regions with lower degree of legal environment development. Notwithstanding the insignificant change in the trend of timely loss recognition after the convergence to IFRS in general, the piecewise analysis suggests a significant decrease in the likelihood of reporting large negative net income for firms located in regions with lower degree of legal environment development. 1

3 1. Introduction Through the transition from a regulated economy to a market-oriented economy, China has pursued a gradual approach in its institutional reform. This approach, however, leads to an uneven institutional development across regions. In this study, we examine how this unevenness in regional institution development affects firms accounting quality after China converges with the International Financial Reporting Standards (IFRS). On February 15, 2006, China announced the adoption of a new accounting standard system. The new system has bought about substantial convergence between Chinese accounting standards and IFRS. It also affirms China, the world largest emerging economy s commitment to convergence. However, some researchers concern that the convergence of accounting standards may not lead to the convergence of accounting practice. Ball (2006) argues that deep-rooted political and institutional factors (i.e. extent and nature of government involvement in the economy, legal system, securities regulation and regulatory bodies and depth of financial markets) influence the incentives of financial statement preparers and inevitably shape actual financial reporting practice. Most importantly, these political or institutional factors primarily remain local (Ball, 2006). Existing accounting literature (e.g. Pope and Walker, 1999; Ball, Kothari and Robin, 2000; Ball, Robin and Wu, 2003; Ali and Hwang, 2000; Leuz, Nanda and Wysocki, 2003; Bushman and Piotroski, 2006; Leuz and Oberholzer, 2006) provides evidence showing the association between the cross-country difference on institutional development and accounting quality. Whereas the effect of within-country difference, or the effect of regional institutions, has not been examined. The World Bank, in its 2009 assessment of the Chinese accounting and auditing practices, commends China on making impressive progress in 2

4 establishing an institutional framework for accounting, auditing, and corporate financial reporting. However, the World Bank also expresses concerns regarding the quality of accounting and auditing practices outside major cities in China 1. In this study, we first investigate whether and to what extent the convergence with IFRS improves Chinese firms accounting quality. Secondly, we examine the role played by the regional institutions in this process. We focus on two aspects of accounting quality: the accrual aggressiveness and the timely loss recognition. Specifically, we test: 1) whether the convergence to IFRS significantly lowers the accrual aggressiveness proxied by the magnitude of discretionary accruals; 2) whether the convergence to IFRS significantly enhances the timely loss recognition proxied by the likelihood of reporting large negative net income; and 3) whether the effects of convergence with IFRS on accounting quality vary with the quality of regional institutions. We focus our study on Chinese firms because in China, the level of government s involvement in business, and market and legal institutions vary significantly across regions. This unevenness provides us a unique opportunity to examine the relation between regional institutions and their effects on accounting quality. Following prior research (i.e. Wang et al. 2008; Chan et al. 2010), we use three institutional indexes issued by the National Economic Research Institute of China (NERI) 2 to measure the quality of regional institutions. The three indexes evaluate the regional institutions quality from the degree of government influence in the local economy, the development of credit market, and the development of legal environment The NERI index is an assessment system for relative progress in marketization level of China s provinces, which is assessed in five fields by a total of 23 basic indicators. Data are either from statistics or enterprise and household surveys. 3

5 We find that, in general, the convergence to IFRS is associated with an increase in aggressive reporting of accruals for Chinese firms. The further analysis on the regional institution development level reveals that there is an incremental increase in accrual aggressiveness for firms located in regions with lower degree of legal environment development. Notwithstanding the insignificant change in the trend of timely loss recognition after the convergence to IFRS in general, the piecewise analysis suggests a significant decrease in the likelihood of reporting large negative net income for firms located in regions with lower degree of legal environment development. This study contributes to the accounting literature in several ways. First, we extend accounting literature regarding institutional factors by examining the association between the regional institutions and accounting quality. Second, by adopting a within-country setting, we preclude cross-country confounding factors caused by country-specific accounting rules and regulations, differences in infrastructure and culture, and other potential endogeneity problems (Chan et al. 2010). Third, while a lot of attention has been focused on the European and U.S. application of IFRS, the application of IFRS taking place in emerging markets is overshadowed. By examining China, the world largest emerging economy in the process of economic transition, this study sheds light on the effect of convergence with IFRS on accounting quality for emerging or transitional economies. The remainder of this proposal is organized as follows. The next section provides background and hypothesis development. Section three provides discussions on the empirical research design. Section four discusses the empirical results, and section five concludes this study. 4

6 2. Background and Hypothesis Development 2.1 Chinese Accounting Standards and Convergence to IFRS The old Chinese accounting system was inherited from the former Soviet Union and was designed to serve the centrally controlled economy (Tang et al. 1996; Tang, 2000). Since the Chinese stock market was reestablished in the early 1990s 3, China has issued four sets of accounting regulation effective in 1993, 1998, 2001 and 2007 respectively. Each set of standards replaced the previous one and was considered to be in greater conformity with IFRS (Chen et al., 2002; Pacter & Yuen, 2001; ISAB 2006; Peng et al., 2008). On February 15, 2006, China announced the issuance of the new Accounting Standards for Business Enterprises (ASBEs), signaling the achievement of a milestone on its convergence to IFRS. Starting from January 1, 2007, all public-listed companies in China are required to comply with the new standards. The new ASBEs, commonly referred to as the 2007 Chinese Generally Accepted Accounting Principles (GAAP), consist of a new basic standard and 38 specific ASBEs and cover almost all of the topics under the current IFRS s literature. Compared to the previous set of accounting standards, the 2007 GAAP not only expands the disclosure requirements, but also makes fundamental changes in many areas. The key differences between the 2007 GAAP and the previous version include but are not limited to the following new provisions 4 : 1) fair value measurement is widely adopted; 2) reversal of impairment loss for fixed assets and intangible assets is prohibited; 3) the tax 3 The two Chinese stock exchanges, Shanghai Stock Exchange (SHSE) and Shenzhen Stock Exchange (SZSE), were established in year 1990 and year 1991 respectively, and expanded rapidly thereafter. 4 The discussion in this section draws from China s New Accounting Standards: A Comparison with Current PRC GAAP and IFRS, prepared by Deloitte Touche Tohmatsu, August

7 effect of accounting method replaces the tax payable method to record temporary differences; 4) the use of the LIFO method to assign the cost of inventory is prohibited; 5) capitalization of the borrowing costs is allowed for a broader scope of assets; and 6) the pooling of interests method is used for business combinations involving entities under common control. The new accounting standards bring about substantial convergence between Chinese GAAP and IFRS, and significantly affect the quality of accounting information in the Chinese stock market. For example, the impairment loss for fixed and intangible assets was abused under the previous version of accounting standards. To manage earnings, firms may accrue such an impairment loss in the good years and reverse it subsequently when they experience operating loss. Zhang et al. (2010) compare the long-term assets impairment loss between the transition period (the period after announcement of the new standard and the before the effective date) and pre-announcement period. They find that the mean long-term assets impairment charge fell by 29.4% after the announcement of the new standards. The study also shows that firms with substantial previous write-downs reversed more impairment charges to achieve their earnings targets in the transition period (Zhang et al., 2010). Another significant impact on the financial reporting comes from the wide adoption of fair value measurement. As Barth (2007) suggests, IASB considers fair value accounting more relevant in many situations because: 1) it reflects the present economic conditions; 2) it can faithfully represents assets and liabilities, therefore, it is unbiased; and 3) it enhances consistency since it reflects the same type of information in every period. Under the 2007 Chinese GAAP, 17 out of 38 ASBEs have used fair value measurement at some level. 6

8 2.2 Hypothesis Development Some accounting literature (i.e. Barth et al., 2006; Gassen and Sellhorn, 2006; Hung and Subramanyam, 2007; Barth et al., 2008) suggest that the application of IFRS will improve accounting quality. Barth et al. (2008) contributes the expected enhancement on accounting quality to three primary features of IFRS. First, a goal of International Accounting Standards Committee (IASC) and IASB is to develop an internationally acceptable set of high quality financial reporting standards (Barth et al., 2008). To ensure that IFRS are applicable to countries with diverse accounting traditions and varying institutional conditions, IFRS are issued as principle-based standards. This principle-based approach makes IFRS hard to circumvent. Second, in the process of developing a set of high quality standards, IASC and IASB take steps to remove allowable accounting alternatives (IASC, 1989). By doing so, IFRS limit the management s opportunistic discretion in determining accounting amounts (Ashbaugh and Pincus, 2001), and thus, improve accounting quality. The last feature of IFRS is the increasingly use of fair value in financial reporting. IASB concludes that in some cases fair value meets the conceptual framework criteria better than other measure bases considered (Barth, 2007), since fair value is relevant, comparable, and unbiased. This wide adoption of fair value measurements makes IFRS better reflect a firm s economic condition and performance, therefore, helps investors make better investment decisions. Alternatively, some researchers concern that the convergence of accounting standards may not lead to the convergence of accounting practice, especially for the emerging or transitional economies. Eccher and Healy (2003) suggest that the standards developed by the IASB are primarily based on those for countries with highly developed capital market. It is 7

9 questionable whether such standards are also optimal for developing and transitional economies that lack the infrastructure for monitoring managers financial reporting decisions (Peng et al., 2008). Though the principle-based approach makes IFRS more applicable, therefore, hard to circumvent. However, the flexibility embedded in IFRS might render it ineffective in restricting earnings management of firms with low incentives to comply (Christensen et al., 2006). Similarly, fair value accounting is not a panacea. Despite its advantages, fair value accounting also brings about many application issues. Some commonly expressed concerns include lack of a clear definition of fair value 5, lack of verifiability, the ability for management to affect fair value estimates, and the potential circularity of reflecting values in financial statements when the objective is to provide financial statement users with information to make economic decisions that include assessing the value of the entity (Barth, 2007). In addition to the conflicting theoretical opinions, overall, the evidence on the association between IFRS application and accounting quality is quite mixed. By examining firms from 21 countries between 1994 and 2003, Barth et al. (2008) finds that firms voluntarily adopting the International Accounting Standards (IAS) show less earnings management, more timely loss recognition, and more value relevance of accounting amounts during the post-adoption periods. Ahmed, Neel, and Wang (2010) focus their cross-country testing on firms that mandatorily adopt IFRS in In contrast, they find that the mandatory IFRS adoption results in smoother earnings, more aggressive reporting of accruals, and a reduction in timeliness of loss recognition. Moreover, the result is more pronounced 5 The IASB defines fair value as the amount for which an asset could be exchanged, or a liability settled between knowledgeable, willing parties in an arm s length transaction (IAS 39 para. 9; IASB, 2004). Although this definition states the measurement objective, it lacks sufficient specificity to ensure consistent application (Barth, 2007). 8

10 for firms in countries with strong legal enforcement. They attribute their findings to different incentives between mandatory adopters and voluntary adopters, since voluntary adopters choose to adopt IFRS and thus are likely to have stronger incentives to report higher quality accounting numbers (Soderstrom and Sun, 2007; Daske et al., 2008; Ahmed et al., 2010). In single country studies, some literature suggest that firms preparing financial statement based on IAS or IFRS show more value relevant in earnings (i.e. Bartov, Goldberg, and Kim, 2005), and less earnings smoothing (i.e. Bagaeva, 2009) than firms using local GAAP. On the other hand, Jeanjean et al. (2008) finds that the pervasiveness of earnings management did not decline after the introduction of IFRS, but instead increased in France. They interpret this result due to the role played by management incentives and national institutional factors in framing financial reporting characteristics. Using Swedish evidence, Paananen (2008) shows that there has been no improvement in financial reporting quality after Sweden adopted IFRS in 2005, and this result even holds for committed adopters. Even for the same country, like China, the evidence of the convergence effects is quite mixed. By comparing the Chinese firms that are issuing A-share and B-share 6 between 1994 to 2000, Zhou et al. (2009) find that firms that adopting IFRS (B-share firms) are less likely to smooth earnings in the post-adoption period, but they fail to find evidence indicating lower tolerance for reporting losses or more timely loss recognition for adopting firms. Liu et at. (2011) directly examine the impact of IFRS on accounting quality after new substantially IFRS-convergent accounting standards became effective for Chinese listed firms in They find that accounting quality improved with decreased earnings 6 A-share investors receive accounting information prepared under the Chinese GAAP and audited by local firm, while B-share investors receive accounting information prepared under the IFRS and audited primarily by international accounting firms (Zhou et al. (2009)). Thus, before China converges its accounting standards to IFRS, there is a substantial difference on the reports provided to A-share investors and B-share investors. 9

11 smoothing and increased value relevance during the period between 2005 to Covering a little longer period (2004 to 2009), Lin et al. (2012) show less earnings smoothing, timelier loss recognition, but higher tendency to manage earnings toward a positive target post the convergence to IFRS. Given the conflicting theoretical opinions and mixed empirical evidence mentioned above, the first two hypotheses we would like to test are: H 1 : The convergence to IFRS significantly lowers the aggressive reporting of accruals for Chinese listed firms. H 2 : The convergence to IFRS significantly enhances the timely loss recognition for Chinese listed firms. The impact of IFRS on accounting quality reflects combined effects of standards, interpretation, enforcement and litigation. Even if IFRS are higher quality standards, the effects of features of the financial reporting system (i.e. institutional development level, enforcement, and litigation) other than the standards themselves could eliminate any improvement in accounting quality arising from adopting IFRS (Barth et al. 2008). Ball (2006) argues that the deep-rooted political and institutional factors (i.e. extent and nature of government involvement in the economy, legal system, securities regulation and regulatory bodies and depth of financial markets) will influence the incentives of financial statement preparers and inevitably shape actual financial reporting practice. Most importantly, these political or institutional factors primarily remain local (Ball, 2006). Existing accounting literature on institutional factors primarily use cross-country settings to investigate the impact of institutional difference on accounting quality (i.e. Pope and Walker, 1999; Ball, Kothari and Robin, 2000; Ball, Robin and Wu, 2003; Ali and Hwang, 2000; Leuz, Nanda and Wysocki, 2003; Bushman and Piotroski, 2006; Leuz and Oberholzer, 10

12 2006). However, the impact on accounting quality due to regional difference on levels of institutional development within a single country has yet to be determined. Through the transition from a regulated economy to a market-oriented economy, China has pursued a gradual approach in its institutional reform. This approach, however, leads to an uneven institutional development across regions. The large disparity in institutional development within China provides us a unique opportunity to examine the role played by regional institutions during the process of convergence, and motivates our third hypothesis: H 3: The impact on accounting quality after China converges with IFRS varies with the quality of regional institutions. 3. Methodology 3.1 Testing period and Sample Selection This study focuses on the last set of Chinese accounting standards effective on January 1, The testing period is from year 2001 to year 2010, which covers the last two sets of accountings standards effective in year 2001 and 2007 respectively (commonly referred to as the 2001 GAAP and 2007 GAAP). We treat year 2001 to year 2006 as the preconvergence period, and year 2007 to year 2010 as the post-convergence period. In the Chinese stock market, there are two types of tradable shares: domestic investment shares (A-shares) and foreign investment shares (B-shares). A-shares can only be issued to Chinese citizens or Chinese institutions in the local currency, while B-shares are originally issued to foreign investors only and traded in U.S. dollars or Hong Kong dollars. Since 2001, domestic investors have been allowed to purchase B-shares with foreign currencies. From a reporting perspective, firms that issue A-share are required to prepare 11

13 financial statements following the Chinese GAAP, and firms that issue B-shares follow IFRS to prepare financial statements. A listed company on the Chinese stock market is authorized to issue A-shares only, B-shares only, or both A-shares and B-shares, which is referred to as dual listing. To examine the effect of adoption of 2007 GAAP on accounting quality, our initial sample includes all non-financial firms that only issue A-shares in Chinese stock market. 3.2 The effect of convergence on accrual aggressiveness Researchers frequently use measures of discretionary accruals in tests for earnings management and market efficiency (Kothari, Leone, and Wasley 2005). Following previous studies, we use the modified Jones Model to calculate discretionary accrual. For every year from 2001 to 2010, we first estimate the following model by fiscal year and industry: TA it = β 0 + β 1 (1/AT it-1 ) + β 2 (ΔREV it ΔAR it ) +β 3 PPE it +ε it (1) Where: TA it AT it-1 =The total accrual of firm i in year t, calculated as the difference between income before extraordinary items and operating cash flows; =Total assets of firm i in year t-1; ΔREV it =The change in sales of firm i from year t-1to year t; ΔAR it =The change in accounts receivable of firm i from year t-1to year t; PPE it =The gross property, plant, and equipment of firm i in year t. Based on the coefficients estimated from model (1), we then calculate discretionary accrual as the difference between the reported and fitted value of total accrual. In the calculation of discretionary accrual, we trim one percent for all independent variables. 12

14 To analyze the effect of convergence on accrual aggressiveness, we estimate the following equation: ABSDA = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE + β 7 AUD + β 8 CONVERG+ e it (2) Where: ABSDA ROA LEV SIZE GROWTH AUD EISSUE DISSUE = absolute value of discretionary accrual estimated by using modified Jones Model; =net income divided by total assets; = total liabilities divided by total assets; = natural logarithm of total assets; = market value of capital divided by total common holder s equity; = 1 if the auditor is Big four auditors, and 0 otherwise; = the percentage change in common stock outstanding; = the percentage change in total liabilities; CONVERG = a dummy variable equals to one if fiscal year is 2007, 2008, 2009, or 2010, and zero otherwise. Following prior research, we include several independent variables to control for differences in firm s economic factors that might impact discretionary accruals. The control variables used in this study include an indicator of firm s performance (ROA), a firm s debt ratio (LEV), firm size (SIZE), a firm s growth opportunities (GROWTH), Big Four auditor (AUD), and the percentage changes in common stock outstanding (EISSUE) and liabilities (DISSUE). Kothari et al. (2005) argue that accruals of firms that have experienced unusual performance are expected to be systematically nonzero and are, therefore, correlated with firm performance. To control for the effect of firm s performance on the level of earnings management, we include return on assets (ROA) in our regression models. Dechow and Dichev (2002) suggest that larger firms have relatively stable operations and better control 13

15 system. Hence, such firms might have lower levels of accruals, on average. We use the natural logarithm of total assets (SIZE) to capture the potential effect of firm size on earnings quality. Prior studies also show that leverage is associated with a firm s incentive to engage in earnings management (e.g., Becker et al. 1998, DeFond and Jiambalvo 1994). To control for the possible effect of leverage on accruals management, we include LEV as a control variable. Skinner and Sloan (2002) document that firms with growth opportunities are penalized more by the stock market when they miss earnings thresholds (Roychowdhury 2006). The relative stronger market reaction to the earnings of growth firms may provide an incentive for high growth firms to manipulate earnings to meet certain earnings thresholds. Following Skinner and Sloan (2002) and Hribar et al. (2004), we use the ratio of the market value of equity to the book value of equity to proxy firm s growth opportunities (GROWTH) in our regression. We also include a dummy variable AUD to capture the effect of audits performed by Big Four auditors on earnings quality. Furthermore, following Barth et al. (2008), we include control variables of percentage change in common stock outstanding (EISSUE) and percentage change in total liability (DISSUE) in our regression models. If the convergence to IFRS lowers the magnitude of earnings management, we expect that the coefficient (β 8 ) for CONVERG is negatively associated with the dependent variable and statistically significant. 3.3 The effect of convergence on timely loss recognition Following Barth et al. (2008), Chen et al. (2010) and Ahmed et al. (2011), we use the frequency of large loss recognition (the loss greater than of average total assets) to measure timely loss recognition. To examine whether the convergence with IFRS enhance the timelier loss recognition, we estimate: 14

16 LNEG = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE (3) Where: LNEG =1 if net income scaled by average total assets is less than -0.20, and 0 otherwise; ROA =net income divided by total assets; LEV = total liabilities divided by total assets; SIZE = natural logarithm of total assets; GROWTH = market value of capital divided by total common holder s equity; AUD = 1 if the auditor is Big four auditors, and 0 otherwise; EISSUE = the percentage change in common stock outstanding; DISSUE = the percentage change in total liabilities; CONVERG = a dummy variable equals to one if fiscal year is 2007, 2008, 2009, or 2010, and zero otherwise. If the convergence with IFRS significantly enhances the timelier loss recognition, we expect the coefficient (β 8 ) for CONVERG is positively associated with the dependent variable and is statistically significant. 3.4 Measure of regional institutions quality Following Wang et al. (2008) and Chan et al. (2010), we use three institutional indexes issued by the NERI of China to measure regional institution development level. The first one is the government decentralization index. In transitional economies, governments (national or regional) possess considerable control over the allocation of resources either through their power of planning or through their control of state-owned enterprises, including banks (Nee 1992; McMillan 1997). The government decentralization index is constructed based on regional government s spending as a percentage of GDP, a survey result regarding 15

17 the time spent by corporate executives in dealing with the government, the tax rates in the region, and the size of local government. Higher government decentralization index suggests lower degree of local government intervention in the region. The second index is the credit market index. Jin and Qian (1998) find that in areas with a lower level of market development, firms that are more closely affiliated with local governments enjoy more credit advantages (Chan et al., 2010). The credit market index is measured by the share of non-state owned financial institutions in total deposits and the share of bank loans issued to non-state owned enterprises. The NERI report shows that China s credit market is more developed in east coastal regions. Higher credit market index indicates higher development level of credit market. The last index is the intermediates and legal environment index. IFRS are principlesbased, which means that auditors and accountants need to follow general principles rather than detailed standards and adapt these principles to specific situations (Ball, 2006). Therefore, the quality of intermediates (i.e. accountant and lawyer) and legal environment is very important in determining accounting quality under situations when an interpretation of the principles is required. The intermediates and legal environment index is measured by the percentage of lawyers and public accountant in the local population, legal environment for businesses as perceived by corporate executives, protection of intellectual property rights in terms of patent applications and research and development (R&D)grants, and protection of consumer rights (Chan et al., 2010). Higher index suggests a higher percentage of CPA or lawyer in local population and a higher development level of for legal environment. These three indexes collectively capture the underlying features of the regional institutional environment. 16

18 To analyze how regional difference on institutional development affects improvement on accounting quality after the convergence with IFRS, we estimate the following piecewise regression model: ABSDA/LNEG = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE + β 7 AUD + β 8 CONVERG+ β 9 LG_CONVERG + β 10 LC_CONVERG + β 11 LL_CONVERG +e it (4) Where: LG_CONVERG LC_CONVERG LL_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the government decentralization index is less than the average; = one if fiscal year is 2007, 2008, 2009, or 2010, and the credit market index is less than the average; = one if fiscal year is 2007, 2008, 2009, or 2010, and the legal environment index is less than the average. For dependent and control variables of model (4), please see model (2) and (3) respectively. If the quality of regional institutions influences the effect of convergence on accounting quality, we expect the coefficients (β 9, β 10, and β 11 ) on the incremental variables LG_CONVERG, LC_CONVERG and LL_CONVERG to be statistically significant. 4. Result The sample of Chinese firms used in this analysis is drawn from those in the CMSAR database. There are two stock exchanges in Chinese stock market: Shanghai stock exchange and Shenzhen stock exchange. Due to the availability of physical location data about the listed firms, we limit our sample to firms listed in Shanghai stock exchange only. Our initial sample consists of 6,791 firm-year observations from CMSAR database. We further 17

19 eliminated observations from banking industry, lacking of Institutional Development Index information, cross-listed in B-share market or H-share market, and with insufficient data in computing regression variables. Our final sample consists of 605 firms with 4,195 firm-year observations. Details of the sampling process are shown in Table 1. <Insert Table 1 Here> Table 2 reports the mean and median for key regression variables between the preconvergence period ( ) and post-convergence period ( ). The mean (median) of ABSDA is 0.044(0.026) during the pre-convergence period, and increases to (0.036) after the convergence. The mean of LNEG decreases from in preconvergence period to after the convergence. The mean and median of ROA, LEV, SIZE, GROWTH, EISSUE, DISSUE increase after the convergence to IFRS, but the mean likelihood of hiring big-four auditors (AUD) decreases from to In addition, the differences in ABSDA, LNEG, ROA, LEV, SIZE, EISSUE, DISSUE, and AUD between pre-convergence and post-convergence periods are mostly significant at one percent levels in both parametric t-test and nonparametric Wilcoxon Z-test. < Insert Table 2 Here > Table 3 displays correlation matrix for the regression variables. As expected, we find that the proxy for accrual aggressiveness (ABSDA) is positively associated with the convergence dummy (CONVERG), while the association between CONVERG and the proxy 18

20 for timely loss recognition (LNEG) is negative. We also find that the likelihood to hire bigfour auditors (AUD) is negatively associated with CONVERG. < Insert Table 3 here > To test H 1 regarding the IFRS convergence effect on accrual aggressiveness, we run regression model (2) for the pooled sample. The results are presented in Table 4. Notably, the coefficient relating post-convergence period to accrual aggressiveness (β 8 ) are positive (0.0112) and significant at the 1% probability level (p-value=0.0000). This result suggests that the convergence to IFRS, in general, is associated with an increase in accrual aggressiveness. Furthermore, we document that the magnitude of discretionary accrual is positively associated with LEV (0.0021, p-value=0.0080), EISSUE (0.0124, p-value=0.0000), DISSUE (0.0087, p-value=0.0000), but negatively associated with ROA ( , p- value=0.0000). Model (4) is a piecewise version of model (2), which allows the coefficient on the experimental variable (CONVERG) to vary with different levels of regional institutional development. As in model (2), CONVERG is the indicator dummy expected to capture the reporting behavior of accrual aggressiveness after the convergence to IFRS. The coefficients on the interaction terms, LG_CONVERG, LC_CONVERG and LL_CONVERG, test the incremental accrual aggressiveness of firms in regions with higher government intervention, lower development level of credit market or legal environment. The results of estimating model (4) are also presented in Table 4. The coefficient for CONVERG is positive and significant (β 8 = , p-value=0.0000). The only interaction term significant in model (4) 19

21 is LL_CONVERG. The coefficient for LL_CONVERG is and significant at 5% probability level (p-value=0.0190). Taken together, these results suggest that the convergence to IFRS is associated with an increase in accrual aggressiveness, and this association is stronger in regions with lower percentage of CPA or lawyer in local population and a lower development level of legal environment. < Insert Table 4 Here > To evaluate whether the convergence to IFRS significantly improves the timely loss recognition, we run model (3). The results are presented in Table 5. The coefficient for the convergence dummy, CONVERG, is negative but not significant, suggesting that the convergence to IFRS does not significantly affect the likelihood of reporting large negative net income for Chinese listed firms. We further run the piecewise regression of model (4). We find that though the experiment variable, CONVERG, is still insignificant, the coefficient for the interaction term LL_CONVERG is negative and significant ( , p- value=0.0290). This result is consistent with a decrease in the timely loss recognition for firms located in regions with lower percentage of CPA or lawyer in local population and a lower development level of legal environment after the convergence to IFRS. < Insert Table 5 Here > 5. Limitation and Conclusion Overall, our findings can be summarized as follows. First, we find evidence of a significant increase in the aggressive reporting of accruals for Chinese firms after the convergence to IFRS. In addition, there is an increase in accrual aggressiveness for firms in 20

22 regions with lower percentage of CPA or lawyer in local population and a lower development level of legal environment incremental to the general upward trend over our sample period. Second, we do not find a significant change in the likelihood of reporting large negative net income due to the convergence to IFRS in general. However, the piecewise regression results suggest that the timely loss recognition decreases for firms located in regions with lower percentage of CPAs or lawyers in local population and a lower development level of legal environment. 21

23 Table 1: Sample Selection Process Sample Selection Process Observations Firms Firm s listed in Shanghai Stock Exchange with financial information for the period of 2001 to 2010 available from the CSMAR database 6,791 Less: Observations from banking industry -73 Observations lack of Institutional Development Index information -642 Observations cross-listed in B-share market or H-share market -729 Observations with insufficient data in computing regression variables -1,152 Final Sample 4,

24 Table 2: Descriptive Statistics of Variables Pre ( ) Post ( ) Difference Mean Median Mean Median Mean Median Variables (A) (B) (C) (D) (A-C) (B-D) ABSDA *** *** LNEG *** 0.000*** ROA *** *** LEV *** *** SIZE *** *** GROWTH EISSUE *** *** DISSUE ** AUD *** 0.000*** TURN *** *** CF *, **, *** Denote significantly different from 0 at the 10, 5, and 1 percent levels, respectively. Variable Definitions: ABSDA = absolute value of discretionary accrual estimated by using modified Jones Model; LNEG = one if net income scaled by average total assets is less than -0.20, and zero otherwise; ROA = net income divided by total assets; LEV = total liabilities divided by total assets; SIZE = natural logarithm of total assets; GROWTH = market value of capital divided by total common holder s equity; EISSUE = the percentage change in common stock outstanding; DISSUE = the percentage change in total liabilities; AUD = one if the auditor is Big four auditors, and zero otherwise; TURN = sales divided by total assets; CF = net cash provided by operating activities divided by total assets. 23

25 Table 3: Pearson Correlation Coefficient Matrix A B C D E F G H I G K L ABSDA A 1.00 LNEG B ROA C LEV D SIZE E GROWTH F EISSUE G DISSUE H AUD I TURN G CF K CONVERG L Variable Definitions: ABSDA = absolute value of discretionary accrual estimated by using modified Jones Model; LNEG = one if net income scaled by average total assets is less than -0.20, and zero otherwise; ROA = net income divided by total assets; LEV = total liabilities divided by total assets; SIZE = natural logarithm of total assets; GROWTH = market value of capital divided by total common holder s equity; EISSUE = the percentage change in common stock outstanding; DISSUE = the percentage change in total liabilities; AUD = one if the auditor is Big four auditors, and zero otherwise; TURN = sales divided by total assets; CF = net cash provided by operating activities divided by total assets; CONVERG = one if fiscal year is 2007, 2008, 2009, or

26 Table 4: Regression Analysis of Magnitude of Discretionary Accruals Model (2) : ABSDA = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE +β 7 AUD + β 8 CONVERG+ e it Model (4): ABSDA = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE + β 7 AUD + β 8 CONVERG+ β 9 LG_CONVERG +β 10 LC_CONVERG + β 11 LL_CONVERG +e it Model (2) Coefficient ( P-Value) Model (4) Coefficient (P-Value) Experimental Variables CONVERG β (0.000) *** (0.0000) *** LG_CONVERG β (0.3560) LC_CONVERG β (0.4380) LL_CONVERG β (0.0190) ** Control Variables ROA β (0.0000) *** (0.0000) *** LEV β (0.0080) *** (0.0110) ** SIZE β (0.3420) (0.3880) GROWTH β (0.1030) (0.0950) * EISSUE β (0.0000) *** (0.0000) *** DISSUE β (0.0000) *** (0.0000) *** AUD β (0.5100) (0.4690) 25

27 Intercept β (0.0630) * (0.0760) * Adjusted R Square 4.27% 3.94% F-Value P-Value (0.0000) *** (0.0000) *** *, **, *** Denote significantly different from 0 at the 10, 5, and 1 percent levels, respectively. Variable Definitions: ABSDA = absolute value of discretionary accrual estimated by using modified Jones Model; ROA = net income divided by total assets; LEV = total liabilities divided by total assets; SIZE = natural logarithm of total assets; GROWTH = market value of capital divided by total common holder s equity; EISSUE = the percentage change in common stock outstanding; DISSUE = the percentage change in total liabilities; AUD = one if the auditor is Big four auditors, and zero otherwise; CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010; LG_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the government decentralization index is less than the average; LC_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the credit market index is less than the average; LL_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the legal environment index is less than the average. 26

28 Table 5: Regression Analysis of Timely Loss Recognition Model (3) : LNEG = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE +β 7 AUD + β 8 CONVERG+ e it Model (4): LNEG = α + β 1 ROA+ β 2 LEV+ β 3 SIZE+ β 4 GROWTH + β 5 EISSUE + β 6 DISSUE + β 7 AUD + β 8 CONVERG+ β 9 LG_CONVERG +β 10 LC_CONVERG + β 11 LL_CONVERG + e it Model (3) Coefficient ( P-Value) Model (4) Coefficient (P-Value) Experimental Variables CONVERG β (0.8680) (0.3810) LG_CONVERG β (0.7440) LC_CONVERG β (0.3260) LL_CONVERG β (0.0290) ** Control Variables ROA β (0.0000) *** (0.0000) *** LEV β (0.0030) *** (0.0020) *** SIZE β (0.6820) (0.7210) GROWTH β (0.0000) *** (0.0000) *** EISSUE β (0.0000) *** (0.0000) *** DISSUE β (0.0440) ** (0.0480) ** AUD β (0.9390) (0.8870) Intercept β (0.8810) (0.8420) 27

29 Adjusted R Square 36.12% 36.15% F-Value P-Value (0.0000) *** (0.0000) *** *, **, *** Denote significantly different from 0 at the 10, 5, and 1 percent levels, respectively. Variable Definitions: LNEG = one if net income scaled by average total assets is less than -0.20, and zero otherwise; ROA = net income divided by total assets; LEV = total liabilities divided by total assets; SIZE = natural logarithm of total assets; GROWTH = market value of capital divided by total common holder s equity; EISSUE = the percentage change in common stock outstanding; DISSUE = the percentage change in total liabilities; AUD = one if the auditor is Big four auditors, and zero otherwise; CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010; LG_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the government decentralization index is less than the average; LC_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the credit market index is less than the average; LL_CONVERG = one if fiscal year is 2007, 2008, 2009, or 2010, and the legal environment index is less than the average. 28

30 References Ali, A., and L. Hwang Country-specific factors related to financial reporting and the value relevance of accounting data. Journal of Accounting Research 38 (1): Ahmed, A. S., M. Neel, and D Wang Does mandatory adoption of IFRS improve accounting quality? Preliminary evidence. Contemporary Accounting Research 30 (4): Bagaeva, A The IFRS and accounting quality in the transitional economy. Working paper, University of Oulu. Ball, R., S.P. Kothari, and A. Robin The effect of international institutional factors on properties of accounting earnings. Journal of Accounting and Economics 29 (1): Ball, R., A. Robin, and J. Wu Incentives versus standards: Properties of accounting income in four East Asian countries. Journal of Accounting and Economics 36: Ball, R International Financial Reporting Standards (IFRS): Pros and cons for investors. Accounting and Business Research 36 (supplement 1): Barth, M., W. Landsman, and M. Lang International accounting standards and accounting quality. Journal of Accounting Research 46 (3): Becker, C. L., M. L. Defond, J. Jiambalvo, and K.R. Subramanyam The effect of audit quality on earnings management. Contemporary Accounting Research 15(1): Bushman, R., and J. Piotroski Financial reporting incentives for conservative accounting: The influence of legal and political institutions. Journal of Accounting and Economics 42 (1-2): Carmona S., and M. Trombetta On the global acceptance of IAS/IFRS accounting standards: The logic and implications of the principles-based system. Journal of Accounting and Public Policy 27 (6): Chan, K.H., K.Z. Lin, and B. Wong The impact of government ownership and institutions on the reporting behavior of local auditors in China. Journal of International Accounting Research 9 (2): Chen, G., M. Firth, and J. Kim The use of accounting information for the valuation of dual-class shares listed on China s stock markets. Accounting and Business Research 32 (No. 3):

31 Christensen, H.B., E. Lee, and M. Walker Incentives or standards: What determines accounting quality changes around IFRS adoption? Working paper. Dechow, P., and I. Dichev The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review 77 (Supplement): DeFond, M., and J. Jiambalvo Debt covenant violation and manipulation of accruals. Journal of Accounting & Economics 17 (January): Fan, G., X. Wang, and H. Zhu The report on the relative process of marketization of each region in China. The Economic Science Press (in Chinese). Fan, G., X. Wang, and H. Zhu NERI index of marketization for China s provinces: 2006 report. The Economic Science Press (in Chinese). Jeanjean T. and H. Stolowy Do accounting standards matter? An exploratory analysis of earnings management before and after IFRS adoption. Journal of Accounting and Public Policy 27 (6): Hribar, P., N.T. Jenkins, and W.B. Johnson The stock repurchases as an earnings management device. Journal of Accounting and Economics 41 (April): Kothari, S.P., A. Leone, and C. Wasley Performance matched discretionary accrual measures. Journal of Accounting and Economics 39 (1): Leuz, C., D. Nanda, and P. Wysocki Earnings management and investor protection: an international comparison. Journal of Financial Economics 69 (3): Leuz, C. and F. Oberholzer-Gee Political relationships, global financing, and corporate transparency: Evidence from Indonesia. Journal of Financial Economics 81 (2): Lin, J., J. Neiva de Figueredo, and A.J. Staglilano The impact of IFRS adoption on earnings quality of Chinese firms. Journal of Accounting and Finance. Liu, C., L.J. Yao, N. Hu, and L. Liu The impact of IFRS on accounting quality in a regulated market: An empirical study of China. Journal of Accounting, Auditing & Finance 26 (4): Lu, J., B. Xu, and X. Liu The effects of corporate governance and institutional environments on export behaviour in emerging economies: Evidence from China. Management International Review 49:

32 McMillan, J Markets in Transition. In Advances in Economics and Econometrics, Volume II, edited by D. Kreps and K. Wallis, Cambridge: Cambridge University Press. Nee, V Organizational dynamics of market transition: Hybrid forms, property rights and mixed economy in China. Administrative Science Quarterly 37 (1): Paananen, M The IFRS adoption s effect on accounting quality in Sweden. University of Hertfordshire. Available at SSRN: or Peng, M., W. Wang, and D. Jiang An institutional-based view of international business strategy: A focus on emerging economies. Journal of International Business Studies (39) 5: Peng, S., R.H. Tondkar, J.L. Smith, and D. Harless Does convergence of accounting standards lead to the convergence of accounting practices? The International Journal of Accounting 43: Pope, P. and M. Walker International differences in the timeliness, conservatism, and classification of earnings. Journal of Accounting Research 37 (Supplement): Roychowdhury, S Earnings management through real activities manipulation. Journal of Accounting & Economics 42(3): Skinner, D.J., and R.G. Sloan Earnings surprises, growth expectations, and stock returns or don t let an earnings torpedo sink your portfolio. Review of Accounting Studies 7 (2-3): Tang, Y Bumpy road leading to internationalization: A review of accounting development in China. Accounting Horizons 14 (March): Wang. Q., T.J. Wong, and L. Xia State ownership, the institutional environment, and auditor choice: Evidence from China. Journal of Accounting Economics 46 (1): Zhang, R., Z. Lu, and K. Ye How do firms react to the prohibition of long-lived asset impairment reversals? Evidence from China. Journal of Accounting and Public Policy 29 (5): Zhou, H., Y. Xiong, and G. Ganguli Does the adoption of international financial reporting standards restrain earnings management? Evidence from an emerging market. Academy of Accounting and Financial Studies Journal 13 (Special Issue):