Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

State ownership and the market pricing of accruals quality

State ownership and the market pricing of accruals quality China Journal of aCC ounting StudieS , 2017 Vol . 5, no . 2, 155–172 https://doi.org/10.1080/21697213.2017.1339432 a b b c Shinong Wu , Yaping Wang , Liansheng Wu and Xianhui Bo a b School of Management, Xiamen university, China; guanghua School of Management, Peking university, China; School of a ccountancy, Central university of f inance and economics, China ABSTRACT KEYWORDS a ccruals quality; investor This paper examines the effect of state ownership on the market protection; market pricing; pricing of accruals quality. We find that the market pricing effect state ownership of accruals quality is larger for non-state-owned enterprises than for state-owned enterprises in China. The results still hold when we divide accruals quality into innate and discretionary accruals quality. Further analyses reveal a higher level of expropriation and a greater fundamental risk in non-state-owned enterprises, leading to a larger market pricing effect of accruals quality. This paper is the first to study the governance role of state ownership from the perspective of market pricing of accruals quality. It also adds to the literature on the market pricing of accruals quality by studying its link with state ownership. 1. Introduction The function of accounting information in a capital market relies on two aspects: the account- ing information quality itself and the influence of the quality of accounting information on the capital market. Absence of either impairs the effectiveness of accounting information. Existing studies mainly focus on the quality of accounting information itself but ignore its effect on capital market. One of the important effects of accounting information quality on capital market is the market pricing of accruals quality. When we examine the market pricing of accruals quality, the governance role of state ownership should also be considered. The governance role of state ownership in China is so important that it has received a consider- able amount of attention. Since an agent’s infringement of a principal’s interest may arise where there is information asymmetry (Jensen & Meckling, 1976), accounting information has been a focus in studying the governance role of state ownership. Specifically, we can investigate the effects of state ownership on accruals quality and the market pricing of accruals quality. Extant literature explores the effect of state ownership on earnings quality and finds that the level of earnings management is lower in state-owned enterprises (SOEs) than in non-state-owned enterprises (NSOEs) (Bo & Wu, 2009; Ding, Zhang, & Zhang, 2007). Previous studies also reveal that the impact of accounting information in SOEs on debt covenants is weaker than that in NSOEs (Chen, Chen, Lobo, & Wang, 2010), and that the CONTACT Xianhui Bo 13466715176@163.com *Paper accepted by Jason Xiao. © 2017 a ccounting Society of China 156 S. WU ET AL. governance role of accounting information is also weaker in SOEs than in NSOEs (Shen & Wu, 2012). Yet no research focuses on how state ownership is associated with the capital market effects of accruals quality. Extant literature suggests that China’s stock market is becoming more and more efficient and it is changing from a weak efficient stock market to a semi-strong-form efficient stock market (Zhang & Li, 2011). So China’s stock market can be used to investigate how state ownership is associated with the capital market effects of accruals quality. In this paper, we examine the effect of state ownership on the market pricing of accruals quality using Chinese market data. Using the CAPM and Fama and French three-factor model augmented with an accrual quality risk factor, we find the market pricing effect of accruals quality is larger for NSOEs than for SOEs. The results hold when we divide accruals quality into innate and discretionary accruals quality. Further analyses reveal a higher level of expro- priation by the largest shareholder and a greater fundamental risk in NSOEs than in SOEs, leading to a larger market pricing effect of accruals quality, which means that poorer accruals quality is associated with higher costs of equity capital. To the best of our knowledge, this paper is the first to study the governance role of state ownership from the perspective of market pricing of accruals quality. The existing literature mainly focuses on the effect of state ownership on accruals quality and neglects its effect on the market pricing of accruals quality. In contrast, this paper investigates the different market effects of accruals quality under different types of corporate ownership. This paper enriches our knowledge of the governance role of state ownership. Our study also adds to the growing literature on the market pricing of accruals quality by studying its link with state ownership. Compared with a developed economy, China’s institutional and regulatory envi- ronment features a number of subtle differences, which potentially impact the market pricing of accruals quality. The result in our study shows that the market pricing effect of accruals quality critically depends on the corporate ownership structure, which extends the growing literature on the market pricing of accruals quality. The remainder of the paper proceeds as follows. Section 2 provides the institutional background and hypothesis development. Section 3 outlines the empirical methodology and sample selection. The empirical results are presented in Section 4. Section 5 reports the robustness checks. Section 6 provides further analysis. Section 7 concludes. 2. Institutional background and hypothesis development In an attempt to revitalise inefficient SOEs, the Chinese government partially privatised more than 1,000 SOEs through share issue privatisation in the course of its reform programme from 1990 (Jian & Wong, 2010). Partial privatisation freed about one-third of a company’s shares, while the remaining two-thirds were owned by the state or transferred to ‘legal persons’. State shares were held by government agencies such as the Bureau of State Property Administration (Bai, Lin, Wang, & Wu, 2013). Legal person shares were often held by repre- sentatives of the original SOEs, who retained close ties with local governments (Jiang, Lee, & Yue, 2010). The remaining shares were sold to mainland China investors (A shares), Hong Kong investors (H shares), and foreign investors (B shares). These listed firms are still con- trolled by the state, either directly or indirectly ( Wang, Wong, & Xia, 2008). Because we exam- ine only listed firms, we still refer to them as SOEs. CHINA JOURNAL OF ACCOUNTING STUDIES 157 In April 2005, the China Securities Regulatory Commission (CSRC) launched a state share reform, aiming at converting the non-tradable state-owned shares to tradable shares. As a result of the reform, the Chinese stock market improved dramatically (Huang, Su, & Chong, 2008). According to the report of the Central People’s Government of China in March 2007, 97% of Chinese firms had complied with the orders to reform their share structure. This figure represents 1,301 of the firms that were ordered to complete the reform by the end of 2006. Their aggregated market value was nearly 98% of the market capitalisation of the Shanghai and Shenzhen stock exchanges at that time. After the split-share structure reform in 2007, the institutional background in China became different. Meanwhile, beginning on 1 January 2007, all Chinese-listed companies were forced to implement the new set of accounting standards released by the Ministry of Finance in 2006. The introduction of the new account- ing standards has resulted in fundamental changes to financial reporting practices in China (Deloitte Touche Tohmatsu, 2006). In China’s unique institutional background, there is a significant difference between SOEs and NSOEs. SOEs receive political and financial support from the government, and govern- ment leaders have incentives to assist SOEs (Brandt & Li, 2003; Kornai, 1993; Li & Zhou, 2005). Similarly, the government, as the stock market regulator, also treats SOEs preferentially, providing listing privileges to them based on political rather than economic objectives (Aharony, Lee, & Wong, 2000). Jian and Wong (2010) show that related sales propping is more prevalent among state-owned firms than among non-state-owned firms when earnings are close to the securities regulators’ earnings targets. Since SOEs receive government support while NSOEs do not, the level of propping by the largest shareholder is higher in SOEs than in NSOEs. Because of the highly concentrated ownership structure of Chinese listed firms, the main agency problem is between controlling and minority shareholders (Jiang et al., 2010). In theory, a divergence between cash flow and control rights will motivate the largest share - holder to expropriate firm resources (Claessens, Djankov, Fan, & Lang, 2002). However, this incentive might differ between state and private owners. The former, who represent the largest shareholder in an SOE, might not have a strong incentive to exploit company wealth, although they may make the firm undertake certain social responsibilities, as they often have different goals from those of private block shareholders (Lin & Tan, 1999). The state shareholder will put more weight on the maximisation of social welfare than on the maxi- misation of the wealth of block shareholders (Bai et al., 2013). On the personal level, managers of SOEs are frequently reviewed by government agencies, and their political advancement might depend on their performance. For example, the State Council has an explicit policy guideline to remove managers from SOEs if they are responsible for losses over three con- secutive years. The potential loss of political reputation and forced demotion due to poor performance discourages management from aggressively exploiting the company and expropriating minority shareholders (Li & Qian, 2013). However, the objective of the largest shareholder in the NSOEs is the maximisation of their own interests. Thus, they are very likely to exploit the company because this could improve their welfare. Meanwhile, the manage- ment’s benefits of NSOEs are largely decided by the largest shareholder. As a result, they are inclined to assist the largest shareholder to exploit the company (Bai et al., 2013). Therefore, See http://www.gov.cn. See http://www.sasac.gov.cn. 158 S. WU ET AL. it is more likely that the largest shareholder in a private company will expropriate the resources or interests of minority shareholders in a firm more than those in an SOE. In sum, due to government support and their own incentives, the level of expropriation by the largest shareholder is lower in SOEs than in NSOEs. Extant literature finds supporting evidence. Qian (2001) finds that private controlling shareholders have greater incentives to expropriate company resources than state ones do. Calomiris, Fisman, and Wang (2010) find a negative effect of government ownership on returns at the announcement date and a symmetric positive effect from the policy’s cancellation. Jiang et al. (2010) show that tun- nelling through inter-corporate loans is worse in NSOEs than that in SOEs. Bai et al. (2013) find that fully privatised firms perform worse than state-controlled enterprises owing to excessive expropriation by controlling shareholders after privatisation. According to Yee (2006), reported earnings have two roles, first as a fundamental attrib - ute and second as a financial reporting attribute. Fundamental earnings is the accounting profitability measure that gauges a firm’s ability to make future dividend payments. In contrast, reported earnings is the imperfect signal of fundamental earnings that a firm announces. Earnings quality refers to how quickly and precisely reported earnings reveal fundamental earnings. The higher the quality of earnings, the more quickly and precisely reported earnings convey shocks to the present value of expected dividends (PVED). Corresponding to the two guises of reported earnings are two sources of earnings risk. The first source is fundamental risk. Fundamental risk is unresolved uncertainty in future divi- dend payments, which is not resolved even if one observes the exact (mean) value of PVED. The second risk source is earnings quality risk – additional uncertainty about the value of PVED caused by limited earnings quality. Earnings quality risk is information risk stemming from deficiencies in the accounting rules and firms’ application of those accounting rules in a financial reporting environment. In the absence of fundamental risk, earnings quality risk has no effect on the cost of capital, and that increasing fundamental risk serves to magnify the effect of earnings quality risk on the cost of capital ( Yee, 2006), thus increasing the market pricing effect of earnings quality. Chen, Dhaliwal, and Trombley (2008) find that there is essentially no relation between accruals quality and cost of capital as measured by future return realisations for firms with the lowest fundamental risk. In contrast, for firms with the highest fundamental risk, there is a strong relation between accruals quality and future return realisations. For the reason that the largest shareholders of SOEs have less incentive to expropriate r fi m resources, and the fundamental risk is lower in low expropriation firms than that in high expropriation firms, SOEs have a lower level of fundamental risk than NSOEs. As fundamental risk is positively related to the market pricing effect of accruals quality, the market pricing effect of accruals quality is lower for SOEs than for NSOEs. Therefore, we posit the following hypothesis: H1: The market pricing effect of accruals quality is lower for SOEs than for NSOEs. Guay, Kothari, and Watts (1996) and Subramanyam (1996) argue that financial reporting outcome can be partitioned into innate and discretionary components. Innate accruals qual- ity is driven by innate features of the firm’s business model and operating environment, and discretionary accruals quality is due to accounting choices, implementation decisions, and managerial error (Francis, LaFond, Olsson, & Schipper, 2005). Discretionary accruals largely contain two subcomponents: performance and opportunism components. As a result, the CHINA JOURNAL OF ACCOUNTING STUDIES 159 discretionary accruals quality reflects a mixture of information-risk-increasing and informa- tion-risk-decreasing eec ff ts (Francis et al., 2005 ). Francis et al. (2005) find that innate accruals quality has a larger effect on cost of capital than discretionary accruals quality. NSOEs have a higher level of expropriation by the largest shareholder, thus have a higher level of fundamental risk. The market pays more attention to the business model and oper- ating environment of NSOEs. As a result, the market pricing effects for innate accruals quality is higher for NSOEs than for SOEs. Similarly, the market pays more attention to the discre- tionary accruals quality caused by accounting choices, implementation decisions, and man- agerial error of NSOEs, and gives a higher market pricing. Because of the information-risk-increasing role of discretionary accruals, we argue that the market will pay more attention to the discretionary accruals quality of NSOEs and give a higher market pricing. According to the above, the market pricing effects for both innate and discretionary accruals quality are lower for SOEs than for NSOEs. Thus, following Hypothesis 1, we posit our second hypothesis as: H2: The market pricing effects of both innate accruals quality and discretionary accruals quality are lower for SOEs than for NSOEs. 3. Methodology and sample selection 3.1. Measures of accruals quality and market pricing of accruals quality The empirical analysis requires a metric of accruals quality and its partition into innate and discretionary components. We adopt the approach developed in Dechow and Dichev (2002) to capture the precision of financial statement information. We follow Francis et al. (2005) by adopting McNichols’ (2002) modification of the Dechow and Dichev (2002) model. In the Dechow and Dichev (2002) model, accruals quality is measured by the extent to which working capital accruals map onto the past, current, and future operating cash flows, con- trolling for changes in revenue and the level of gross property, plant and equipment. Specifically, we measure accruals quality as the time-series standard deviation of the resid- uals, (v) calculated over years t–2 to t using the following equation: j,t TCA =  +  CFO +  CFO +  CFO +  ΔRev j,t 0,j 1,j j,t−1 2,j j,t 3,j j,t+1 4,j j,t (1) +  PPE + v 5,j j,t j,t where TCA is the total current accruals measured as income before depreciation and amor- tisation minus operating cash flow, CFO is the cash flow from operations, ΔRev is the change in revenue, and PPE is the level of property, plant, and equipment, all scaled by average total assets. Equation (1) is estimated using an industry with at least 10 firms in year t. Because approximately half of Chinese listed companies are in the manufacturing industry, we use a two-digit code (according to the Chinese Listed Company Classification) to classify man- ufacturing companies, while we use a one-digit code to classify firms in other industries. From the definition of accruals quality, we can see that a high accruals quality score reflects poor accruals quality and vice versa. f rancis et al. (2005) calculate the a Q measure using firm-specific residuals in year t –4 to year t. Because of the data restriction, we use three annual residuals to calculate a Q. 160 S. WU ET AL. Dechow and Dichev (2002) identify five innate factors associated with accruals quality: firm size, standard deviation of cash flow from operations, standard deviation of sales reve - nues, length of operating cycle, and incidence of negative earnings realisations. They suggest that these five innate variables capture economic fundamentals, as opposed to managerial discretion, which drive accruals quality. Following Dechow and Dichev (2002) and Francis et al. (2005), we expect smaller firms and those with greater cash flow volatility, longer operating cycles, and a greater incidence of losses to have poorer accruals quality. Then, the innate and discretionary components of accruals quality can be calculated from annual cross-sectional estimations of the following equation: AQ =  +  Size +  (CFO) +  (Sale) +  log(Opercycle) j,t 0 1 j,t 2 j,t 3 j,t 4 j,t (2) +  NegEarn + 5 j,t j,t where AQ is accruals quality, firm size (Size) is measured as the natural log of total assets, the standard deviation of cash flow from operations (σ(CFO )) is measured over the previous five years, the standard deviation of sales revenues (σ(Sales)) is measured over the previous five years, the length of the operating cycle (Opercycle) is measured as the log of the sum of days accounts receivable and days inventory, and the incidence of negative earnings realisation (NegEarn) is measured by the number of years out of the last five with negative reported net income. The predicted values from Equation (2) constitute the estimate of the innate com- ponent of firm j’s accruals quality (IAQ ) in year t, IAQ =  +  Size +  (CFO) +  (Sale) +  log(Opercycle) j,t 0 1 j,t 2 j,t 3 j,t 4 j,t +  NegEarn 5 j,t The residual from Equation (2) is the estimate of the discretionary component of firm j’s accruals quality (DAQ) in year t, DAQ = j,t j,t As for the market pricing of accruals quality, most of the extant research uses the ex post return in the stock market (Ecker, Francis, Kim, & Olsson, 2006; Francis et al., 2005). That is, it uses a capital asset pricing model, such as the Capital Asset Pricing Model (CAPM) or Fama- French three-factor model, and realised yearly (monthly) average asset return to estimate the future return. The literature shows that both of the foregoing models hold in the Chinese stock market (e.g. Drew, Naughton, & Veeraraghavan, 2003; Sun & Tong, 2000). Following Francis et al. (2005), we use both the CAPM and the Fama-French three-factor model, and construct accruals quality risk factors, AQfactor, IAQfactor and DAQfactor, which we include in the asset pricing model as additional factors to investigate whether accruals quality is priced in the Chinese market. 3.2. The model Francis et al. (2005) use augmented versions of the standard one-factor and three-factor models to examine the market pricing of accruals quality. Following Francis et al. (2005), we form an AQ (IAQ, DAQ) factor-mimicking portfolio that is equal to the difference between the monthly excess returns of the best two AQ (IAQ, DAQ) and the worst two AQ (IAQ, DAQ) CHINA JOURNAL OF ACCOUNTING STUDIES 161 quintiles, which is similar to that used by Fama and French (1993) to construct size and book-to-market factor-mimicking portfolios. The following firm-specific asset-pricing models are estimated: R − R =  +  (R − R )+ e AQfactor + (3) j,m F ,m j j M,m F ,m j m j,m R − R =  +  (R − R )+ s SMB + h HML + e AQfactor + (4) j,m F ,m j j M,m F ,m j m j m j m j,m R − R =  +  (R − R )+ e IAQfactor + e DAQfactor + (5) j,m F,m j j M,m F,m j,1 m j,2 m j,m R − R =  +  (R − R )+ s SMB + h HML + e IAQfactor j,m F,m j j M,m F,m j m j m j,1 m (6) + e DAQfactor + j,2 m j,m where R is firm j ’s stock return in the mth month, R is the risk-free return in the same time j,m F,m period, R is the value-weighted market index return in the mth month, SMB is the size M,m m factor mimic portfolio return, and HML is the value factor mimic portfolio return in the mth month. Both the size and the value factor portfolio returns are constructed following Fama and French (1992). AQfactor , IAQfactor , and DAQfactor are respectively the accruals qual- m m m ity, innate accruals quality, and discretionary accruals quality factors in the mth month, the construction of which is discussed in Section 3.1. In Equations (3) and (4), parameter e cap- tures the market pricing of accruals quality for firm j. Similarly, parameters e and e in j1 j2 Equations (5) and (6) respectively capture the market pricing of innate and discretionary accruals quality for firm j. We run regression models (3) to (6) for each firm using monthly data, where a firm’s number of monthly return observations is at least 18. After getting the parameter estimates for each firm, we use the approach of Fama and MacBeth (1973) to aggregate them to report the average parameter estimates. We also divide the whole sample into two subsamples, SOEs and NSOEs, and report the average parameter estimates for each subsample. According to Hypotheses 1 and 2, the market pricing effects of accruals (innate, discretionary) quality are larger for NSOEs than for SOEs. We should find that e , e and e are significantly larger j j1 j2 for NSOEs than for SOEs. 3.3. Data and sample We divide the listed companies into two categories: SOEs, the ultimate owner of which is the local government (any department in the local government, such as the Bureau of State Assets Management or Finance Bureau) or the central government (any central government unit, such as the Ministry of Finance or Central Industrial Enterprises Administration Committee); and NSOEs, which are owned by non-government units (individuals, townships and villages, or foreign companies). Information on the ownership type of each firm is obtained from firm annual reports, for which mandatory disclosure of the ultimate owner has been required since 2001. Because new accounting standards were implemented in China since 2007, our financial data end at 2006. f or example, for the months from May 2002 to april 2003, firms are ranked into quintiles based on the a Q signals calculated using annual data for the fiscal year ends between d ecember 2001 and april 2002. 162 S. WU ET AL. The sample studied in this paper was collected from the SinoFin database, which contains the accounting and financial information of all Chinese listed firms. Our original sample period is 1997–2008, but since we need to use the previous five years’ observations to cal- culate AQfactor, and the future 18 months’ observations to calculate return, our final sample for regression analyses is from 2001 to 2006. We have 7,687 firm-year observations from 2001 to 2006 for non-financial companies. To be included in any of the market-based tests, each firm-year observation must have data on AQ and the necessary market measures. We delete 2,797 firm-years without data of five-consecutive-year cash flows, 18 firm-years with - out data of receivables, and three firm-years without days sales of inventory. To construct AQfactor, IAQfactor and DAQfactor, which are used in the asset-pricing regressions (Models 3 through 6), we also require that firms have data on at least 18 monthly returns. The return accumulation period begins on the next 1 May to ensure the complete dissemination of the accounting information in the financial statements of the previous fiscal year. We also delete 418 firm-year observations without data of 18 future month returns. Finally, we have a sample of 973 firms, which includes 4,451 firm-year observations. Among the final sample, 3,311 are for SOEs, which account for 74% of all observations. Variable definitions are provided in Table 1, and Table 2 reports the sample distributions. There are 449 observations in 2001, whereas there are 919 observations in 2006. This is consistent with the fact that the size of the stock market is increasing in China. There are 695 observations in machinery, equipment, and instrument manufacturing, whereas there are only 28 observations in the mining industry. 4. Empirical results Table 3 provides the descriptive statistics. Panel A reports those for the full sample, and Panels B and C report those for the SOE and NSOE subsamples, respectively. Panel D reports the results of tests for differences in the characteristics of the SOE and NSOE subsamples. The average firm size (Size) is 21.27, and the maximum and minimum are 24.49 and 19.08, respectively, which shows that firm size varies greatly. The mean of the standard deviation of operating cash flow (σ(CFO )) is 0.06 and that of sales (σ(Sales)) is 0.15, which shows that sales are more volatile than cash flows. The average of NegEarn is 0.11, which means that 11% of firms reported a loss in the last five years. The minimum of MB (market value to book value) is less than 1 (0.67) and its maximum is much greater than 1 (49.13), which shows that the market value of some firms is much higher than their book value, while that of other firms is even lower than their book value. The summary statistics for the subsamples show that compared with NSOEs, SOEs are larger (Size) and have a more stable operating cash o fl w ( σ(CFO)), a shorter operating cycle (log(Opercycle)), lower incidence of negative earnings realisation (NegEarn), and lower level of growth (MB). Table 4 reports the estimation of AQ. Panel A presents the regression results for 2006. The estimated coefficients for σ(CFO), σ(Sales) and NegEarn are significantly positive, and the estimated coefficient of Size is significantly negative. The results are similar to those of Dechow and Dichev (2002) and Francis et al. (2005). The explanatory power of the summary indicators of the innate factors is 25.7%. The regression results for the years 2001 through 2005 are quantitatively similar to those for 2006. CHINA JOURNAL OF ACCOUNTING STUDIES 163 Table 1. Variable definition. Variable Definition Size natural logarithm of total assets σ(CFO) t he standard deviation of operation cash flows from operations over the past 5 years σ(Sales) t he standard deviation of operation sales from operations over the past 5 years l og(Opercycle) natural logarithm of operating cycle (the sum of days accounts receivable and days inventory) NegEarn t he incidence of negative earnings over the past 5 years Beta one-year beta estimated from firm-specific C aPM estimation MB t he market value to the book value AQ t he measure of accruals quality which is defined as standard deviation of residuals from years t–2 to t by annual cross-sectional estimations of the modified d echow and dichev (2002) model AQ1 t he measure of accruals quality which is defined as the absolute value of abnormal accruals generated by the modified Jones (1991) approach. IAQ t he measure of innate accruals quality which is the predicted value obtained from the annual parameter estimates from f rancis et al. (2005) model DAQ t he measure of discretionary accruals quality which is the residual from f rancis et al. (2005) model SMB return to size factor-mimicking portfolio HML return to book-to-market factor-mimicking portfolio AQfactor return to the AQ factor-mimicking portfolio AQfactor1 return to the AQ1 factor-mimicking portfolio IAQfactor return to IAQ factor-mimicking portfolio DAQfactor return to DAQ factor-mimicking portfolio ORECTA o ther receivables scaled by total assets SOE an indicator variable that equals 1 if the firm is ultimately controlled by central or local governments, and 0 otherwise. Lev t otal liabilities divided by total assets. First t he percentage of shares held by the largest shareholder. Secfive t he sum of the percentage of shareholding held by the second to the fifth largest shareholders Inddir t he percentage of the number of independent directors on the board. ROA return on total assets L_ORECTA t he lag of ORECTA FRisk t he first principal component of Turnover, ReturnVol, FirmAge and MktValue Turnover t he average daily trading volume as a percentage of shares outstanding over the last six months ReturnVol t he standard deviation of daily returns for the last 25 business days FirmAge t he number of months since the initial Sinof in monthly return for the firm MktValue t he total market value of shares outstanding as of the most recent d ecember 31 Table 2. Sample distributions. Panel A: Distribution across years Year 2001 2002 2003 2004 2005 2006 t otal nSoe 90 153 181 225 226 265 1,140 Soe 359 492 564 590 652 654 3,311 t otal 449 645 745 815 878 919 4,451 Panel B: Distribution across industries industry n o. industry n o. f arming, forestry, animal husbandry, and fishing 83 utilities 198 Mining 28 Construction 67 f ood & beverage 204 t ransportation and Warehousing 154 t extile, apparel, fur and l eather 173 information t echnology 261 Paper and allied Products; Printing 64 Wholesale and retail t rades 417 Petroleum, Chemical, Plastics, and rubber Products 514 real estate 213 Manufacturing electronics 130 Public f acilities and o ther Services 157 Metal, non-metal 376 Communication and Cultural industries 37 Machinery, equipment, and instrument Manufacturing 695 Conglomerates 382 Medicine and Biological Products 268 t otal 4,451 o ther Manufacturing 30 164 S. WU ET AL. Table 3. d escriptive statistics. Mean Standard deviation Median Minimum Maximum Panel A: Full sample (n = 4,451) Size 21.27 0.94 21.22 19.08 24.49 σ(CFO) 0.06 0.05 0.05 0.01 0.27 σ(Sales) 0.15 0.15 0.11 0.01 0.96 l og(Opercycle) 5.10 1.01 5.06 2.42 8.15 NegEarn 0.11 0.17 0.00 0.00 0.60 Beta 1.10 0.31 1.12 0.01 1.87 MB 3.51 3.90 2.42 0.67 49.13 Panel B: SOE sample (n = 3,311) Size 21.38 0.94 21.35 19.08 24.49 σ(CFO) 0.06 0.04 0.05 0.01 0.27 σ(Sales) 0.15 0.15 0.11 0.01 0.96 log Opercycle) 5.01 0.98 4.97 2.42 8.15 NegEarn 0.10 0.16 0.00 0.00 0.60 Beta 1.10 0.30 1.11 0.06 1.87 MB 3.31 3.71 2.36 0.67 49.13 Panel C: NSOE sample (n = 1,140) Size 20.96 0.86 20.94 19.08 24.02 σ(CFO) 0.07 0.05 0.06 0.01 0.27 σ(Sales) 0.15 0.14 0.10 0.01 0.96 l og(Opercycle) 5.35 1.06 5.31 2.42 8.15 NegEarn 0.15 0.19 0.00 0.00 0.60 Beta 1.10 0.35 1.14 0.01 1.87 MB 4.09 4.37 2.61 0.67 49.13 Panel D: SOE sample-NSOE sample Mean difference t-test Median difference Wilcoxon test Size 0.42 13.78*** 0.41 12.36*** σ(CFO) –0.01 –6.30*** –0.01 –7.45*** σ(Sales) 0.01 1.24 0.01 0.70 l og(Opercycle) –0.34 –9.56*** –0.35 –10.07*** NegEarn –0.05 –7.58*** –0.00 –7.91*** Beta –0.00 –0.06 –0.03 –0.04 MB –0.78 –5.39*** –0.26 –5.46*** *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. CHINA JOURNAL OF ACCOUNTING STUDIES 165 Table 4. estimations and descriptive statistics of accruals quality. Panel A: Regression result of AQ on innate factors (2006) Predicted sign estimate t value intercept 0.081 3.71*** Size – –0.003 –2.91*** σ(CFO) + 0.119 5.63*** σ(Sales) + 0.014 2.18** l og(Opercycle) + –0.001 –1.11 NegEarn + 0.074 13.18*** a dj. R 0.257 Panel B: Descriptive statistics Mean Standard deviation Median Minimum Maximum (1) f ull sample AQ 0.03 0.03 0.02 0.00 0.19 IAQ 0.03 0.02 0.03 0.01 0.10 DAQ 0.00 0.03 –0.00 –0.07 0.14 (2) Soe sample AQ 0.03 0.03 0.02 0.00 0.19 IAQ 0.03 0.02 0.03 0.01 0.10 DAQ 0.00 0.03 –0.00 –0.07 0.14 (3) nSoe sample AQ 0.04 0.04 0.03 0.00 0.19 IAQ 0.04 0.02 0.03 0.01 0.10 DAQ –0.00 0.03 –0.00 –0.07 0.14 note: t he dependent variable of Panel a is AQ . *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. Based on these parameters, we report the AQ decompositions in Panel B. For the full sample, the mean and median of AQ are 0.03 and 0.02, respectively. The minimum AQ is 0.00, while the maximum AQ is 0.19, which shows that there is a big difference in AQ among our sample. A big difference is also found between IAQ and DAQ. The minimums of IAQ and DAQ are 0.01 and –0.07, respectively, and the corresponding maximums are 0.10 and 0.14, respec- tively. The mean of DAQ is 0, because DAQ is the residual from Equation (2). The average AQ and IAQ for NSOEs are 0.04 and 0.04, respectively, and both are slightly larger than those for SOEs (0.03 and 0.03). Table 5 reports the mean results of firm-specific market-pricing regressions using the CAPM. Panel A reports the average coefficient estimates across the 973 firm-specific estima - tions. Panels B and C report the average coefficient estimates across the 667 SOEs and 306 NSOEs, respectively. Panel D reports the differences in the mean results between the SOE and NSOE subsamples, respectively. Model 1 is the benchmark model. Model 2 is for the market pricing of accruals quality and Model 3 is for that of innate accruals quality and discretionary accruals quality. For the full sample, the adjusted R increases from 0.546 to 0.556 when AQfactor is added (Models 1 and 2 in Panel A). The mean coefficient for AQfactor is 0.083 and is significant at the 1% level, which suggests that accruals quality has been significantly priced (Model 2 in Panel A). When we break down the sample into the SOE and NSOE subsamples, we find that the mean coefficient for AQfactor is –0.050 and insignificant for the former companies (Model 2 in Panel B), while it is 0.374 and significant at the 1% level for the latter companies (Model 2 in Panel C). The statistical results in Panel D show that Panels a, B and C in t ables 4 to 7 report the mean values of firm-specific asset-pricing regressions using the approach of f ama and MacBeth (1973). t herefore we present only the results of mean test and median test in Panel d . 166 S. WU ET AL. Table 5. Mean results of firm-specific market-pricing regressions: C aPM model. Model 1 Model 2 Model 3 R –R R –R R –R j F j F j F Estimate t value Estimate t value Estimate t value Panel A: Full sample R –R 0.963 136.76*** 0.955 145.51*** 0.933 140.91*** M F AQfactor 0.083 4.31*** IAQfactor 0.228 5.91*** DAQfactor 0.087 1.86* a dj. R 0.546 0.556 0.582 Panel B: SOE sample R –R 0.945 119.16*** 0.941 119.94*** 0.934 131.02*** M F AQfactor –0.050 –1.41 IAQfactor 0.046 1.15 DAQfactor –0.019 –0.39 a dj. R 0.522 0.535 0.565 Panel C: NSOE sample R –R 1.002 71.63*** 0.985 73.59*** 0.931 65.43*** M F AQfactor 0.374 5.40*** IAQfactor 0.624 7.64*** DAQfactor 0.319 3.19*** a dj. R 0.487 0.498 0.534 Panel D: SOE sample-NSOE sample Mean diff t value Mean diff t value AQfactor –0.424 –5.45*** IAQfactor –0.578 –6.36*** DAQfactor –0.338 –2.98*** notes: t he dependent variable is R –R , which is excess return on risk-free return. R –R is excess return on the market portfolio. j F M F *** ** * Significant at the 1% level; significant at the 5% level; significant at the 10% level. NSOEs have a significantly larger mean coefficient for the AQfactor than SOEs. The results support Hypothesis 1, i.e. the market pricing effect of accruals quality is greater for NSOEs than for SOEs. When we divide AQ into IAQ and DAQ, we find that for the full sample, the mean coefficient for IAQfactor is 0.228 and significant at the 1% level, whereas that for DAQfactor is just 0.087 and significant at the 10% level (Model 3 in Panel A). However, neither of the coefficient means of IAQfactor and DAQfactor is significant for the SOE sample (Model 3 in Panel B). In contrast, both of the coefficient means of IAQfactor and DAQfactor for NSOEs are significant at the 1% level: the mean coefficient for IAQfactor is 0.624 and that for DAQfactor is 0.319 (Model 3 in Panel C). The loading of IAQfactor is significantly greater than that of DAQfactor , which suggests that innate accruals quality has a greater effect on market pricing than discretionary accruals quality among NSOEs. The coefficient for DAQfactor is significantly positive in NSOEs and suggests that the opportunistic role dominates the informational role of discretionary accruals in NSOEs. The statistical comparison in Panel D shows that the mean difference of the coefficients for IAQfactor is –0.578 and significant at the 1% level between the SOE and NSOE subsamples. In addition, the mean coefficient for DAQfactor for SOEs is significantly smaller than that for NSOEs. This is consistent with Hypothesis 2, which proposes that the market pricing effects of both innate and discretionary accruals quality are larger for NSOEs than for SOEs. We use the Fama–French three-factor model to investigate the market pricing of accruals quality. Table 6 reports the mean results of firm-specific market-pricing regressions. As in Table CHINA JOURNAL OF ACCOUNTING STUDIES 167 5, Panel A reports the average coefficient estimates across the 973 firm-specific estimations. Panels B and C report the average coefficient estimates across the 667 SOEs and 306 NSOEs, respectively. Model 1 is the benchmark model. The adjusted R increases from 0.577 to 0.582 when AQfactor is considered (Models 1 and 2 in Panel A). The coefficient means for AQfactor are –0.064 and 0.208 for the SOE and the NSOE subsample, respectively (Panels B and C, Model 2). The statistical result in Model 2 of Panel D shows that NSOEs have a significantly larger mean than SOEs. The coefficient means for IAQfactor and DAQfactor are respectively 0.085 and –0.023 for the SOE subsample, and they are respectively 0.497 and 0.253 for the NSOE subsample. A comparison of the results (Model 3 in Panel D) shows that between the SOE and NSOE subsamples, the mean differences of the coefficients for IAQfactor and DAQfactor are –0.412 and –0.275, respectively, both of which are significant at the 1% level. The results show larger loadings for both IAQfactor and DAQfactor for NSOEs than for SOEs. Hence, the results in Table 6 are qualitatively similar to those in Table 5 and support our two hypotheses. To provide some perspective on economic significance, we calculate the standard devi- ations of AQfactor, IAQfactor, DAQfactor, R – R , SMB, HML for NSOEs, which are 0.03, 0.04, j F Table 6. Mean results of firm-specific market-pricing regressions: f ama–f rench model. Model 1 Model 2 Model 3 R –R R –R R –R j F j F j F Estimate t value Estimate t value Estimate t value Panel A: Full sample R –R 0.947 147.47*** 0.943 147.57*** 0.931 142.21*** M F SMB 0.100 5.70*** 0.080 4.55*** 0.005 0.23 HML –0.070 –3.06*** –0.073 –3.27*** –0.022 –0.99 AQfactor 0.021 0.65 IAQfactor 0.214 4.80*** DAQfactor 0.064 1.49 a dj. R 0.577 0.582 0.593 Panel B: SOE sample R –R 0.941 128.92*** 0.935 129.34*** 0.932 126.42*** M F SMB 0.020 0.97 0.015 0.70 -0.032 –1.16 HML –0.017 –0.65 –0.030 –1.18 -0.004 –0.16 AQfactor –0.064 –1.69 IAQfactor 0.085 1.71* DAQfactor –0.023 –0.47 a dj. R 0.560 0.557 0.562 Panel C: NSOE sample R –R 0.959 75.03*** 0.953 74.69*** 0.928 70.08*** M F SMB 0.277 8.42*** 0.223 7.16*** 0.087 1.98** HML –0.185 –4.17*** –0.166 –3.87*** –0.059 –1.43 AQfactor 0.208 3.06*** IAQfactor 0.497 5.50*** DAQfactor 0.253 2.95*** a dj. R 0.529 0.573 0.579 Panel D: SOE sample-NSOE sample Mean diff t value Mean diff t value AQfactor –0.272 –3.54*** IAQfactor –0.412 –4.00*** DAQfactor –0.275 –2.80*** notes: t he dependent variable is R –R , which is excess return on risk-free return. R –R is excess return on the market j F M F portfolio. *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. 168 S. WU ET AL. 0.03, 0.18, 0.06, 0.04, respectively. According to Panel C of Table 6, a one standard deviation increase in AQfactor corresponds to a 3% increase in standard deviation of R –R j F (0.208 × 0.03/0.18 = 3%). Similarly, the ee ff cts are 12% and 4% respectively for IAQfactor and DAQfactor while the impact for SMB (HML) is 8% (4%). Relative to SMB and HML, the market pricing effect of accruals quality is economically significant for NSOEs. 5. Robustness checks We also use the modied fi Jones ( 1991) model to compute an alternative measure of accruals quality and use it as a robustness check. We estimate the following cross-sectional regression for each of the industry groups with at least 10 firms in year t. TAC ∕A =  [1∕A ]+  [ΔREV ∕A ]+  [PPE ∕A ]+ (7) j,t j,t−1 0 j,t−1 1 j,t j,t−1 2 j,t j,t−1 j,t The industry- and year-specific parameter estimates obtained from Equation (7) are used to estimate firm-specific normal accruals (NA) as a percentage of lagged total assets: NA =  [1∕A ]+  [(ΔREV −ΔAR )∕A ]+  [PPE ∕A ] j,t 0 j,t−1 1 j,t j,t j,t−1 2 j,t j,t−1 where ΔAR is firm j’s change in accounts receivable between year t–1 and year t, and j,t AA = TAC /A − NA is the abnormal accruals (AA) in year t. The absolute value AQ1 = |AA | j,t j,t j,t-1 j,t j,t j,t is our alternative proxy for accruals quality, with larger values of AQ1 indicating poorer j,t accruals quality. The results in Tables 7 and 8 show that all the estimated coefficients on AQfactor1 for SOEs are insignificant while those for NSOEs are significantly positive. The coefficients for SOEs are much smaller than those for NSOEs. Overall, using an alternative measure of accruals quality does not change the results. Table 7. an alternative measure of accruals quality: C aPM model. Model 1 Model 2 R –R R –R j F j F Estimate t value Estimate t value Panel A: Full sample R –R 0.932 92.35*** 0.920 90.44*** M F AQfactor1 0.190 4.71*** a dj. R 0.456 0.468 Panel B: SOE sample R –R 0.931 80.05*** 0.930 79.62*** M F AQfactor1 0.042 1.13 a dj. R 0.462 0.472 Panel C: NSOE sample R –R 0.994 60.80*** 0.959 57.02*** M F AQfactor1 0.447 6.07*** a dj. R 0.455 0.468 Panel D: SOE sample-NSOE sample Mean diff t value AQfactor1 –0.404 –4.90*** notes: t he dependent variable is R –R , which is excess return on risk-free return. R –R is excess return on the market j F M F portfolio. *** ** * Significant at the 1% level; significant at the 5% level; significant at the 10% level. CHINA JOURNAL OF ACCOUNTING STUDIES 169 Table 8. an alternative measure of accruals quality: f ama–f rench model. Model 1 Model 2 R –R R –R j F j F Estimate t value Estimate t value Panel A: Full sample R –R 0.888 114.62*** 0.885 108.61*** M F SMB 0.125 6.83*** 0.120 6.28*** HML –0.096 –5.03*** –0.087 –4.52*** AQfactor1 0.051 1.24 a dj. R 0.496 0.505 Panel B: SOE sample R –R 0.916 101.00*** 0.916 98.64*** M F SMB 0.059 3.46*** 0.059 3.10*** HML –0.043 –2.75*** –0.043 –2.79*** AQfactor1 –0.006 –0.13 a dj. R 0.499 0.507 Panel C: NSOE sample R –R 0.887 63.42*** 0.884 59.67*** M F SMB 0.299 9.57*** 0.270 8.77*** HML –0.201 –5.81*** –0.201 –5.87*** AQfactor1 0.142 2.08** a dj. r 0.507 0.511 Panel D: SOE sample-NSOE sample Mean diff t value AQfactor1 –0.147 –1.84* note: t he dependent variable is rj–rf , which is excess return on risk-free return. rM–rf is excess return on the market portfolio. *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. 6. Further analysis We propose in our hypothesis development that the largest shareholders of SOEs have less incentive to expropriate firm resources, so SOEs have a lower level of fundamental risk than NSOEs, which leads to different market pricing. We test in this part whether this mechanism exists. The expropriation by the largest shareholder is just one of the factors that may affect fundamental risk. By analysing the expropriation, we aim to prove that the fundamental risk of NSOEs is different from that of SOEs. We have stated that the market pricing effect is higher in NSOEs than in SOEs. This is due to the fact that NSOEs have a higher level of expropriation by the largest shareholder, and thus have a higher level of fundamental risk. Although it is a common sense, we provide direct evidence here. First, we examine the impact of state ownership on expropriation by the largest shareholder. Following Jiang et al. (2010), we measure expropriation using ORECTA (Other Receivable scaled by Total Assets). SOE is an indicator variable that equals 1 if the client is ultimately controlled by central or local governments, and 0 otherwise. Size is the natural logarithm of total assets. Lev is the total liabilities divided by total assets. First is the percentage of shares held by the largest shareholder. Secfive is the sum of the percentage of shareholding held by the second to the fifth largest shareholders. Inddir is the percentage of the number of independent directors on the board. ROA is the return on total assets. L_ORECTA is the lag of ORECTA. We regress ORECTA on SOE and Table 9 presents the regression result. We find that the coefficient on 170 S. WU ET AL. Table 9. t he effects of Soe on expropriation and fundamental risk. Model 1 Model 2 Model 3 ORECTA FRisk FRisk Estimate t value Estimate t value Estimate t value ORECTA 0.190 3.11*** SOE –0.012 –6.24*** –0.032 –2.91*** Size 0.163 4.20*** –0.035 –6.43*** –0.031 –5.62*** Lev 0.006 1.30 0.131 4.56*** 0.118 4.10*** First 0.566 4.52*** –0.307 –11.57*** –0.282 –11.04*** Secfive 13.769 3.36*** 0.077 11.58*** 0.071 11.05*** Inddir –0.017 –1.98** 0.093 2.26** 0.094 2.28** ROA –0.250 –17.33*** –0.385 –4.36*** –0.323 –3.54*** L_ORECTA 0.638 60.23*** σ(CFO) –0.065 –0.61 –0.112 –1.06 σ(Sales) 0.180 5.41*** 0.187 5.63*** l og(Opercycle) 0.030 6.01*** 0.028 5.57*** NegEarn 0.195 6.01*** 0.186 5.75*** intercept –3.163 –4.17*** 1.084 9.08*** 0.980 8.18*** N 4,450 4,440 4,440 a dj. R 0.569 0.127 0.127 *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. SOE is significantly negative (–0.012, t = –6.24, Model 1), which suggests a higher level of expropriation by the largest shareholder in NSOEs than in SOEs. Then we examine the impact of state ownership on fundamental risk. Chen et al. (2008) measure fundamental risk using market capitalisation, firm age, return volatility and trading volume. As in Chen et al. (2008), we combine the four proxies into a summary fundamental risk measure (FRisk) using principal component analysis. We regress FRisk on SOE and Table 9 presents the regression result. We can find that the coefficient on SOE is significantly neg- ative (–0.032, t = –2.91, Model 2), which suggests that the fundamental risk is lower in SOEs than that in NSOEs. Finally, we examine the effect of expropriation on fundamental risk. Similarly, we regress FRisk on ORECTA and Table 9 presents the regression result. We can see that the coefficient on ORECTA is significant positive (0.190, t = 3.11, Model 3), which is consistent with the fundamental risk being lower in low expropriation firms than in high expropriation firms. 7. Conclusions Existing literature has studied the role of state ownership in investor protection from the perspective of the effect of state ownership on accruals quality but ignores the effect of state ownership on the market pricing of accruals quality. Our paper investigates this problem using data of Chinese listed firms. We find that the market pricing effect of accruals quality is larger for NSOEs than for SOEs. The results hold when we divide accruals quality into innate and discretionary accruals quality, which suggests that state ownership improves the level of investor protection. Our study adds to the growing literature on the governance role of state ownership and the market pricing of accruals quality. The findings of this paper also have important practical implications. The results of our study again show that SOEs provide better investor protection than NSOEs. Therefore, supervisory departments should pay closer attention to the investor protection problems in NSOEs. Meanwhile, as the market is more concerned with the CHINA JOURNAL OF ACCOUNTING STUDIES 171 accounting information of NSOEs, supervisory departments should care more about the quality of accounting information of NSOEs. Auditors should also devote more effort and maintain a higher level of professional scepticism when auditing NSOEs. Disclosure statement No potential conflict of interest was reported by the authors. Funding Xianhui Bo acknowledges financial support from the National Natural Science Foundation of China [grant number 71302130]. References Aharony, J., Lee, C.-W.J., & Wong, T.J. (2000). Financial packaging of IPO firms in China. Journal of Accounting Research, 38(1), 103–126. Bai, Y., Lin, B., Wang, Y., & Wu, L. (2013). Full privatization, expropriation, and firm performance: Evidence from China. Applied Economics, 45, 1857–1867. Bo, X., & Wu, L. (2009). The governance roles of state controlling and institution investors: A perspective of earnings management. Economic Research Journal, 2, 81–91 (In Chinese). Brandt, L., & Li, H. (2003). Bank discrimination in transition economies: Ideology, information, or incentives? Journal of Comparative Economics, 31, 387–413. Calomiris, C., Fisman, R., & Wang, Y. (2010). Profiting from government stakes in a command economy: Evidence from Chinese asset sales. Journal of Financial Economics, 96, 399–412. Chen, H., Chen, J.Z., Lobo, G.J., & Wang, Y. (2010). Association between borrower and lender state ownership and accounting conservatism. Journal of Accounting Research, 48, 973–1014. Chen, L., Dhaliwal, D., & Trombley, M. (2008). The effect of fundamental risk on the market pricing of accruals quality. Journal of Accounting, Auditing & Finance, 23, 471–492. Claessens, S., Djankov, S., Fan, J.P.H., & Lang, L.H.P. (2002). Disentangling the incentive and entrenchment effects of large shareholdings. The Journal of Finance, 57, 2741–2771. Dechow, P., & Dichev, I. (2002). The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77(Supplement), 35–59. Deloitte Touche Tohmatsu. (2006). China’s new accounting standards: A comparison with current PRC GAAP and IFRS. http://www.iasplus.com/country/china.htm. Ding, Y., Zhang, H., & Zhang, J. (2007). Private vs. state ownership and earnings management: Evidence from Chinese listed companies. Corporate Governance: An International Review, 15, 224–239. Drew, M.E., Naughton, T., & Veeraraghavan, M. (2003). Firm size, book-to-market equity and security returns: Evidence from the Shanghai Stock Exchange. Australian Journal of Management, 28, 119–139. Ecker, F., Francis, J., Kim, I., Olsson, P.M., & Schipper, K. (2006). A returns-based representation of earnings quality. The Accounting Review, 81, 749–780. Fama, E.F., & French, K.R. (1992). The cross-section of expected stock returns. Journal of Finance, 47, 427–465. Fama, E.F., & French, K.R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56. Fama, E.F., & MacBeth, J. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81, 607–636. Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2005). The market pricing of accruals quality. Journal of Accounting and Economics, 39, 295–327. Guay, W.R., Kothari, S.P., & Watts, R.L. (1996). A market-based evaluation of discretionary accrual models. Journal of Accounting Research, 34 (Supplement), 83–105. 172 S. WU ET AL. Huang, F., Su, J., & Chong, T. (2008). Testing for structural change in the nontradable share reform of the Chinese stock market. The Chinese Economy, 41, 24–33. Jensen, M.C., & Meckling, M.H. (1976). Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3, 305–360. Jian, M., & Wong, T.J. (2010). Propping through related party transactions. Review of Accounting Studies, 15(1), 70–105. Jiang, G., Lee, C., & Yue, H. (2010). Tunneling through inter-corporate loans: The China experience. Journal of Financial Economics, 98(1), 1–20. Jones, J. (1991). Earnings management during import relief investigations. Journal of Accounting Research, 29, 193–228. Kornai, J. (1993). The evolution of financial discipline under the postsocialist system. Kyklos , 46, 315–336. Li, J., & Qian, C. (2013). Principal-principal conflicts under weak institutions: A study of corporate takeovers in China. Strategic Management Journal, 34, 498–508. Li, H., & Zhou, L.-A. (2005). Political turnover and economic performance: The incentive role of personnel control in China. Journal of Public Economics, 89, 1743–1762. Lin, Y., & Tan, G. (1999). Policy burdens, accountability, and the soft budget constraint. American Economic Review, 89, 426–431. McNichols, M.F. (2002). Discussion of the quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77 (Supplement), 61–69. Qian, Y. (2001). Government control in corporate governance as a transitional institution: lessons from China. In J. Stiglitz & S. Yusuf (Eds.), Rethinking the East Asian miracle (pp. 295–321). New York, NY: Oxford University Press and the World Bank. Shen, H., & Wu, L. (2012). Corporate ownership, environment uncertainty and the governance role of accounting information. Accounting Research, 8, 8–16. (In Chinese). Subramanyam, K.R. (1996). The pricing of discretionary accruals. Journal of Accounting and Economics, 22, 249–281. Sun, Q., & Tong, W.H.S. (2000). The effect of market segmentation on stock prices: The China syndrome. Journal of Banking and Finance, 24, 1875–1902. Wang, Q., Wong, T.J., & Xia, L. (2008). State ownership, the institutional environment, and auditor choice: Evidence from China. Journal of Accounting and Economics, 46, 112–134. Yee, K.K. (2006). Earnings quality and the equity risk premium: A benchmark model. Contemporary Accounting Research, 23, 833–877. Zhang, H., & Li, G.J. (2011). Study on the efficient of Chinese stock market. Journal of Data Analysis, 6(1), 63–82. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png China Journal of Accounting Studies Taylor & Francis

State ownership and the market pricing of accruals quality

Download PDF
18 pages

Loading next page...
 
/lp/taylor-francis/state-ownership-and-the-market-pricing-of-accruals-quality-jWnuOqDBY5

References (36)

Publisher
Taylor & Francis
Copyright
© 2017 Accounting Society of China
ISSN
2169-7221
eISSN
2169-7213
DOI
10.1080/21697213.2017.1339432
Publisher site
See Article on Publisher Site

Abstract

China Journal of aCC ounting StudieS , 2017 Vol . 5, no . 2, 155–172 https://doi.org/10.1080/21697213.2017.1339432 a b b c Shinong Wu , Yaping Wang , Liansheng Wu and Xianhui Bo a b School of Management, Xiamen university, China; guanghua School of Management, Peking university, China; School of a ccountancy, Central university of f inance and economics, China ABSTRACT KEYWORDS a ccruals quality; investor This paper examines the effect of state ownership on the market protection; market pricing; pricing of accruals quality. We find that the market pricing effect state ownership of accruals quality is larger for non-state-owned enterprises than for state-owned enterprises in China. The results still hold when we divide accruals quality into innate and discretionary accruals quality. Further analyses reveal a higher level of expropriation and a greater fundamental risk in non-state-owned enterprises, leading to a larger market pricing effect of accruals quality. This paper is the first to study the governance role of state ownership from the perspective of market pricing of accruals quality. It also adds to the literature on the market pricing of accruals quality by studying its link with state ownership. 1. Introduction The function of accounting information in a capital market relies on two aspects: the account- ing information quality itself and the influence of the quality of accounting information on the capital market. Absence of either impairs the effectiveness of accounting information. Existing studies mainly focus on the quality of accounting information itself but ignore its effect on capital market. One of the important effects of accounting information quality on capital market is the market pricing of accruals quality. When we examine the market pricing of accruals quality, the governance role of state ownership should also be considered. The governance role of state ownership in China is so important that it has received a consider- able amount of attention. Since an agent’s infringement of a principal’s interest may arise where there is information asymmetry (Jensen & Meckling, 1976), accounting information has been a focus in studying the governance role of state ownership. Specifically, we can investigate the effects of state ownership on accruals quality and the market pricing of accruals quality. Extant literature explores the effect of state ownership on earnings quality and finds that the level of earnings management is lower in state-owned enterprises (SOEs) than in non-state-owned enterprises (NSOEs) (Bo & Wu, 2009; Ding, Zhang, & Zhang, 2007). Previous studies also reveal that the impact of accounting information in SOEs on debt covenants is weaker than that in NSOEs (Chen, Chen, Lobo, & Wang, 2010), and that the CONTACT Xianhui Bo 13466715176@163.com *Paper accepted by Jason Xiao. © 2017 a ccounting Society of China 156 S. WU ET AL. governance role of accounting information is also weaker in SOEs than in NSOEs (Shen & Wu, 2012). Yet no research focuses on how state ownership is associated with the capital market effects of accruals quality. Extant literature suggests that China’s stock market is becoming more and more efficient and it is changing from a weak efficient stock market to a semi-strong-form efficient stock market (Zhang & Li, 2011). So China’s stock market can be used to investigate how state ownership is associated with the capital market effects of accruals quality. In this paper, we examine the effect of state ownership on the market pricing of accruals quality using Chinese market data. Using the CAPM and Fama and French three-factor model augmented with an accrual quality risk factor, we find the market pricing effect of accruals quality is larger for NSOEs than for SOEs. The results hold when we divide accruals quality into innate and discretionary accruals quality. Further analyses reveal a higher level of expro- priation by the largest shareholder and a greater fundamental risk in NSOEs than in SOEs, leading to a larger market pricing effect of accruals quality, which means that poorer accruals quality is associated with higher costs of equity capital. To the best of our knowledge, this paper is the first to study the governance role of state ownership from the perspective of market pricing of accruals quality. The existing literature mainly focuses on the effect of state ownership on accruals quality and neglects its effect on the market pricing of accruals quality. In contrast, this paper investigates the different market effects of accruals quality under different types of corporate ownership. This paper enriches our knowledge of the governance role of state ownership. Our study also adds to the growing literature on the market pricing of accruals quality by studying its link with state ownership. Compared with a developed economy, China’s institutional and regulatory envi- ronment features a number of subtle differences, which potentially impact the market pricing of accruals quality. The result in our study shows that the market pricing effect of accruals quality critically depends on the corporate ownership structure, which extends the growing literature on the market pricing of accruals quality. The remainder of the paper proceeds as follows. Section 2 provides the institutional background and hypothesis development. Section 3 outlines the empirical methodology and sample selection. The empirical results are presented in Section 4. Section 5 reports the robustness checks. Section 6 provides further analysis. Section 7 concludes. 2. Institutional background and hypothesis development In an attempt to revitalise inefficient SOEs, the Chinese government partially privatised more than 1,000 SOEs through share issue privatisation in the course of its reform programme from 1990 (Jian & Wong, 2010). Partial privatisation freed about one-third of a company’s shares, while the remaining two-thirds were owned by the state or transferred to ‘legal persons’. State shares were held by government agencies such as the Bureau of State Property Administration (Bai, Lin, Wang, & Wu, 2013). Legal person shares were often held by repre- sentatives of the original SOEs, who retained close ties with local governments (Jiang, Lee, & Yue, 2010). The remaining shares were sold to mainland China investors (A shares), Hong Kong investors (H shares), and foreign investors (B shares). These listed firms are still con- trolled by the state, either directly or indirectly ( Wang, Wong, & Xia, 2008). Because we exam- ine only listed firms, we still refer to them as SOEs. CHINA JOURNAL OF ACCOUNTING STUDIES 157 In April 2005, the China Securities Regulatory Commission (CSRC) launched a state share reform, aiming at converting the non-tradable state-owned shares to tradable shares. As a result of the reform, the Chinese stock market improved dramatically (Huang, Su, & Chong, 2008). According to the report of the Central People’s Government of China in March 2007, 97% of Chinese firms had complied with the orders to reform their share structure. This figure represents 1,301 of the firms that were ordered to complete the reform by the end of 2006. Their aggregated market value was nearly 98% of the market capitalisation of the Shanghai and Shenzhen stock exchanges at that time. After the split-share structure reform in 2007, the institutional background in China became different. Meanwhile, beginning on 1 January 2007, all Chinese-listed companies were forced to implement the new set of accounting standards released by the Ministry of Finance in 2006. The introduction of the new account- ing standards has resulted in fundamental changes to financial reporting practices in China (Deloitte Touche Tohmatsu, 2006). In China’s unique institutional background, there is a significant difference between SOEs and NSOEs. SOEs receive political and financial support from the government, and govern- ment leaders have incentives to assist SOEs (Brandt & Li, 2003; Kornai, 1993; Li & Zhou, 2005). Similarly, the government, as the stock market regulator, also treats SOEs preferentially, providing listing privileges to them based on political rather than economic objectives (Aharony, Lee, & Wong, 2000). Jian and Wong (2010) show that related sales propping is more prevalent among state-owned firms than among non-state-owned firms when earnings are close to the securities regulators’ earnings targets. Since SOEs receive government support while NSOEs do not, the level of propping by the largest shareholder is higher in SOEs than in NSOEs. Because of the highly concentrated ownership structure of Chinese listed firms, the main agency problem is between controlling and minority shareholders (Jiang et al., 2010). In theory, a divergence between cash flow and control rights will motivate the largest share - holder to expropriate firm resources (Claessens, Djankov, Fan, & Lang, 2002). However, this incentive might differ between state and private owners. The former, who represent the largest shareholder in an SOE, might not have a strong incentive to exploit company wealth, although they may make the firm undertake certain social responsibilities, as they often have different goals from those of private block shareholders (Lin & Tan, 1999). The state shareholder will put more weight on the maximisation of social welfare than on the maxi- misation of the wealth of block shareholders (Bai et al., 2013). On the personal level, managers of SOEs are frequently reviewed by government agencies, and their political advancement might depend on their performance. For example, the State Council has an explicit policy guideline to remove managers from SOEs if they are responsible for losses over three con- secutive years. The potential loss of political reputation and forced demotion due to poor performance discourages management from aggressively exploiting the company and expropriating minority shareholders (Li & Qian, 2013). However, the objective of the largest shareholder in the NSOEs is the maximisation of their own interests. Thus, they are very likely to exploit the company because this could improve their welfare. Meanwhile, the manage- ment’s benefits of NSOEs are largely decided by the largest shareholder. As a result, they are inclined to assist the largest shareholder to exploit the company (Bai et al., 2013). Therefore, See http://www.gov.cn. See http://www.sasac.gov.cn. 158 S. WU ET AL. it is more likely that the largest shareholder in a private company will expropriate the resources or interests of minority shareholders in a firm more than those in an SOE. In sum, due to government support and their own incentives, the level of expropriation by the largest shareholder is lower in SOEs than in NSOEs. Extant literature finds supporting evidence. Qian (2001) finds that private controlling shareholders have greater incentives to expropriate company resources than state ones do. Calomiris, Fisman, and Wang (2010) find a negative effect of government ownership on returns at the announcement date and a symmetric positive effect from the policy’s cancellation. Jiang et al. (2010) show that tun- nelling through inter-corporate loans is worse in NSOEs than that in SOEs. Bai et al. (2013) find that fully privatised firms perform worse than state-controlled enterprises owing to excessive expropriation by controlling shareholders after privatisation. According to Yee (2006), reported earnings have two roles, first as a fundamental attrib - ute and second as a financial reporting attribute. Fundamental earnings is the accounting profitability measure that gauges a firm’s ability to make future dividend payments. In contrast, reported earnings is the imperfect signal of fundamental earnings that a firm announces. Earnings quality refers to how quickly and precisely reported earnings reveal fundamental earnings. The higher the quality of earnings, the more quickly and precisely reported earnings convey shocks to the present value of expected dividends (PVED). Corresponding to the two guises of reported earnings are two sources of earnings risk. The first source is fundamental risk. Fundamental risk is unresolved uncertainty in future divi- dend payments, which is not resolved even if one observes the exact (mean) value of PVED. The second risk source is earnings quality risk – additional uncertainty about the value of PVED caused by limited earnings quality. Earnings quality risk is information risk stemming from deficiencies in the accounting rules and firms’ application of those accounting rules in a financial reporting environment. In the absence of fundamental risk, earnings quality risk has no effect on the cost of capital, and that increasing fundamental risk serves to magnify the effect of earnings quality risk on the cost of capital ( Yee, 2006), thus increasing the market pricing effect of earnings quality. Chen, Dhaliwal, and Trombley (2008) find that there is essentially no relation between accruals quality and cost of capital as measured by future return realisations for firms with the lowest fundamental risk. In contrast, for firms with the highest fundamental risk, there is a strong relation between accruals quality and future return realisations. For the reason that the largest shareholders of SOEs have less incentive to expropriate r fi m resources, and the fundamental risk is lower in low expropriation firms than that in high expropriation firms, SOEs have a lower level of fundamental risk than NSOEs. As fundamental risk is positively related to the market pricing effect of accruals quality, the market pricing effect of accruals quality is lower for SOEs than for NSOEs. Therefore, we posit the following hypothesis: H1: The market pricing effect of accruals quality is lower for SOEs than for NSOEs. Guay, Kothari, and Watts (1996) and Subramanyam (1996) argue that financial reporting outcome can be partitioned into innate and discretionary components. Innate accruals qual- ity is driven by innate features of the firm’s business model and operating environment, and discretionary accruals quality is due to accounting choices, implementation decisions, and managerial error (Francis, LaFond, Olsson, & Schipper, 2005). Discretionary accruals largely contain two subcomponents: performance and opportunism components. As a result, the CHINA JOURNAL OF ACCOUNTING STUDIES 159 discretionary accruals quality reflects a mixture of information-risk-increasing and informa- tion-risk-decreasing eec ff ts (Francis et al., 2005 ). Francis et al. (2005) find that innate accruals quality has a larger effect on cost of capital than discretionary accruals quality. NSOEs have a higher level of expropriation by the largest shareholder, thus have a higher level of fundamental risk. The market pays more attention to the business model and oper- ating environment of NSOEs. As a result, the market pricing effects for innate accruals quality is higher for NSOEs than for SOEs. Similarly, the market pays more attention to the discre- tionary accruals quality caused by accounting choices, implementation decisions, and man- agerial error of NSOEs, and gives a higher market pricing. Because of the information-risk-increasing role of discretionary accruals, we argue that the market will pay more attention to the discretionary accruals quality of NSOEs and give a higher market pricing. According to the above, the market pricing effects for both innate and discretionary accruals quality are lower for SOEs than for NSOEs. Thus, following Hypothesis 1, we posit our second hypothesis as: H2: The market pricing effects of both innate accruals quality and discretionary accruals quality are lower for SOEs than for NSOEs. 3. Methodology and sample selection 3.1. Measures of accruals quality and market pricing of accruals quality The empirical analysis requires a metric of accruals quality and its partition into innate and discretionary components. We adopt the approach developed in Dechow and Dichev (2002) to capture the precision of financial statement information. We follow Francis et al. (2005) by adopting McNichols’ (2002) modification of the Dechow and Dichev (2002) model. In the Dechow and Dichev (2002) model, accruals quality is measured by the extent to which working capital accruals map onto the past, current, and future operating cash flows, con- trolling for changes in revenue and the level of gross property, plant and equipment. Specifically, we measure accruals quality as the time-series standard deviation of the resid- uals, (v) calculated over years t–2 to t using the following equation: j,t TCA =  +  CFO +  CFO +  CFO +  ΔRev j,t 0,j 1,j j,t−1 2,j j,t 3,j j,t+1 4,j j,t (1) +  PPE + v 5,j j,t j,t where TCA is the total current accruals measured as income before depreciation and amor- tisation minus operating cash flow, CFO is the cash flow from operations, ΔRev is the change in revenue, and PPE is the level of property, plant, and equipment, all scaled by average total assets. Equation (1) is estimated using an industry with at least 10 firms in year t. Because approximately half of Chinese listed companies are in the manufacturing industry, we use a two-digit code (according to the Chinese Listed Company Classification) to classify man- ufacturing companies, while we use a one-digit code to classify firms in other industries. From the definition of accruals quality, we can see that a high accruals quality score reflects poor accruals quality and vice versa. f rancis et al. (2005) calculate the a Q measure using firm-specific residuals in year t –4 to year t. Because of the data restriction, we use three annual residuals to calculate a Q. 160 S. WU ET AL. Dechow and Dichev (2002) identify five innate factors associated with accruals quality: firm size, standard deviation of cash flow from operations, standard deviation of sales reve - nues, length of operating cycle, and incidence of negative earnings realisations. They suggest that these five innate variables capture economic fundamentals, as opposed to managerial discretion, which drive accruals quality. Following Dechow and Dichev (2002) and Francis et al. (2005), we expect smaller firms and those with greater cash flow volatility, longer operating cycles, and a greater incidence of losses to have poorer accruals quality. Then, the innate and discretionary components of accruals quality can be calculated from annual cross-sectional estimations of the following equation: AQ =  +  Size +  (CFO) +  (Sale) +  log(Opercycle) j,t 0 1 j,t 2 j,t 3 j,t 4 j,t (2) +  NegEarn + 5 j,t j,t where AQ is accruals quality, firm size (Size) is measured as the natural log of total assets, the standard deviation of cash flow from operations (σ(CFO )) is measured over the previous five years, the standard deviation of sales revenues (σ(Sales)) is measured over the previous five years, the length of the operating cycle (Opercycle) is measured as the log of the sum of days accounts receivable and days inventory, and the incidence of negative earnings realisation (NegEarn) is measured by the number of years out of the last five with negative reported net income. The predicted values from Equation (2) constitute the estimate of the innate com- ponent of firm j’s accruals quality (IAQ ) in year t, IAQ =  +  Size +  (CFO) +  (Sale) +  log(Opercycle) j,t 0 1 j,t 2 j,t 3 j,t 4 j,t +  NegEarn 5 j,t The residual from Equation (2) is the estimate of the discretionary component of firm j’s accruals quality (DAQ) in year t, DAQ = j,t j,t As for the market pricing of accruals quality, most of the extant research uses the ex post return in the stock market (Ecker, Francis, Kim, & Olsson, 2006; Francis et al., 2005). That is, it uses a capital asset pricing model, such as the Capital Asset Pricing Model (CAPM) or Fama- French three-factor model, and realised yearly (monthly) average asset return to estimate the future return. The literature shows that both of the foregoing models hold in the Chinese stock market (e.g. Drew, Naughton, & Veeraraghavan, 2003; Sun & Tong, 2000). Following Francis et al. (2005), we use both the CAPM and the Fama-French three-factor model, and construct accruals quality risk factors, AQfactor, IAQfactor and DAQfactor, which we include in the asset pricing model as additional factors to investigate whether accruals quality is priced in the Chinese market. 3.2. The model Francis et al. (2005) use augmented versions of the standard one-factor and three-factor models to examine the market pricing of accruals quality. Following Francis et al. (2005), we form an AQ (IAQ, DAQ) factor-mimicking portfolio that is equal to the difference between the monthly excess returns of the best two AQ (IAQ, DAQ) and the worst two AQ (IAQ, DAQ) CHINA JOURNAL OF ACCOUNTING STUDIES 161 quintiles, which is similar to that used by Fama and French (1993) to construct size and book-to-market factor-mimicking portfolios. The following firm-specific asset-pricing models are estimated: R − R =  +  (R − R )+ e AQfactor + (3) j,m F ,m j j M,m F ,m j m j,m R − R =  +  (R − R )+ s SMB + h HML + e AQfactor + (4) j,m F ,m j j M,m F ,m j m j m j m j,m R − R =  +  (R − R )+ e IAQfactor + e DAQfactor + (5) j,m F,m j j M,m F,m j,1 m j,2 m j,m R − R =  +  (R − R )+ s SMB + h HML + e IAQfactor j,m F,m j j M,m F,m j m j m j,1 m (6) + e DAQfactor + j,2 m j,m where R is firm j ’s stock return in the mth month, R is the risk-free return in the same time j,m F,m period, R is the value-weighted market index return in the mth month, SMB is the size M,m m factor mimic portfolio return, and HML is the value factor mimic portfolio return in the mth month. Both the size and the value factor portfolio returns are constructed following Fama and French (1992). AQfactor , IAQfactor , and DAQfactor are respectively the accruals qual- m m m ity, innate accruals quality, and discretionary accruals quality factors in the mth month, the construction of which is discussed in Section 3.1. In Equations (3) and (4), parameter e cap- tures the market pricing of accruals quality for firm j. Similarly, parameters e and e in j1 j2 Equations (5) and (6) respectively capture the market pricing of innate and discretionary accruals quality for firm j. We run regression models (3) to (6) for each firm using monthly data, where a firm’s number of monthly return observations is at least 18. After getting the parameter estimates for each firm, we use the approach of Fama and MacBeth (1973) to aggregate them to report the average parameter estimates. We also divide the whole sample into two subsamples, SOEs and NSOEs, and report the average parameter estimates for each subsample. According to Hypotheses 1 and 2, the market pricing effects of accruals (innate, discretionary) quality are larger for NSOEs than for SOEs. We should find that e , e and e are significantly larger j j1 j2 for NSOEs than for SOEs. 3.3. Data and sample We divide the listed companies into two categories: SOEs, the ultimate owner of which is the local government (any department in the local government, such as the Bureau of State Assets Management or Finance Bureau) or the central government (any central government unit, such as the Ministry of Finance or Central Industrial Enterprises Administration Committee); and NSOEs, which are owned by non-government units (individuals, townships and villages, or foreign companies). Information on the ownership type of each firm is obtained from firm annual reports, for which mandatory disclosure of the ultimate owner has been required since 2001. Because new accounting standards were implemented in China since 2007, our financial data end at 2006. f or example, for the months from May 2002 to april 2003, firms are ranked into quintiles based on the a Q signals calculated using annual data for the fiscal year ends between d ecember 2001 and april 2002. 162 S. WU ET AL. The sample studied in this paper was collected from the SinoFin database, which contains the accounting and financial information of all Chinese listed firms. Our original sample period is 1997–2008, but since we need to use the previous five years’ observations to cal- culate AQfactor, and the future 18 months’ observations to calculate return, our final sample for regression analyses is from 2001 to 2006. We have 7,687 firm-year observations from 2001 to 2006 for non-financial companies. To be included in any of the market-based tests, each firm-year observation must have data on AQ and the necessary market measures. We delete 2,797 firm-years without data of five-consecutive-year cash flows, 18 firm-years with - out data of receivables, and three firm-years without days sales of inventory. To construct AQfactor, IAQfactor and DAQfactor, which are used in the asset-pricing regressions (Models 3 through 6), we also require that firms have data on at least 18 monthly returns. The return accumulation period begins on the next 1 May to ensure the complete dissemination of the accounting information in the financial statements of the previous fiscal year. We also delete 418 firm-year observations without data of 18 future month returns. Finally, we have a sample of 973 firms, which includes 4,451 firm-year observations. Among the final sample, 3,311 are for SOEs, which account for 74% of all observations. Variable definitions are provided in Table 1, and Table 2 reports the sample distributions. There are 449 observations in 2001, whereas there are 919 observations in 2006. This is consistent with the fact that the size of the stock market is increasing in China. There are 695 observations in machinery, equipment, and instrument manufacturing, whereas there are only 28 observations in the mining industry. 4. Empirical results Table 3 provides the descriptive statistics. Panel A reports those for the full sample, and Panels B and C report those for the SOE and NSOE subsamples, respectively. Panel D reports the results of tests for differences in the characteristics of the SOE and NSOE subsamples. The average firm size (Size) is 21.27, and the maximum and minimum are 24.49 and 19.08, respectively, which shows that firm size varies greatly. The mean of the standard deviation of operating cash flow (σ(CFO )) is 0.06 and that of sales (σ(Sales)) is 0.15, which shows that sales are more volatile than cash flows. The average of NegEarn is 0.11, which means that 11% of firms reported a loss in the last five years. The minimum of MB (market value to book value) is less than 1 (0.67) and its maximum is much greater than 1 (49.13), which shows that the market value of some firms is much higher than their book value, while that of other firms is even lower than their book value. The summary statistics for the subsamples show that compared with NSOEs, SOEs are larger (Size) and have a more stable operating cash o fl w ( σ(CFO)), a shorter operating cycle (log(Opercycle)), lower incidence of negative earnings realisation (NegEarn), and lower level of growth (MB). Table 4 reports the estimation of AQ. Panel A presents the regression results for 2006. The estimated coefficients for σ(CFO), σ(Sales) and NegEarn are significantly positive, and the estimated coefficient of Size is significantly negative. The results are similar to those of Dechow and Dichev (2002) and Francis et al. (2005). The explanatory power of the summary indicators of the innate factors is 25.7%. The regression results for the years 2001 through 2005 are quantitatively similar to those for 2006. CHINA JOURNAL OF ACCOUNTING STUDIES 163 Table 1. Variable definition. Variable Definition Size natural logarithm of total assets σ(CFO) t he standard deviation of operation cash flows from operations over the past 5 years σ(Sales) t he standard deviation of operation sales from operations over the past 5 years l og(Opercycle) natural logarithm of operating cycle (the sum of days accounts receivable and days inventory) NegEarn t he incidence of negative earnings over the past 5 years Beta one-year beta estimated from firm-specific C aPM estimation MB t he market value to the book value AQ t he measure of accruals quality which is defined as standard deviation of residuals from years t–2 to t by annual cross-sectional estimations of the modified d echow and dichev (2002) model AQ1 t he measure of accruals quality which is defined as the absolute value of abnormal accruals generated by the modified Jones (1991) approach. IAQ t he measure of innate accruals quality which is the predicted value obtained from the annual parameter estimates from f rancis et al. (2005) model DAQ t he measure of discretionary accruals quality which is the residual from f rancis et al. (2005) model SMB return to size factor-mimicking portfolio HML return to book-to-market factor-mimicking portfolio AQfactor return to the AQ factor-mimicking portfolio AQfactor1 return to the AQ1 factor-mimicking portfolio IAQfactor return to IAQ factor-mimicking portfolio DAQfactor return to DAQ factor-mimicking portfolio ORECTA o ther receivables scaled by total assets SOE an indicator variable that equals 1 if the firm is ultimately controlled by central or local governments, and 0 otherwise. Lev t otal liabilities divided by total assets. First t he percentage of shares held by the largest shareholder. Secfive t he sum of the percentage of shareholding held by the second to the fifth largest shareholders Inddir t he percentage of the number of independent directors on the board. ROA return on total assets L_ORECTA t he lag of ORECTA FRisk t he first principal component of Turnover, ReturnVol, FirmAge and MktValue Turnover t he average daily trading volume as a percentage of shares outstanding over the last six months ReturnVol t he standard deviation of daily returns for the last 25 business days FirmAge t he number of months since the initial Sinof in monthly return for the firm MktValue t he total market value of shares outstanding as of the most recent d ecember 31 Table 2. Sample distributions. Panel A: Distribution across years Year 2001 2002 2003 2004 2005 2006 t otal nSoe 90 153 181 225 226 265 1,140 Soe 359 492 564 590 652 654 3,311 t otal 449 645 745 815 878 919 4,451 Panel B: Distribution across industries industry n o. industry n o. f arming, forestry, animal husbandry, and fishing 83 utilities 198 Mining 28 Construction 67 f ood & beverage 204 t ransportation and Warehousing 154 t extile, apparel, fur and l eather 173 information t echnology 261 Paper and allied Products; Printing 64 Wholesale and retail t rades 417 Petroleum, Chemical, Plastics, and rubber Products 514 real estate 213 Manufacturing electronics 130 Public f acilities and o ther Services 157 Metal, non-metal 376 Communication and Cultural industries 37 Machinery, equipment, and instrument Manufacturing 695 Conglomerates 382 Medicine and Biological Products 268 t otal 4,451 o ther Manufacturing 30 164 S. WU ET AL. Table 3. d escriptive statistics. Mean Standard deviation Median Minimum Maximum Panel A: Full sample (n = 4,451) Size 21.27 0.94 21.22 19.08 24.49 σ(CFO) 0.06 0.05 0.05 0.01 0.27 σ(Sales) 0.15 0.15 0.11 0.01 0.96 l og(Opercycle) 5.10 1.01 5.06 2.42 8.15 NegEarn 0.11 0.17 0.00 0.00 0.60 Beta 1.10 0.31 1.12 0.01 1.87 MB 3.51 3.90 2.42 0.67 49.13 Panel B: SOE sample (n = 3,311) Size 21.38 0.94 21.35 19.08 24.49 σ(CFO) 0.06 0.04 0.05 0.01 0.27 σ(Sales) 0.15 0.15 0.11 0.01 0.96 log Opercycle) 5.01 0.98 4.97 2.42 8.15 NegEarn 0.10 0.16 0.00 0.00 0.60 Beta 1.10 0.30 1.11 0.06 1.87 MB 3.31 3.71 2.36 0.67 49.13 Panel C: NSOE sample (n = 1,140) Size 20.96 0.86 20.94 19.08 24.02 σ(CFO) 0.07 0.05 0.06 0.01 0.27 σ(Sales) 0.15 0.14 0.10 0.01 0.96 l og(Opercycle) 5.35 1.06 5.31 2.42 8.15 NegEarn 0.15 0.19 0.00 0.00 0.60 Beta 1.10 0.35 1.14 0.01 1.87 MB 4.09 4.37 2.61 0.67 49.13 Panel D: SOE sample-NSOE sample Mean difference t-test Median difference Wilcoxon test Size 0.42 13.78*** 0.41 12.36*** σ(CFO) –0.01 –6.30*** –0.01 –7.45*** σ(Sales) 0.01 1.24 0.01 0.70 l og(Opercycle) –0.34 –9.56*** –0.35 –10.07*** NegEarn –0.05 –7.58*** –0.00 –7.91*** Beta –0.00 –0.06 –0.03 –0.04 MB –0.78 –5.39*** –0.26 –5.46*** *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. CHINA JOURNAL OF ACCOUNTING STUDIES 165 Table 4. estimations and descriptive statistics of accruals quality. Panel A: Regression result of AQ on innate factors (2006) Predicted sign estimate t value intercept 0.081 3.71*** Size – –0.003 –2.91*** σ(CFO) + 0.119 5.63*** σ(Sales) + 0.014 2.18** l og(Opercycle) + –0.001 –1.11 NegEarn + 0.074 13.18*** a dj. R 0.257 Panel B: Descriptive statistics Mean Standard deviation Median Minimum Maximum (1) f ull sample AQ 0.03 0.03 0.02 0.00 0.19 IAQ 0.03 0.02 0.03 0.01 0.10 DAQ 0.00 0.03 –0.00 –0.07 0.14 (2) Soe sample AQ 0.03 0.03 0.02 0.00 0.19 IAQ 0.03 0.02 0.03 0.01 0.10 DAQ 0.00 0.03 –0.00 –0.07 0.14 (3) nSoe sample AQ 0.04 0.04 0.03 0.00 0.19 IAQ 0.04 0.02 0.03 0.01 0.10 DAQ –0.00 0.03 –0.00 –0.07 0.14 note: t he dependent variable of Panel a is AQ . *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. Based on these parameters, we report the AQ decompositions in Panel B. For the full sample, the mean and median of AQ are 0.03 and 0.02, respectively. The minimum AQ is 0.00, while the maximum AQ is 0.19, which shows that there is a big difference in AQ among our sample. A big difference is also found between IAQ and DAQ. The minimums of IAQ and DAQ are 0.01 and –0.07, respectively, and the corresponding maximums are 0.10 and 0.14, respec- tively. The mean of DAQ is 0, because DAQ is the residual from Equation (2). The average AQ and IAQ for NSOEs are 0.04 and 0.04, respectively, and both are slightly larger than those for SOEs (0.03 and 0.03). Table 5 reports the mean results of firm-specific market-pricing regressions using the CAPM. Panel A reports the average coefficient estimates across the 973 firm-specific estima - tions. Panels B and C report the average coefficient estimates across the 667 SOEs and 306 NSOEs, respectively. Panel D reports the differences in the mean results between the SOE and NSOE subsamples, respectively. Model 1 is the benchmark model. Model 2 is for the market pricing of accruals quality and Model 3 is for that of innate accruals quality and discretionary accruals quality. For the full sample, the adjusted R increases from 0.546 to 0.556 when AQfactor is added (Models 1 and 2 in Panel A). The mean coefficient for AQfactor is 0.083 and is significant at the 1% level, which suggests that accruals quality has been significantly priced (Model 2 in Panel A). When we break down the sample into the SOE and NSOE subsamples, we find that the mean coefficient for AQfactor is –0.050 and insignificant for the former companies (Model 2 in Panel B), while it is 0.374 and significant at the 1% level for the latter companies (Model 2 in Panel C). The statistical results in Panel D show that Panels a, B and C in t ables 4 to 7 report the mean values of firm-specific asset-pricing regressions using the approach of f ama and MacBeth (1973). t herefore we present only the results of mean test and median test in Panel d . 166 S. WU ET AL. Table 5. Mean results of firm-specific market-pricing regressions: C aPM model. Model 1 Model 2 Model 3 R –R R –R R –R j F j F j F Estimate t value Estimate t value Estimate t value Panel A: Full sample R –R 0.963 136.76*** 0.955 145.51*** 0.933 140.91*** M F AQfactor 0.083 4.31*** IAQfactor 0.228 5.91*** DAQfactor 0.087 1.86* a dj. R 0.546 0.556 0.582 Panel B: SOE sample R –R 0.945 119.16*** 0.941 119.94*** 0.934 131.02*** M F AQfactor –0.050 –1.41 IAQfactor 0.046 1.15 DAQfactor –0.019 –0.39 a dj. R 0.522 0.535 0.565 Panel C: NSOE sample R –R 1.002 71.63*** 0.985 73.59*** 0.931 65.43*** M F AQfactor 0.374 5.40*** IAQfactor 0.624 7.64*** DAQfactor 0.319 3.19*** a dj. R 0.487 0.498 0.534 Panel D: SOE sample-NSOE sample Mean diff t value Mean diff t value AQfactor –0.424 –5.45*** IAQfactor –0.578 –6.36*** DAQfactor –0.338 –2.98*** notes: t he dependent variable is R –R , which is excess return on risk-free return. R –R is excess return on the market portfolio. j F M F *** ** * Significant at the 1% level; significant at the 5% level; significant at the 10% level. NSOEs have a significantly larger mean coefficient for the AQfactor than SOEs. The results support Hypothesis 1, i.e. the market pricing effect of accruals quality is greater for NSOEs than for SOEs. When we divide AQ into IAQ and DAQ, we find that for the full sample, the mean coefficient for IAQfactor is 0.228 and significant at the 1% level, whereas that for DAQfactor is just 0.087 and significant at the 10% level (Model 3 in Panel A). However, neither of the coefficient means of IAQfactor and DAQfactor is significant for the SOE sample (Model 3 in Panel B). In contrast, both of the coefficient means of IAQfactor and DAQfactor for NSOEs are significant at the 1% level: the mean coefficient for IAQfactor is 0.624 and that for DAQfactor is 0.319 (Model 3 in Panel C). The loading of IAQfactor is significantly greater than that of DAQfactor , which suggests that innate accruals quality has a greater effect on market pricing than discretionary accruals quality among NSOEs. The coefficient for DAQfactor is significantly positive in NSOEs and suggests that the opportunistic role dominates the informational role of discretionary accruals in NSOEs. The statistical comparison in Panel D shows that the mean difference of the coefficients for IAQfactor is –0.578 and significant at the 1% level between the SOE and NSOE subsamples. In addition, the mean coefficient for DAQfactor for SOEs is significantly smaller than that for NSOEs. This is consistent with Hypothesis 2, which proposes that the market pricing effects of both innate and discretionary accruals quality are larger for NSOEs than for SOEs. We use the Fama–French three-factor model to investigate the market pricing of accruals quality. Table 6 reports the mean results of firm-specific market-pricing regressions. As in Table CHINA JOURNAL OF ACCOUNTING STUDIES 167 5, Panel A reports the average coefficient estimates across the 973 firm-specific estimations. Panels B and C report the average coefficient estimates across the 667 SOEs and 306 NSOEs, respectively. Model 1 is the benchmark model. The adjusted R increases from 0.577 to 0.582 when AQfactor is considered (Models 1 and 2 in Panel A). The coefficient means for AQfactor are –0.064 and 0.208 for the SOE and the NSOE subsample, respectively (Panels B and C, Model 2). The statistical result in Model 2 of Panel D shows that NSOEs have a significantly larger mean than SOEs. The coefficient means for IAQfactor and DAQfactor are respectively 0.085 and –0.023 for the SOE subsample, and they are respectively 0.497 and 0.253 for the NSOE subsample. A comparison of the results (Model 3 in Panel D) shows that between the SOE and NSOE subsamples, the mean differences of the coefficients for IAQfactor and DAQfactor are –0.412 and –0.275, respectively, both of which are significant at the 1% level. The results show larger loadings for both IAQfactor and DAQfactor for NSOEs than for SOEs. Hence, the results in Table 6 are qualitatively similar to those in Table 5 and support our two hypotheses. To provide some perspective on economic significance, we calculate the standard devi- ations of AQfactor, IAQfactor, DAQfactor, R – R , SMB, HML for NSOEs, which are 0.03, 0.04, j F Table 6. Mean results of firm-specific market-pricing regressions: f ama–f rench model. Model 1 Model 2 Model 3 R –R R –R R –R j F j F j F Estimate t value Estimate t value Estimate t value Panel A: Full sample R –R 0.947 147.47*** 0.943 147.57*** 0.931 142.21*** M F SMB 0.100 5.70*** 0.080 4.55*** 0.005 0.23 HML –0.070 –3.06*** –0.073 –3.27*** –0.022 –0.99 AQfactor 0.021 0.65 IAQfactor 0.214 4.80*** DAQfactor 0.064 1.49 a dj. R 0.577 0.582 0.593 Panel B: SOE sample R –R 0.941 128.92*** 0.935 129.34*** 0.932 126.42*** M F SMB 0.020 0.97 0.015 0.70 -0.032 –1.16 HML –0.017 –0.65 –0.030 –1.18 -0.004 –0.16 AQfactor –0.064 –1.69 IAQfactor 0.085 1.71* DAQfactor –0.023 –0.47 a dj. R 0.560 0.557 0.562 Panel C: NSOE sample R –R 0.959 75.03*** 0.953 74.69*** 0.928 70.08*** M F SMB 0.277 8.42*** 0.223 7.16*** 0.087 1.98** HML –0.185 –4.17*** –0.166 –3.87*** –0.059 –1.43 AQfactor 0.208 3.06*** IAQfactor 0.497 5.50*** DAQfactor 0.253 2.95*** a dj. R 0.529 0.573 0.579 Panel D: SOE sample-NSOE sample Mean diff t value Mean diff t value AQfactor –0.272 –3.54*** IAQfactor –0.412 –4.00*** DAQfactor –0.275 –2.80*** notes: t he dependent variable is R –R , which is excess return on risk-free return. R –R is excess return on the market j F M F portfolio. *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. 168 S. WU ET AL. 0.03, 0.18, 0.06, 0.04, respectively. According to Panel C of Table 6, a one standard deviation increase in AQfactor corresponds to a 3% increase in standard deviation of R –R j F (0.208 × 0.03/0.18 = 3%). Similarly, the ee ff cts are 12% and 4% respectively for IAQfactor and DAQfactor while the impact for SMB (HML) is 8% (4%). Relative to SMB and HML, the market pricing effect of accruals quality is economically significant for NSOEs. 5. Robustness checks We also use the modied fi Jones ( 1991) model to compute an alternative measure of accruals quality and use it as a robustness check. We estimate the following cross-sectional regression for each of the industry groups with at least 10 firms in year t. TAC ∕A =  [1∕A ]+  [ΔREV ∕A ]+  [PPE ∕A ]+ (7) j,t j,t−1 0 j,t−1 1 j,t j,t−1 2 j,t j,t−1 j,t The industry- and year-specific parameter estimates obtained from Equation (7) are used to estimate firm-specific normal accruals (NA) as a percentage of lagged total assets: NA =  [1∕A ]+  [(ΔREV −ΔAR )∕A ]+  [PPE ∕A ] j,t 0 j,t−1 1 j,t j,t j,t−1 2 j,t j,t−1 where ΔAR is firm j’s change in accounts receivable between year t–1 and year t, and j,t AA = TAC /A − NA is the abnormal accruals (AA) in year t. The absolute value AQ1 = |AA | j,t j,t j,t-1 j,t j,t j,t is our alternative proxy for accruals quality, with larger values of AQ1 indicating poorer j,t accruals quality. The results in Tables 7 and 8 show that all the estimated coefficients on AQfactor1 for SOEs are insignificant while those for NSOEs are significantly positive. The coefficients for SOEs are much smaller than those for NSOEs. Overall, using an alternative measure of accruals quality does not change the results. Table 7. an alternative measure of accruals quality: C aPM model. Model 1 Model 2 R –R R –R j F j F Estimate t value Estimate t value Panel A: Full sample R –R 0.932 92.35*** 0.920 90.44*** M F AQfactor1 0.190 4.71*** a dj. R 0.456 0.468 Panel B: SOE sample R –R 0.931 80.05*** 0.930 79.62*** M F AQfactor1 0.042 1.13 a dj. R 0.462 0.472 Panel C: NSOE sample R –R 0.994 60.80*** 0.959 57.02*** M F AQfactor1 0.447 6.07*** a dj. R 0.455 0.468 Panel D: SOE sample-NSOE sample Mean diff t value AQfactor1 –0.404 –4.90*** notes: t he dependent variable is R –R , which is excess return on risk-free return. R –R is excess return on the market j F M F portfolio. *** ** * Significant at the 1% level; significant at the 5% level; significant at the 10% level. CHINA JOURNAL OF ACCOUNTING STUDIES 169 Table 8. an alternative measure of accruals quality: f ama–f rench model. Model 1 Model 2 R –R R –R j F j F Estimate t value Estimate t value Panel A: Full sample R –R 0.888 114.62*** 0.885 108.61*** M F SMB 0.125 6.83*** 0.120 6.28*** HML –0.096 –5.03*** –0.087 –4.52*** AQfactor1 0.051 1.24 a dj. R 0.496 0.505 Panel B: SOE sample R –R 0.916 101.00*** 0.916 98.64*** M F SMB 0.059 3.46*** 0.059 3.10*** HML –0.043 –2.75*** –0.043 –2.79*** AQfactor1 –0.006 –0.13 a dj. R 0.499 0.507 Panel C: NSOE sample R –R 0.887 63.42*** 0.884 59.67*** M F SMB 0.299 9.57*** 0.270 8.77*** HML –0.201 –5.81*** –0.201 –5.87*** AQfactor1 0.142 2.08** a dj. r 0.507 0.511 Panel D: SOE sample-NSOE sample Mean diff t value AQfactor1 –0.147 –1.84* note: t he dependent variable is rj–rf , which is excess return on risk-free return. rM–rf is excess return on the market portfolio. *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. 6. Further analysis We propose in our hypothesis development that the largest shareholders of SOEs have less incentive to expropriate firm resources, so SOEs have a lower level of fundamental risk than NSOEs, which leads to different market pricing. We test in this part whether this mechanism exists. The expropriation by the largest shareholder is just one of the factors that may affect fundamental risk. By analysing the expropriation, we aim to prove that the fundamental risk of NSOEs is different from that of SOEs. We have stated that the market pricing effect is higher in NSOEs than in SOEs. This is due to the fact that NSOEs have a higher level of expropriation by the largest shareholder, and thus have a higher level of fundamental risk. Although it is a common sense, we provide direct evidence here. First, we examine the impact of state ownership on expropriation by the largest shareholder. Following Jiang et al. (2010), we measure expropriation using ORECTA (Other Receivable scaled by Total Assets). SOE is an indicator variable that equals 1 if the client is ultimately controlled by central or local governments, and 0 otherwise. Size is the natural logarithm of total assets. Lev is the total liabilities divided by total assets. First is the percentage of shares held by the largest shareholder. Secfive is the sum of the percentage of shareholding held by the second to the fifth largest shareholders. Inddir is the percentage of the number of independent directors on the board. ROA is the return on total assets. L_ORECTA is the lag of ORECTA. We regress ORECTA on SOE and Table 9 presents the regression result. We find that the coefficient on 170 S. WU ET AL. Table 9. t he effects of Soe on expropriation and fundamental risk. Model 1 Model 2 Model 3 ORECTA FRisk FRisk Estimate t value Estimate t value Estimate t value ORECTA 0.190 3.11*** SOE –0.012 –6.24*** –0.032 –2.91*** Size 0.163 4.20*** –0.035 –6.43*** –0.031 –5.62*** Lev 0.006 1.30 0.131 4.56*** 0.118 4.10*** First 0.566 4.52*** –0.307 –11.57*** –0.282 –11.04*** Secfive 13.769 3.36*** 0.077 11.58*** 0.071 11.05*** Inddir –0.017 –1.98** 0.093 2.26** 0.094 2.28** ROA –0.250 –17.33*** –0.385 –4.36*** –0.323 –3.54*** L_ORECTA 0.638 60.23*** σ(CFO) –0.065 –0.61 –0.112 –1.06 σ(Sales) 0.180 5.41*** 0.187 5.63*** l og(Opercycle) 0.030 6.01*** 0.028 5.57*** NegEarn 0.195 6.01*** 0.186 5.75*** intercept –3.163 –4.17*** 1.084 9.08*** 0.980 8.18*** N 4,450 4,440 4,440 a dj. R 0.569 0.127 0.127 *** ** * Significant at the 1% level; Significant at the 5% level; Significant at the 10% level. SOE is significantly negative (–0.012, t = –6.24, Model 1), which suggests a higher level of expropriation by the largest shareholder in NSOEs than in SOEs. Then we examine the impact of state ownership on fundamental risk. Chen et al. (2008) measure fundamental risk using market capitalisation, firm age, return volatility and trading volume. As in Chen et al. (2008), we combine the four proxies into a summary fundamental risk measure (FRisk) using principal component analysis. We regress FRisk on SOE and Table 9 presents the regression result. We can find that the coefficient on SOE is significantly neg- ative (–0.032, t = –2.91, Model 2), which suggests that the fundamental risk is lower in SOEs than that in NSOEs. Finally, we examine the effect of expropriation on fundamental risk. Similarly, we regress FRisk on ORECTA and Table 9 presents the regression result. We can see that the coefficient on ORECTA is significant positive (0.190, t = 3.11, Model 3), which is consistent with the fundamental risk being lower in low expropriation firms than in high expropriation firms. 7. Conclusions Existing literature has studied the role of state ownership in investor protection from the perspective of the effect of state ownership on accruals quality but ignores the effect of state ownership on the market pricing of accruals quality. Our paper investigates this problem using data of Chinese listed firms. We find that the market pricing effect of accruals quality is larger for NSOEs than for SOEs. The results hold when we divide accruals quality into innate and discretionary accruals quality, which suggests that state ownership improves the level of investor protection. Our study adds to the growing literature on the governance role of state ownership and the market pricing of accruals quality. The findings of this paper also have important practical implications. The results of our study again show that SOEs provide better investor protection than NSOEs. Therefore, supervisory departments should pay closer attention to the investor protection problems in NSOEs. Meanwhile, as the market is more concerned with the CHINA JOURNAL OF ACCOUNTING STUDIES 171 accounting information of NSOEs, supervisory departments should care more about the quality of accounting information of NSOEs. Auditors should also devote more effort and maintain a higher level of professional scepticism when auditing NSOEs. Disclosure statement No potential conflict of interest was reported by the authors. Funding Xianhui Bo acknowledges financial support from the National Natural Science Foundation of China [grant number 71302130]. References Aharony, J., Lee, C.-W.J., & Wong, T.J. (2000). Financial packaging of IPO firms in China. Journal of Accounting Research, 38(1), 103–126. Bai, Y., Lin, B., Wang, Y., & Wu, L. (2013). Full privatization, expropriation, and firm performance: Evidence from China. Applied Economics, 45, 1857–1867. Bo, X., & Wu, L. (2009). The governance roles of state controlling and institution investors: A perspective of earnings management. Economic Research Journal, 2, 81–91 (In Chinese). Brandt, L., & Li, H. (2003). Bank discrimination in transition economies: Ideology, information, or incentives? Journal of Comparative Economics, 31, 387–413. Calomiris, C., Fisman, R., & Wang, Y. (2010). Profiting from government stakes in a command economy: Evidence from Chinese asset sales. Journal of Financial Economics, 96, 399–412. Chen, H., Chen, J.Z., Lobo, G.J., & Wang, Y. (2010). Association between borrower and lender state ownership and accounting conservatism. Journal of Accounting Research, 48, 973–1014. Chen, L., Dhaliwal, D., & Trombley, M. (2008). The effect of fundamental risk on the market pricing of accruals quality. Journal of Accounting, Auditing & Finance, 23, 471–492. Claessens, S., Djankov, S., Fan, J.P.H., & Lang, L.H.P. (2002). Disentangling the incentive and entrenchment effects of large shareholdings. The Journal of Finance, 57, 2741–2771. Dechow, P., & Dichev, I. (2002). The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77(Supplement), 35–59. Deloitte Touche Tohmatsu. (2006). China’s new accounting standards: A comparison with current PRC GAAP and IFRS. http://www.iasplus.com/country/china.htm. Ding, Y., Zhang, H., & Zhang, J. (2007). Private vs. state ownership and earnings management: Evidence from Chinese listed companies. Corporate Governance: An International Review, 15, 224–239. Drew, M.E., Naughton, T., & Veeraraghavan, M. (2003). Firm size, book-to-market equity and security returns: Evidence from the Shanghai Stock Exchange. Australian Journal of Management, 28, 119–139. Ecker, F., Francis, J., Kim, I., Olsson, P.M., & Schipper, K. (2006). A returns-based representation of earnings quality. The Accounting Review, 81, 749–780. Fama, E.F., & French, K.R. (1992). The cross-section of expected stock returns. Journal of Finance, 47, 427–465. Fama, E.F., & French, K.R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56. Fama, E.F., & MacBeth, J. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81, 607–636. Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2005). The market pricing of accruals quality. Journal of Accounting and Economics, 39, 295–327. Guay, W.R., Kothari, S.P., & Watts, R.L. (1996). A market-based evaluation of discretionary accrual models. Journal of Accounting Research, 34 (Supplement), 83–105. 172 S. WU ET AL. Huang, F., Su, J., & Chong, T. (2008). Testing for structural change in the nontradable share reform of the Chinese stock market. The Chinese Economy, 41, 24–33. Jensen, M.C., & Meckling, M.H. (1976). Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3, 305–360. Jian, M., & Wong, T.J. (2010). Propping through related party transactions. Review of Accounting Studies, 15(1), 70–105. Jiang, G., Lee, C., & Yue, H. (2010). Tunneling through inter-corporate loans: The China experience. Journal of Financial Economics, 98(1), 1–20. Jones, J. (1991). Earnings management during import relief investigations. Journal of Accounting Research, 29, 193–228. Kornai, J. (1993). The evolution of financial discipline under the postsocialist system. Kyklos , 46, 315–336. Li, J., & Qian, C. (2013). Principal-principal conflicts under weak institutions: A study of corporate takeovers in China. Strategic Management Journal, 34, 498–508. Li, H., & Zhou, L.-A. (2005). Political turnover and economic performance: The incentive role of personnel control in China. Journal of Public Economics, 89, 1743–1762. Lin, Y., & Tan, G. (1999). Policy burdens, accountability, and the soft budget constraint. American Economic Review, 89, 426–431. McNichols, M.F. (2002). Discussion of the quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77 (Supplement), 61–69. Qian, Y. (2001). Government control in corporate governance as a transitional institution: lessons from China. In J. Stiglitz & S. Yusuf (Eds.), Rethinking the East Asian miracle (pp. 295–321). New York, NY: Oxford University Press and the World Bank. Shen, H., & Wu, L. (2012). Corporate ownership, environment uncertainty and the governance role of accounting information. Accounting Research, 8, 8–16. (In Chinese). Subramanyam, K.R. (1996). The pricing of discretionary accruals. Journal of Accounting and Economics, 22, 249–281. Sun, Q., & Tong, W.H.S. (2000). The effect of market segmentation on stock prices: The China syndrome. Journal of Banking and Finance, 24, 1875–1902. Wang, Q., Wong, T.J., & Xia, L. (2008). State ownership, the institutional environment, and auditor choice: Evidence from China. Journal of Accounting and Economics, 46, 112–134. Yee, K.K. (2006). Earnings quality and the equity risk premium: A benchmark model. Contemporary Accounting Research, 23, 833–877. Zhang, H., & Li, G.J. (2011). Study on the efficient of Chinese stock market. Journal of Data Analysis, 6(1), 63–82.

Journal

China Journal of Accounting Studies – Taylor & Francis

Published: Apr 3, 2017

Keywords: Accruals quality; investor protection; market pricing; state ownership

There are no references for this article.