DeepDyve requires Javascript to function. Please enable Javascript on your browser to continue.
Accounting conservatism, and bank loan contracts: Evidence from China Accounting conservatism, and bank loan contracts: Evidence from China
Zhao, Gang; Liang, Shangkun; Wang, Yutao
2014-07-03 00:00:00
China Journal of Accounting Studies, 2014 Vol. 2, No. 3, 200–227, http://dx.doi.org/10.1080/21697213.2014.945131 Accounting conservatism, and bank loan contracts: Evidence from China a b b Gang Zhao , Shangkun Liang * and Yutao Wang School of Economics and Management, Changzhou University, Changzhou, 213164, China; School of Accountancy, Central University of Finance and Economics, Beijing, 100081, China Accounting conservatism plays an important role in bank loan contracts. A high degree of accounting conservatism can reduce the risks faced by bank lenders through reporting losses in a timely manner. Banks may then offer preferential poli- cies to borrowers. This paper investigates this hypothesis, using hand-collected data on bank loans granted to Chinese A-share listed companies from 2000 to 2007. The empirical results show that companies with stronger conservatism tend to obtain larger loan amounts with lower interest rates and longer maturities. In addition, we find that the effect of accounting conservatism on bank loans is more significant in regions with better legal environments, more preferential policies, and less intervention from the local government. Keywords: accounting conservatism; bank loan contract; institutional environment 1. Introduction Allen and Carletti (2008) have proposed that the efficiency of the process of channel- ling bank savings into productive activities is crucial for growth and general welfare. Tan (1999), Li, Ran, and Wen (2007), and Fang, Sun, and Cao (2010) have also found that bank loans are positively associated with promoting economic development, based on Chinese data. Bank loans have always been an important driving force in the devel- opment of China’s economy. Compared with most European countries, China’s stock market is not sufficiently developed. Indirect financing (e.g., bank loans) is the main method of financing for the non-financial sector in China. Yin, Wang, Guo, Li, and Wang (2012) show that in recent years, due to the strong development of the stock and bond markets, the dependence of financing on banks has declined, but from 2007 to 2011 the average proportion of bank loans for all financing was still more than 76%. The current financial system in China is dominated by a large banking sector. The stability of the bank system is important to the development of China’s economy and has become even more so since the 2008 global financial crisis (Allen, Qian, Qian, & Zhao, 2009). Bank credit risk plays a critical role in the stability of the financial system. Hence, the factors influencing bank loans have become an important concern for academics, regulatory agencies, and market participants. However, there is little in the literature about this issue. Relatively few studies consider bank loans an economic consequence of accounting conservatism (Ahmed, Billings, Morton, & Harris, 2002; Zhang, 2008), especial for developing countries. *Corresponding author. Email: jevonacc@gmail.com. Paper accepted by Kangtao Ye. © 2014 Accounting Society of China China Journal of Accounting Studies 201 Compared with developed countries, credit allocation in developing countries relies heavily on non-market mechanisms, such as ‘relationships’ and government interven- tion (e.g. Berger, Hasan, & Zhou, 2009). Ball, Robin, and Wu (2003) find that the low accounting quality of four East Asian countries is due to guanxi. Guanxi between com- panies and banks can reduce incentives to provide high-quality accounting information. The government has intervened in loan allocations for many years in China (Berger et al., 2009). Within the context of a society with more informal relationship and gov- ernment intervention, whether and how the characteristics of accounting information affect bank loans has not yet been made clear. Therefore, this paper analyses this issue and provides empirical evidence from China. This paper conducts a comprehensive study on the characteristics of bank loans. Previous studies on loan contracts often focus on a particular point. For example, Zhang (2008) focuses on the effect of accounting conservatism on lending rates. Liu, Wu, and Jiang (2010) and Xu and Shen (2010) investigate loan amounts and maturity. Unlike these studies, this paper focuses on the three characteristics of bank loans, namely, loan amounts, maturity structures, and lending rates. The results across the three characteristics are consistent, reinforcing the reliability of our conclusions. This paper analyses the relation between accounting conservatism and bank loans. This extension is more important under China’s restricted regulation of interest rates (Wang, 2001; Wang & Zhang, 2007). On the one hand the lending rate reflects the cost of a loan; it is the most important feature of the loan contract. For example, a company obtains a loan for a large amount and with a long maturity but may have to pay higher interest, so the combined effect is difficult to evaluate. On the other hand, because lending rates are strictly regulated in China, accounting conservatism may affect loan amounts and maturity structures but not lending rates. Therefore, focusing only on loan rates will hinder a complete understanding of the correlation between accounting conservatism and bank loans. Accounting conservatism is an important indicator of accounting information qual- ity. It refers to asymmetry in the recognition of losses (bad news) and earnings (good news) in accounting. Companies will be timelier at recognizing losses and will have sufficient supporting evidence before recognising earnings (Basu, 1997). Accounting conservatism can decrease the information asymmetry between borrowers (companies) and creditors (banks), thus improving the efficiency of loan contracts (Watts, 2003a, 2003b; Watts & Zimmerman, 1986; Zhang, 2008). Hence, decreasing information asymmetry can help banks better understand the firms’ situation, thus finally reducing the transaction costs. Based on these arguments, we predict that banks are more willing to provide conservative companies with larger loan amounts, lower interest rates, and longer maturities. We use manually collected information on bank loans granted to China’s A-share listed companies between 2000 and 2007 to empirically verify the influence of accounting conservatism on banks’ loan contracts. We use the C-Score developed by Khan and Watts (2009) to measure the accounting conservatism of com- panies. We find that when a company exhibits stronger accounting conservatism, it tends to obtain larger loan amounts with longer maturities and lower interest rates. We further analyse institutional environments and find that companies in regions with a bet- ter legal environment enjoy more preferential policies and less government intervention and that the effect of accounting conservatism is more significant in these cases. This paper contributes to the literature in the following ways. First, it provides addi- tional comprehensive evidence on how accounting information affects bank loans. On the one hand, bank loans play a vital role in corporate financing. On the other hand, 202 Zhao et al. against a background of relationships and government intervention, it is unclear whether accounting conservatism affects bank loans. Therefore, this study can provide more evidence on the relation between accounting information and bank loans. Second, institutional environment analysis can help us better understand the role of accounting conservatism in different situations, such as the legal environment, preferential policies, and government intervention. The findings of this paper show that a better institutional environment can increase the importance of accounting conservatism and decrease transaction costs for contracting parties to a greater degree, thereby improving contract efficiency. Developmental differences among regions in China provide a good experi- mental environment for our analysis. Third, this paper focuses on unique individual loan data that offer us more measurements than prior studies to proxy for the main variables. For example, prior studies often calculate loan interest rates in terms of finan- cial costs and total borrowings and investigate the compositions of long- and short- maturity loans. We directly use the exact interest rates and loan maturity of each loan to analyse the relation between accounting conservatism and bank loans, which can improve the accuracy of the results. The rest of the paper is structured as follows. Section 2 presents the institutional background. Section 3 presents a review of the literature. Section 4 contains the hypothesis development and research design. Section 5 presents the sample selection along with descriptive statistics and an empirical test. The robustness tests are con- ducted in Section 6. Section 7 concludes and acknowledges limitations. 2. Institutional background According to Lin (2005), in shifting from a centrally planned economy to a market economy, China has carried out its regulatory reforms in the banking system since 1984 in three stages. Lin (2005) also notes that, prior to 1984, the banking industry in China was dominated by a single financial institution, the People’s Bank of China (PBC). Lin explains that the first stage of the reform was to separate commercial banks from the PBC. These commercial banks included the Industrial and Commercial Bank of China, the Bank of China, China Construction Bank, and the Agricultural Bank of China (Lin, 2005). After this stage, the PBC specialised in performing central bank functions, forming the Chinese dual banking system. Lin (2005) continues the description by explaining that the four state-owned com- mercial banks mentioned above played a significant role in the second stage of the reform from 1994 to 2003. In 1994, the establishment of three policy banks marked the separation of policy banks from commercial banks. In 1995, the PRC Commercial Banking Law was promulgated, mandating that the four state-owned commercial banks begin independent operations and become responsible for their own profit/loss. The Asian financial crisis in 1997 then promoted their market-oriented reform, resulting in such measures as cancellation of the loan size system, strengthening a unified legal sys- tem, development of strict authorisation and credit systems, and establishment of risk control mechanisms (Lin, 2005). Finally, Lin (2005) notes that the third stage of the reform was the establishment of a national joint-stock commercial bank holding structure, in 2003. On 11 December, 2001, China formally joined the World Trade Organization, promoting further banking reform. The China Banking Regulatory Commission was established in 2003 to strengthen bank supervision (Lin, 2005). Alicia, Sergio, and Daniel (2006) show that a series of regulations were implemented, such as the establishment of a loan China Journal of Accounting Studies 203 classification system and a risk assessment system, increased penalties for breaches of contract, and restrictions on association loans. At the same time, the reforms stimulated the development of small and medium-sized banks, increased competition between for- eign and domestic banks, and encouraged domestic banks to attract foreign strategic investors. China’s commercial banks have now fully achieved the goals of the above-men- tioned reforms. Market competition among banks has become even more intense and by the end of 2012, there were 12 joint-stock commercial banks, 144 city commercial banks, 337 rural commercial banks, 147 rural cooperative banks, and 1927 rural credit cooperatives in China (Guo, Ma, Bai, & Huo, 2014). In such a competitive market, banks pay more attention to borrower solvency, profitability, risk control, and the qual- ity of accounting information. 3. Literature review Accounting conservatism research has a long history (Watts & Zimmerman, 1986). However, systematic studies were introduced relatively recently, by Watts (1993), who points out that accounting conservatism stems from the contract function of accounting and is subject to regulation and laws. Researchers therefore began to conduct empirical tests, and studies of accounting conservatism developed. Overall, these studies can be divided into three categories dealing respectively with the measurement methods, causes, and economic consequences of accounting conservatism (Watts, 2003a, 2003b). This paper focuses on one of the economic consequences of accounting conserva- tism, specifically its role in bank loan contracts. Therefore, the literature review focuses on studies of the third type, which deal with the economic consequences of accounting conservatism. The other two categories are, however, briefly discussed. Watts (2003b) summarises the current accounting conservatism measures as earnings–stock return relation measures (Ball, Kothari, & Robin, 2000; Basu, 1997; Givoly & Hayn, 2000; Holthausen & Watts, 2001; Pope & Walker, 1999), net asset measures (Ahmed, Morton, & Schaefer, 2000; Beaver & Ryan, 2000; Dechow, Hutton, & Sloan, 1999; Myers, 1999), and earnings–accrual measures (Elgers & Lo, 1994; Givoly & Hayn, 2000). Ball and Shivakumar’s(2005) cash flow–accrued profits regression model and that of Basu (1997) have been widely implemented. However, both models make it dif- ficult to measure the accounting conservatism of a single company. To resolve this issue, Khan and Watts (2009) proposed the C Score method, which can measure any company’s level of accounting conservatism. Our study adopts their measurement method. Watts (2003a, 2003b) argues that accounting conservatism is derived from account- ing contracts (loan and compensation contracts), legal proceedings, regulation and taxa- tion, and notes that loan contracts play a vitally important role. Ball and Shivakumar (2005), Ball, Robin, and Sadka (2008), and Chen, Chen, Lobo, and Wang (2010) con- duct empirical tests on these issues. Chen et al. (2010), using Chinese data, find that state-owned companies that obtain loans exhibit lower accounting conservatism than private companies. Since the requirements of accounting conservatism for state-owned companies are lower, bank decisions are important factors in determining accounting conservatism. Accounting conservatism usually has two kinds of economic effects: those related to financing and those related to investment activities. Ahmed et al. (2002) find that companies in which greater creditor–shareholder conflicts exist also exhibit higher 204 Zhao et al. accounting conservatism. This means that accounting conservatism eases creditor– shareholder conflicts and thereby decreases the cost of debt. Lara, Osma, and Penalva (2007) apply a variety of cost-of-equity measures and verify that higher accounting conservatism is associated with a lower cost of equity. Moerman (2008) studies accounting conservatism and bid–ask spreads on the secondary loan market and con- cludes that accounting conservatism can significantly decrease the spreads, indicating the importance of decreasing the information asymmetry between companies and sec- ondary loan buyers. Zhang (2008) proposes that accounting conservatism helps decrease transaction costs in loan contracts and thus improves their efficiency. Further, higher accounting conservatism has been found to increase the possibility of debt con- tract violation and the timely delivery of this information can help creditors adjust their risk. As compensation, these companies could receive lower loan interest rates. Chen et al. (2010) find that when a company adopts more conservative accounting policies, it receives more loans from foreign banks. Using data from 25 countries, Bushman, Smith, and Piotroski (2011) address the effect of accounting conservatism on invest- ment activity and conclude that business investment behaviour based on accounting conservatism is asymmetric. When a company’s investment opportunities decrease, its business investments become more sensitive to accounting conservatism. Many Chinese studies focus on the causes of accounting conservatism. Liu et al. (2010) and Xu and Shen (2010) address the supervisory role of loan contracts under different circumstances. Liu et al. (2010) show that companies with many short-matu- rity loans but few long-maturity loans exhibit a higher degree of conservatism and that banks have higher requirements for the conservatism of non-state-owned companies. Xu and Shen (2010) examine the data of Chinese listed companies between 1999 and 2004 and find that larger bank loans are associated with higher accounting conserva- tism. Further, they find that short-maturity loans are more important in monitoring com- pared with long-maturity loans. From the perspective of internal supervision, Zhao, Zeng, and Tan (2008) find that the supervision of board directors, measured by number of board members, board member professional accounting competence, their rewards/ reputation, and the supervision environment have a correlation with firms’ accounting conservatism that increases with improvement in corporate governance. From the per- spective of external supervision, Zhu, Xi, and Chen, (2010) examine the influence of audit tenure on accounting conservatism. They believe that relatively long audit tenure forces audit firms to decrease their investment in audit activities and lower their audit standards. They also find the conservatism of firms that hired audit firms for a longer time is relatively low. Because Big Four audit firms face a greater loss of reputation, the impact of extended audit tenure on conservatism is less pronounced for them. In addition, Chen and Huang (2008) suggest that the accruals of companies change along their life cycles, together with corresponding changes in accounting conservatism. Accordingly, in the early stages of the corporate cycle, a company’s accrual is mostly positive. It will overstate positive accruals and hide more bad news, so its accounting conservatism is kept relatively low. In contrast, at the end of the corporate life cycle, a company’s accrual is mostly negative and its accounting conservatism is relatively high. Xu and Sun (2006) examine the relation between stock market cycles and accounting conservatism from a macroeconomic point of view. They show that, during a stock market downturn, companies must meet the needs of creditors and therefore provide more conservative accounting information. In contrast, when the stock market is boom- ing, the result is the opposite and their empirical results support this. Rao and Jiang (2011) consider the effect of a fluctuating monetary policy on corporate accounting China Journal of Accounting Studies 205 conservatism. When the monetary policy is tightening, companies enforce more conser- vative accounting policies to obtain bank loans. The authors also find that non-state- owned companies that rely more on external financing and have higher debt levels and less monetary capital have higher degrees of accounting conservatism. In terms of investment, Yang, Wang, and Ye (2011) use asset impairment as a proxy for corporate accounting conservatism. They examine its effect on investment behaviour and find that it has a restraining effect on excessive investment behaviour yet also introduces the problem of insufficient company investments. Given the findings of the foreign and domestic literature on bank loan contracting, further research on the role of accounting conservatism in loan contracts should be con- ducted. First, few studies consider bank loans as an economic consequence of conser- vatism (Zhang, 2008). For example, in China, Liu et al. (2010) and Xu and Shen (2010) consider conservatism as the result of loan contract supervision; however, this is not their main point. Second, current concerns over loan contracts are often not com- prehensive. For example, Zhang (2008) focuses on the effect of accounting conserva- tism on loan interest rates and Liu et al. (2010) and Xu and Shen (2010) focus on loan amounts and maturity. Third, prior measures of loan interest rates and maturity used by Chinese researchers have been comparatively approximate. Fourth, while the previous literature addresses external pressures, such as accounting conservatism due to bank requirements (Chen et al., 2010), it is the internal governance of companies that often results in changes in corporate conservatism and thus balances banks’ accounting con- servatism requirements. Therefore, in contrast to the previous literature, we begin our analysis from a company angle and then analyse the idea that different levels of accounting conservatism have different economic consequences for companies. In addition, we provide a wide range of empirical evidence. The results encourage com- mercial banks to take into account the accounting conservatism issues of customer companies to improve their decision-making efficiency in relation to loans. We aim to address the potential insufficiencies of past studies and thus advance the related research. 4. Hypothesis development and research design 4.1. Hypothesis development Accounting conservatism plays a vital role in bank loan contracts and can protect the interests of creditors. Watts (2003a) believes that the existence of accounting conserva- tism benefits the contracting bodies in two ways. First, it limits the opportunistic report- ing behaviour of managers. Accounting information has long been the basis on which companies sign and implement various contracts. Because various contracting bodies have asymmetric information, limited vision, and limited liability, contract fulfilment faces a serious moral hazard problem. Owing to the asymmetric recognition of gains and losses, a downward conservative bias can offset any potential upward bias in man- agers’ accounting information. Second, the contractual function of accounting conserva- tism is also reflected in the restrictions it places on the opportunistic payment behaviour of managers. By underestimating corporate net assets and total net income, accounting conservatism effectively limits managerial opportunistic behaviour and its effect on each contracting body (e.g., creditors), thereby increasing firm value. Further, accounting conservatism can decrease information asymmetry. LaFond and Watts (2008) believe that accounting conservatism is an effective corporate governance mechanism that decreases the motives and abilities of managers in manipulating 206 Zhao et al. accounting reports and facilitates an information environment that allows other active reporting channels. In this case, information asymmetry between managers and external creditors and the deadweight loss due to asymmetric information are decreased, ultimately improving corporate value. The authors believe that conservatism may decrease information asymmetry in two ways. First, conservative accounting can better provide so-called ‘hard’ comprehensive information (information with strong verifiability) about a firm’s current performance to creditors, who are at an informa- tional disadvantage. A relatively high requirement for gain recognition can be used to offset managerial motives to overestimate gains and a relatively low requirement for loss recognition can convey information that managers are unwilling to provide. As a result, conservative accounting can provide larger amounts of more reliable hard information compared with unbiased accounting systems. Second, hard accounting information can constrain and monitor a company’s ‘soft’ information (i.e. information with poor verifiability) in relation to current and future performance and thus serve as a benchmark for evaluating the reliability of sources of mutually competitive and diverse information. Banks have basic goals as creditors: to lend capital and to receive the loan principal and interest on time within the bearable scope of risk. Risk control is a bank’s top pri- ority. However, due to the information asymmetry between companies and banks, banks often find it hard to fully acquire company-related information and to fully and accurately assess the financial situations of companies (Bharath, Dahiya, Saunders, & Srinivasan, 2011). Companies with low degrees of information transparency may pres- ent greater risks and banks should pay special attention and tighten their loan require- ments to counter possible losses. On the other hand, banks could lower their loan requirements for those companies that exhibit higher degrees of information transpar- ency and present lower risks (Zhang, 2008). If companies were to use appropriate accounting methods to decrease information asymmetry and send signals, they could win the trust of banks. An improvement in accounting conservatism is a kind of signal. An improvement in accounting conservatism can help decrease the information asym- metry between companies and banks and decrease managerial opportunistic behaviour so that companies can report bad news in a timely manner and publicise their poor situ- ations sooner (Watts, 2003a, 2003b; Zhang, 2008). In such cases, banks can then make timely adjustments and decrease their lending risks. Further, companies that exhibit high degrees of conservatism offer banks more control over their lending risks. Banks would then be willing to lend to these companies while providing more preferential policies, specifically greater loan amounts, longer maturities, and lower interest rates. This leads to our hypothesis, which has three elements. Hypothesis: The higher a company’s level of accounting conservatism, the more preferen- tial loan policies (greater loan amounts, longer maturities and lower interest rates) it can receive from the banks. 4.2. Research design In this paper, we investigate the influence of accounting conservatism on bank loans. Hence, we calculate accounting conservatism for every firm-year. Based on Basu (1997), Khan and Watts (2009) create a new measurement known as Cscore to proxy for accounting conservatism. We adopt this measurement to conduct a regression. The calculations are presented in the Appendix. China Journal of Accounting Studies 207 To test the hypothesis, we set up the following model involving accounting conser- vatism (Model I), based on Graham, Li, and Qiu (2008): LoanFeature ¼ a þ a Cscore þ ControlVariables þ eðModel IÞ t 0 1 t where the dependent variables Lnloans, Lnterm, and Adjust Rate refer to the loan amount, maturity, and interest rate, respectively. The main explanatory variable, Cscore, measures accounting conservatism. The hypothesis predicts that firms with greater accounting conservatism will obtain larger loan amounts, longer maturities (i.e., α is positive), and lower interest rates (i.e., α is negative). Since not all three features of the loans are reported in the annual financial report, the numbers of observations of the loan amounts, maturity and interest rates are different. Following prior studies (e.g., Bharath, Dahiya, Saunders & Srinivasan, 2011; Graham et al., 2008), we control for other factors that may affect bank loans: Loan , t–1 Size, Leverage, ROE, Tangibility, and MTB. The variable Loan refers to the amount t–1 of bank loans for the previous year. The larger this measure, the higher the financial risk to the company and the more negative the effect on the current year’s bank loans. The variable Size refers to firm size and equals the logarithmic value of total assets at the end of the current year. Large firms exhibit more stable operations, disclose more information, and exhibit less information asymmetry than smaller firms. Hence, they are more likely to obtain larger loan amounts, longer loan maturities, and lower interest rates. The term Leverage refers to the debt ratio. Firms with higher debt ratios may experience higher financial risks and may therefore be charged higher interest by banks. The variable ROE refers to the return of net assets. Firms with a higher ROE decrease their risks and are more likely to obtain bank loans. The term Tangibility refers to the tangible assets ratio, which is equal to the sum of fixed assets, land, and other tangible assets divided by total assets. The higher the ratio, the more mortgage assets the firm can use and the less risky its bank loan. The variable MTB refers to growth and is the market-to-book ratio at the end of the year. Firms experiencing higher growth may face more risks and have more chances to increase income. We add the largest shareholder’s holding (Shareholding), firm ownership type (SOE), and bank type (Bank4) to the model to control for the effects of ownership structure on loan maturity. Dennis, Nandy, and Sharpe (2000) and Bharath et al. (2011) determine that non-interest rate provisions are decided earlier than the interest rate. Hence, we also control for loan amounts and maturity in the regression of interest rates. We also control year and industry effects in the regression. 5. Sample selection, descriptive statistics and empirical analysis 5.1. Sample selection We focus on a sample of A-share listed companies in China during 2000–2007. The data stop at 2007, because of the effect of the financial crisis following 2007 on firms. The financial crisis in 2008 may involve more noise in our analysis and thus mislead our inference. To avoid any potential influence from the crisis, our analyses are based on the sample before 2008. Data related to loan amounts, maturities, and interest rates were manually collected from the appendix notes of the relevant companies’ annual reports. The financial data and industrial classification of listed companies were taken from the China Stock Market & Accounting Research database. The market index was taken from ‘Chinese Market Index – 2009’ (Fan, Wang, & Zhu, 2009) and a study by Demruger et al. (2002). Observations with missing values were deleted. We also 208 Zhao et al. removed companies in the financial industry. We winsorise the main variables at the 1% and 99% levels to alleviate the influence of outliers. Of the three factors, that is, loan amounts, maturity, and interest rates, we take our sample based on the loan amount with the largest number of observations and arrive at a final sample of 10,180 observations. Panel A of Table 1 lists the yearly distribution of the sample during 2000–2007. There are some differences between the numbers of observations in each year. This may be due to self-selection of disclosure of borrowing information. Panel B of Table 1 summarises the industrial distribution of the sample and shows that most observations are from the manufacturing industry. 5.2. Descriptive statistics Table 2 presents the descriptive statistics of the main variables and the univariate tests. For the entire sample, the average bank loan value is 16.58 and the median is 16.76. The average loan maturity is 6.44 years and the median value is 6.42 years. The adjusted loan interest rate varies from –0.18 to 0.37 and the actual loan interest rates are 5% higher than the basic rate, on average. About 77% of loans in the sample were issued to state-owned companies. Loans from the Big Four state-owned banks account for 55% of the total loans. The values of Size, Leverage, ROE and Tangibility are also reasonable. We divide the full sample into two groups according to the median value of Cscore to conduct univariate tests. Table 2 shows that there are higher loan amounts, longer loan maturities, and lower interest rates in the high-Cscore group than in the low-Cscore group and the difference between the two groups is statistically significant. These results imply that firms that exhibit higher accounting conservatism are more likely to obtain higher loan amounts, longer maturities, and lower interest rates, there- fore supporting our prediction. In addition, the high-Cscore group shows larger firm sizes, higher debt ratios, higher ROE values, and greater growth than the low-Cscore group. Table 1. Sample distribution. Panel A: Yearly distribution 2000 2001 2002 2003 2004 2005 2006 2007 Total N 1599 878 916 1484 1744 1404 1147 1008 10180 Percent 15.71 8.62 9.00 14.58 17.13 13.79 11.27 9.9 100 Panel B: Industrial distribution N Percent Agriculture, forestry, livestock farming, fishery 268 2.63 Mining 165 1.62 Manufacturing 6,244 61.34 Electric power, steam and hot water production and supply 591 5.81 Construction 354 3.48 Transportation and storage 314 3.08 Information technology industry 441 4.33 Wholesale and retail trade 311 3.06 Real estate 468 4.60 Social services 201 1.97 Communication and cultural industry 25 0.25 Comprehensive 798 7.84 Total 10,180 100 China Journal of Accounting Studies 209 Table 2. Descriptive statistic. All sample Low Cscore sample High Cscore sample Standard N Mean Median deviation Minimum Maximum N Mean Median N Mean Median Lnloans 10,180 16.58 16.76 1.15 14.40 18.58 5,090 16.32 16.30 5090 16.83*** 16.81*** Lnterm 9,815 6.44 6.42 0.75 5.21 7.91 4,885 6.39 6.30 4930 6.49*** 6.58*** Adjust Rate 7,788 0.05 0.02 0.14 -0.18 0.37 3,986 0.06 0.05 3802 0.04*** 0.02*** Loan Balance 10,180 19.99 19.92 1.07 18.04 21.81 5,090 19.26 19.32 5090 20.73*** 20.8*** Size 10,180 21.45 21.33 0.87 20.06 23.23 5,090 20.89 20.85 5090 22.01*** 22.00*** Leverage 10,180 0.56 0.57 0.15 0.28 0.84 5,090 0.48 0.48 5090 0.64*** 0.64*** ROE 10,180 0.04 0.07 0.12 -0.35 0.20 5,090 0.03 0.06 5090 0.06*** 0.08*** Tangibility 10,180 0.44 0.45 0.20 0.06 0.75 5,090 0.43 0.43 5090 0.45*** 0.49*** SOE 10,180 0.77 1.00 0.42 0.00 1.00 5,090 0.74 1.00 5090 0.81*** 1.00*** MTB 10,180 4.58 3.70 2.62 1.65 11.41 5,090 5.04 3.94 5090 4.12*** 3.64*** Shareholding 10,180 0.40 0.38 0.17 0.04 0.85 5,090 0.41 0.38 5090 0.40 0.39 Bank4 10,180 0.55 1.00 0.50 0.00 1.00 5,090 0.55 1.00 5090 0.54* 1.00* Note: The main variables are defined as follows. Lnloans, namely Loan amounts, is the logarithm value of loan amounts. Lnterm equals to the logarithm of the loan maturity in days. Adjust Rate, adjusted loan interest rates, equals the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic interest rate. Loan Balance is the logarithm value of the loan balance from the previous year. Size is the logarithm value of the total assets. Leverage equals the total liabilities divided by the total assets. ROE equals the net income divided by the net assets. Tangibility, namely tangible assets ratio, equals the tangible assets divided by the total assets. SOE is dummy variable, and equal to 1 if the firm is a state-owned company and 0 otherwise. MTB, refers to growth, and is the market-to-book ratio at the end of the year. Share- holding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, and equal to 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. ***, **, *denote significance at 1%, 5%, 10% levels respectively. 210 Zhao et al. Table 3 presents a correlation matrix for the main variables. Loan size is signifi- cantly positively correlated with accounting conservatism (Cscore), the loan balance from the previous year (Loan Balance), firm size (Size), return on net assets (ROE), and the tangible assets ratio (Tangibility), but negatively correlated with the adjusted interest rate (Adjust Rate) and debt ratio (Leverage). Loan maturity is significantly posi- tively correlated with accounting conservatism (Cscore), the adjusted interest rate (Adjust Rate), the loan balance from the previous year (Loan Balance), firm size (Size), return on net assets (ROE), and the tangible assets ratio (Tangibility), but negatively correlated with the debt ratio (Leverage). The interest rate is significantly negatively correlated with accounting conservatism (Cscore), the loan balance from the previous year (Loan Balance), firm size (Size), return on net assets (ROE), and the tangible assets ratio (Tangibility), but positively correlated with the debt ratio (Leverage). All of these results are consistent with our predictions. Correlation among most of control vari- ables is less than 0.5, is not very strong and therefore does not cause multicollinearity. 5.3. Multivariate regression analysis Table 4 reports the regression results of Model I to test the hypothesis. Columns 1 to 3 use the loan amounts, maturities, and interest rates, respectively, as the dependent vari- ables. The coefficient of Cscore in Column 1 is 3.24 (significant at the 1% level), with an increase of one standard deviation in a company’s Cscore contributing to a 6.69% incremental loan amount. This means that a company with higher accounting conserva- tism can obtain a larger amount of bank loans. The coefficient of Cscore in Column 2 is 2.21 (significantly positive at the 1% level), with an increase of one standard deviation in a company’s Cscore contributing to an extension of the loan maturity by 4.52%. This means that a company that exhibits higher accounting conservatism can obtain a longer loan maturity. The coefficient of Cscore in Column 3 is –0.60 (significantly negative at the 5% level), with an increase of one standard deviation in a company’s Cscore contributing to a decline of 23.8% in the adjusted loan interest rate. This means that a company that exhibits higher accounting conservatism can obtain bank loans at lower interest rates. All three results support the hypothesis, sug- gesting the higher a company’s level of accounting conservatism, the more favourable the loan policies it can enjoy. Regarding the control variables, the loan balance for the previous year (Loan Balance) is significantly negatively correlated with loan maturity (in Column 2) and significantly positively correlated with the interest rate (in Column 3). The coefficients of both Size and ROE are significantly positive at the 1% level in Columns 1 and 2, which means that larger firms with higher earnings capacities can obtain higher loan amounts with longer maturity. However, Size and ROE are significantly negative in Column 3, indicating that firms with large size and higher earnings performance can obtain bank loans with lower interest rates. The debt ratio is significantly negative in Columns 1 and 2 but significantly positive in Column 3, indicating that firms with higher debt ratios obtain lower loan amounts with shorter maturity and higher interest rates. The coefficient of Tangibility is positively significant across the three columns, indicating that firms with more tangible assets can obtain bank loans with higher amounts, longer maturities, and higher interest rates. The variable SOE is negatively significant, indicating that, compared with private firms, state-owned companies can obtain lower bank loan amounts with longer maturities. The variable MTB is positively significant in Column 2 but negative in Column 3, indicating that companies that China Journal of Accounting Studies 211 Table 3. Correlation of variables. Lnloans Lnterm Adjust Rate Cscore Loan balance Size Leverage ROE Tangibility SOE MTB Shareholding Bank4 Lnloans 1.00 Lnterm 0.22*** 1.00 Adjust Rate –0.22*** 0.08*** 1.00 Cscore 0.26*** 0.10*** –0.06*** 1.00 Loan Balance 0.25*** 0.10*** –0.04*** 0.79*** 1.00 Size 0.38*** 0.22*** –0.16*** 0.75*** 0.76*** 1.00 Leverage –0.03*** –0.12*** 0.14*** 0.59*** 0.47*** 0.10*** 1.00 ROE 0.15*** 0.16*** –0.10*** 0.16*** 0.07*** 0.27*** –0.19*** 1.00 Tangibility 0.07*** 0.16*** –0.06*** 0.11*** 0.19*** 0.19*** –0.06*** 0.06*** 1.00 SOE 0.04*** 0.15*** –0.10*** 0.09*** 0.09*** 0.22*** –0.18*** 0.06*** 0.22*** 1.00 MTB –0.18*** –0.06*** 0.24*** –0.26*** –0.12*** –0.30*** 0.35*** –0.15*** –0.20*** –0.19*** 1.00 Shareholding 0.11*** 0.10*** –0.16*** –0.01 –0.06*** 0.19*** –0.29*** 0.13*** 0.09*** 0.33*** –0.17*** 1.00 Bank4 0.01 0.02** –0.05*** 0.00 0.02** 0.04*** –0.05*** 0.08*** 0.07*** 0.06*** 0.01 0.09*** 1.00 Note: The main variables are defined as follows. Lnloans, namely Loan amounts, is the logarithm value of loan amounts. Lnterm equals the logarithm of the loan maturity in days. Adjust Rate, adjusted loan interest rates, is equal to the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic inter- est rate. Cscore is the proxy for accounting conservatism based on Khan and Watts (2009). Loan Balance is the logarithm value of the loan balance from the previous year. Size is the logarithm value of the total assets. Leverage equals the total liabilities divided by the total assets. ROE equals the net income divided by the net assets. Tangibility, namely tangible assets ratio, equals the tangible assets divided by the total assets. SOE is a dummy variable, and is equal to 1 if the firm is a state-owned company and 0 otherwise. MTB, which refers to growth, is the market-to-book ratio at the end of the year. Shareholding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, and is equal to 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. ***, **, *denote significance at 1%, 5%, 10% levels respectively. 212 Zhao et al. Table 4. Basic regression of loan contract terms on conservatism. (1) (2) (3) Lnloans Lnterm Adjust Rate Conservatism Cscore 3.24*** 2.21*** –0.60** (3.15) (2.78) (–2.01) Loan characteristics Loan Balance –0.01 –0.07*** 0.02*** (–0.61) (–5.00) (3.13) Lnloans –0.04*** (–15.18) Lnterm 0.06*** (16.55) Firm characteristics Size 0.44*** 0.20*** –0.03*** (14.73) (9.72) (–3.88) Leverage –0.85*** –0.39*** 0.23*** (–6.74) (–4.47) (6.76) ROE 0.29*** 0.39*** –0.13*** (2.86) (5.96) (–4.99) Tangibility 0.09 0.54*** 0.01 (1.39) (12.84) (0.46) SOE –0.15*** 0.09*** –0.02*** (–5.69) (4.93) (–3.39) MTB –0.01 0.06*** –0.05*** (–0.33) (2.79) (–5.33) Shareholding 0.48*** –0.15*** –0.02 (6.52) (–2.92) (–1.27) Bank4 0.02 –0.02 –0.01** (0.83) (–1.56) (–2.01) Constant 7.82*** 3.74*** 1.20*** (15.40) (9.53) (8.09) Year Dummy Yes Yes Yes Industry Dummy Yes Yes Yes Observations 10,180 9,815 7,788 Adj-R 0.18 0.15 0.19 Note: The main variables are defined as follows. Lnloans, namely Loan amounts is the logarithm value of loan amounts. Lnterm is equal to the logarithm of the loan maturity in days. Adjust Rate, adjusted loan inter- est rates, equals the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic interest rate. Cscore is the proxy for accounting conservatism based on Khan and Watts (2009). Loan Balance is the logarithm value of the loan balance from the previous year. Size is the log- arithm value of the total assets. Leverage is the total liabilities divided by the total assets. ROE denotes the net income divided by the net assets. Tangibility, namely tangible assets ratio, is equal to the tangible assets divided by the total assets. SOE is a dummy variable, equal to 1 if the firm is a state-owned company and 0 otherwise. MTB, refers to growth, and is the market-to-book ratio at the end of the year. Shareholding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, and equals 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. ***, **, *denote significance at 1%, 5%, 10% levels respectively. experience higher growth can obtain higher bank loan amounts with longer maturities and lower interest rates. The term Shareholding is positively correlated with loan size, but negatively correlated with loan maturity and interest rate, indicating that firms whose largest shareholders have higher shareholdings can obtain higher bank loan amounts with shorter maturities and lower interest rates. China Journal of Accounting Studies 213 6. Robustness tests 6.1. Discussion of the endogeneity problem The preceding results may be affected by the endogeneity problem. For example, the higher a company’s conservatism, the higher the bank loan amounts it can obtain and the more likely those amounts will further increase its conservatism. To mitigate a potential endogeneity problem, we select the previous year’s accounting conservatism as an explanatory variable to conduct the tests. As shown in Table 5, the results are basically the same. In addition, we use the amount of every new loan obtained in each year rather than the loan balance at each year-end. If we were to use the loan balances, the levels of conservatism would likely reflect the bank’s supervisory role. However, we use the new loans obtained each year, which reflect the bank’s risk-averse decisions ex ante. These features decrease the influence of endogeneity on the results. 6.2. Controlling for self-selection using the Heckman two-stage approach Our data are from the notes of A-share listed companies’ annual reports from 2000 to 2007. However, only about 20% of the companies disclosed detailed information on their loan covenants, which introduces a self-selection problem. To alleviate the problem, we implement the Heckman two-stage method and conduct additional tests. Following Doyle, Ge, and McVay (2007), factors that affect the voluntary disclosure of accounting information include firm size (Size), debt ratio (Leverage), whether a loss is being reported (Loss), sales fluctuations (Sale volatility), cash flow volatility (Cash Vol- atility), the business turnover ratio (Operating Cycle), firm age (Age), and the largest shareholder stake (Shareholding). In addition, as the first stage of the model, we control for the type of the ultimate controller to construct a model of the determinants of information disclosure level. From the first-stage regression, we estimate the likelihood of corporate disclosure of bank information to calculate the inverse Mills ratio (IMR) in Model I to re-examine our hypothesis: Disclose ¼ a þ a Size þ a Leverage þ a Loss þ a SaleVolatility þ a CashVolatility 0 1 2 3 4 5 þ a OperatingCycle þ a Age þ a Shareholding þ a SOE þ e ðModel IIÞ 6 7 8 9 Panel A of Table 6 presents the results of the firm-level bank loan disclosure regres- sion. We find that state-owned companies with higher debt ratios are more likely to dis- close their bank loan information. However, firms with greater sales volatility and longer operating cycles are less likely to disclose their bank loan information. We place the IMR calculated from Model II into Model I. Panel B presents the results of the effect of accounting conservatism on bank loans. The IMR in some columns is signifi- cant. However, the results of accounting conservatism are basically unchanged. Taken together, these results indicate that the self-selecting disclosure of bank loan informa- tion does not significantly affect our main findings. 6.3. Different effects of accounting conservatism on various loan maturities We perform regressions using the sub-samples of short- and long-maturity loans, that is, loans with maturities of less than one year and of more than one year, respectively. The results are shown in Table 7. In the long-maturity loan regression, accounting conserva- tism has a significant influence on the loan amount and interest rate, but this influence does not hold in the short-maturity loan regression. Accounting conservatism as a disciplining mechanism is likely to be more effective over a longer period. 214 Zhao et al. Table 5. Accounting conservatism in the previous year and bank loans. (1) (2) (3) Lnloans Lnterm Adjust Rate Conservatism Cscore_lag 3.19*** 0.58** -0.52*** (3.81) (2.01) (-2.82) Loan characteristics Loan Balance 0.25*** 1.24*** 0.01 (11.01) (145.08) (1.02) Lnloans –0.04*** (–14.94) Lnterm 0.05*** (7.61) Firm characteristics Size 0.37*** 0.02** –0.01** (16.47) (2.36) (–2.31) Leverage –0.68*** –0.14*** 0.25*** (–6.43) (–3.81) (9.22) ROE 0.26** 0.07** –0.14*** (2.45) (2.06) (–5.26) Tangibility 0.30*** 0.27*** –0.01 (4.45) (10.43) (–0.36) SOE –0.15*** 0.03*** –0.02*** (–5.42) (2.84) (–2.93) MTB 0.08** 0.02* –0.06*** (2.20) (1.79) (–6.49) Shareholding 0.50*** –0.03 –0.04** (6.74) (–0.91) (–2.37) Bank4 –0.01 –0.04*** –0.01 (–0.38) (–4.84) (–1.50) Constant 8.09*** 5.24*** 1.20*** (15.93) (28.87) (9.88) Year Dummy Yes Yes Yes Industry Dummy Yes Yes Yes Observations 9,805 9,805 7,778 Adj-R 0.18 0.72 0.18 Note: The main variables are defined as follows. Lnloans, namely Loan amounts, is the logarithm value of loan amounts. Lnterm is the logarithm of the loan maturity in days. Adjust Rate, adjusted loan interest rates, is equal to the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic interest rate. Cscore_lag is the proxy for accounting conservatism of previous year based on Khan and Watts (2009). Loan Balance is the logarithm value of the loan balance from the previous year. Size is the logarithm value of the total assets. Leverage equals the total liabilities divided by the total assets. ROE equals the net income divided by the net assets. Tangibility, namely tangible assets ratio, is equal to the tangible assets divided by the total assets. SOE is a dummy variable, equal to 1 if the firm is a state-owned company and 0 otherwise. MTB refers to growth and is the market-to-book ratio at the end of the year. Shareholding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, which is equal to 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. ***, **, *denote significance at 1%, 5%, 10% levels respectively. 6.4. Alternative measurement of accounting conservatism To ensure the robustness of the results, we use another method (to substitute for Cscore) to calculate accounting conservatism. Zhang (2008) points out that when the criteria for bad news recognition are low, corporate profits usually decrease and have negative skewness. We can measure accounting conservatism by dividing a company’s time series skewness of net profits by the skewness of net cash flow. To ensure the same direction as Cscore, we follow Zhang’s(2008) method and multiply the China Journal of Accounting Studies 215 Table 6. Bank loans and accounting conservatism. Panel A: Factors affecting bank loan disclosure Disclose Size 0.01 (0.62) Leverage 0.45*** (5.34) Loss –0.03 (–0.66) Sale Volatility –0.43*** (–3.89) Cash Volatility 0.25 (0.61) Operating Cycle –0.10*** (–7.01) Age 0.00 (0.36) Shareholding 0.00 (1.47) SOE 0.16*** (4.52) Constant –1.16*** (–3.17) Year Dummy Yes Industry Dummy Yes Observations 9,990 Pseudo R-squared 0.13 Panel B: Bank loans and accounting conservatism (1) (2) (3) Lnloans Lnterm Adjust Rate Conservatism Cscore 3.27*** 2.12*** –0.67** (3.16) (2.66) (–2.09) Loan characteristics Loan Balance –0.02 –0.08*** 0.01*** (–0.95) (–5.74) (2.92) Lnloans –0.04*** (–13.27) Lnterm 0.06*** (14.81) Firm characteristics Size 0.41*** 0.22*** –0.02*** (13.88) (11.06) (–2.83) Leverage –0.54*** –0.64*** 0.19*** (–4.05) (–6.91) (5.25) ROE 0.27*** 0.39*** –0.12*** (2.61) (5.95) (–4.55) Tangibility 0.17** 0.66*** 0.02 (2.49) (14.66) (0.89) SOE –0.06** –0.02 –0.04*** (–2.08) (–0.85) (–5.01) MTB –0.06 0.09*** –0.04*** (–1.61) (3.96) (–4.59) (Continued) 216 Zhao et al. Table 6. (Continued). Panel B: Bank loans and accounting conservatism (1) (2) (3) Lnloans Lnterm Adjust Rate Shareholding 0.47*** –0.18*** –0.03 (6.44) (–3.55) (–1.54) Bank4 0.01 –0.03** –0.01** (0.63) (–2.08) (–2.03) IMR 0.88*** –0.75*** –0.14*** (7.60) (–9.39) (–4.84) Constant 6.94*** 4.64*** 1.33*** (13.26) (11.60) (8.15) Year Dummy Yes Yes Yes Industry Dummy Yes Yes Yes Observations 10,180 9,815 7,788 Adj-R 0.19 0.17 0.19 Note: The main variables are defined as follows. Lnloans, namely Loan amounts, is the logarithm value of loan amounts. Lnterm is equal to the logarithm of the loan maturity in days. Adjust Rate, adjusted loan inter- est rates, equals the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic interest rate. Cscore is the proxy for accounting conservatism based on Khan and Watts (2009). Loan Balance is the logarithm value of the loan balance from the previous year. Size is the log- arithm value of the total assets. Leverage denotes the total liabilities divided by the total assets. ROE is the net income divided by the net assets. Tangibility, namely tangible assets ratio, equals the tangible assets divided by the total assets. SOE is a dummy variable, which is equal to 1 if the firm is a state-owned com- pany and 0 otherwise. MTB refers to growth and is the market-to-book ratio at the end of the year. Sharehold- ing is the shareholding of the biggest stockholder. Bank4 is a dummy variable and is equal to 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. Loss is a dummy variable, equal to 1 if the firm reports loss and 0 otherwise. Sale volatility, is the standard deviation of the ratio of the sale to the total assets. Oper- ating Cycle is the log of [(Sales)/(Average Accounts Receivable) + (Cost of Goods Sold)/Average Inventory)]. Age is the number of years from the establishing year to the current year. IMR, the Inverse Mills ratio, is cal- culated following the Heckman approach. See Table 2 for the definition of the other variables. ***, **, *denote significance at 1%, 5%, 10% levels respectively. calculated result by –1 to obtain Consv_negskew. The larger this value, the more con- servative the accounting figures. We use Consv_negskew to perform the regression and the results are shown in Table 8. Table 8 presents the basic regression of accounting conservatism on a bank loan, in which Consv_negskew is significantly positively correlated with loan amount and maturity, but significantly negatively correlated with the adjusted loan rate. These results are consistent with our previous findings. 6.5. Firm-level test We use loan-level data in the main regression. Here, we consider each firm’s loan information collectively by year and perform regressions on the basic model using firm-year-level data. The results are basically the same. For brevity, the results are not reported here. 6.6. Further analysis Different institutional environments can have different effects on bank loan contracts (Djankov, McLiesh, & Shleifer, 2007; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1998). These institutional environments include legalisation, preferential policies, and China Journal of Accounting Studies 217 Table 7. The impact of conservatism on long- and short-maturity loans. Long-term Loan Short-term Loan (1) (2) (3) (4) Lnloans Adjust Rate Lnloans Adjust Rate Conservatism Cscore 6.28*** –0.84* 1.10 0.36 (2.99) (–1.92) (0.75) (0.75) Loan characteristics Loan Balance 0.01 –0.00 –0.01 0.05*** (0.37) (–0.13) (–0.37) (6.27) Lnloans –0.03*** –0.05*** (–8.13) (–12.54) Lnterm 0.03*** 0.11*** (4.76) (7.30) Firm characteristics Size 0.39*** –0.02* 0.44*** –0.07*** (7.91) (–1.94) (10.00) (–5.01) Leverage –0.82*** 0.26*** –0.77*** 0.09* (–3.47) (5.40) (–4.53) (1.89) ROE 0.59*** –0.22*** –0.09 –0.07* (3.18) (–5.64) (–0.69) (–1.68) Tangibility 0.08 0.01 –0.05 –0.05* (0.90) (0.42) (–0.48) (–1.70) SOE –0.14*** –0.04*** –0.21*** –0.03*** (–3.19) (–4.23) (–5.86) (–3.52) MTB 0.02 –0.04*** –0.05 –0.03** (0.32) (–3.55) (–0.93) (–2.20) Shareholding 0.29*** –0.02 0.71*** –0.06* (2.80) (–1.13) (6.40) (–1.83) Bank4 –0.08*** –0.01* 0.07** –0.01 (–2.74) (–1.65) (2.18) (–1.25) Constant 8.52*** 1.26*** 7.57*** 1.32*** (8.58) (5.83) (10.11) (5.18) Year Dummy Yes Yes Yes Yes Industry Dummy Yes Yes Yes Yes Observations 5,334 4,630 4,481 3,158 Adj-R 0.18 0.13 0.17 0.39 Note: The main variables are defined as follows. Lnloans, namely Loan amounts, is the logarithm value of loan amounts. Lnterm is equal to the logarithm of the loan maturity in days. Adjust Rate, adjusted loan interest rates, is equal to the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic interest rate. Cscore is the proxy for accounting conservatism based on Khan and Watts (2009). Loan Balance is the logarithm value of the loan balance from the previous year. Size is the logarithm value of the total assets. Leverage equals the total liabilities divided by the total assets. ROE equals the net income divided by the net assets. Tangibility, namely tangible assets ratio, is equal to the tangible assets divided by the total assets. SOE is a dummy variable that equals 1 if the firm is a state-owned company and 0 otherwise. MTB refers to growth and is the market-to-book ratio at the end of the year. Shareholding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, which equals 1 if the lending bank is a Big Four state-owned bank and 0 otherwise. ***, **, *denote significance at 1%, 5%, 10% levels respectively. government interference in the market. Based on China’s unique institutional background, this paper investigates how institutional environments influence the rela- tion between accounting conservatism and bank loans with regards to the laws, govern- ment preferential policies, and levels of government intervention in different regions. A region’s legalisation level can be measured based on its development of market intermediary organisations, the protection of the legitimate rights and interests of 218 Zhao et al. Table 8. Bank loans and accounting conservatism. (1) (2) (3) Lnloans Lnterm Adjust Rate Conservatism Consv_negskew 0.01*** 0.01*** –0.00*** (4.43) (8.29) (–8.43) Loan characteristics Loan Balance 0.00 –0.06*** 0.02*** (0.04) (–4.43) (3.47) Lnloans –0.04*** (–13.61) Lnterm 0.06*** (16.18) Firm Characteristics Size 0.47*** 0.24*** –0.04*** (19.58) (16.00) (–7.19) Leverage –0.41*** –0.20*** 0.12*** (–3.88) (–2.74) (4.66) ROE 0.40*** 0.52*** –0.19*** (4.34) (8.61) (–7.26) Tangibility –0.16*** 0.57*** –0.00 (–2.62) (13.35) (–0.04) SOE –0.20*** 0.07*** –0.02*** (–7.11) (3.59) (–2.71) MTB –0.01* 0.01 0.00*** (–1.93) (1.42) (2.90) Shareholding 0.32*** –0.16*** –0.01 (4.43) (–3.17) (–0.50) Bank4 –0.00 –0.03* –0.01** (–0.13) (–1.84) (–2.42) Constant 6.65*** 2.55*** 1.44*** (23.32) (12.52) (17.96) Year Dummy Yes Yes Yes Industry Dummy Yes Yes Yes Observations 10,146 9,781 7,754 Adj-R 0.17 0.16 0.19 Note: The main variables are defined as follows. Lnloans, namely loan amounts, is the logarithm value of loan amounts. Lnterm equals the logarithm of the loan maturity in days. Adjust Rate, adjusted loan interest rates, is equal to the difference between the interest rate for every loan and the basic interest rate of the cur- rent year, divided by the basic interest rate. Consv_negskew is the proxy for accounting conservatism. Loan Balance is the logarithm value of the loan balance from the previous year. Size is the logarithm value of the total assets. Leverage equals the total liabilities divided by the total assets. ROE equals the net income divided by the net assets. Tangibility, namely tangible assets ratio, equals the tangible assets divided by the total assets. SOE is a dummy variable and is equal to 1 if the firm is a state-owned company and 0 otherwise. MTB refers to growth and is the market-to-book ratio at the end of the year. Shareholding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, which equals 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. ***, **, *denote significance at 1%, 5%, 10% levels. producers, and the protection of intellectual property rights. A higher level of legalisa- tion usually contributes to the better development of market intermediary organisations and market order, greater law enforcement efficiency, and effective intellectual property protection. Banks can credibly rely on the financial information of companies in these areas and conservative accounting information can help companies gain the trust of banks, since laws can improve the effectiveness of creditor contract implementation (Qian & Cao, 2013; Zhang & Wang, 2012). Laws punish default actions and decrease China Journal of Accounting Studies 219 the amount of risk for banks, which then have the option to offer more company loans. Based on a company’s conservative accounting information, banks can rely on the law to fulfil a contract and decrease the amount of risk in the case of company fraud. Jappelli, Pagano, and Bianco (2005) find that stronger creditor protection from the legal and judicial systems helps to increase the likelihood that a bank will lend to a com- pany. Therefore, the higher the legal system benchmark, the more often a bank will use accounting information to judge a debtor’s profitability and solvency and the stronger the effect of conservative accounting information on bank loan contracts. Preferential government policies in a region imply that the region’s banks have more abundant credit resources. Companies that exhibit high accounting conservatism may be more likely to receive large loan amounts from banks, with lower interest rates. This paper uses the method of Demruger et al. (2002) to measure the level of govern- ment preferential policies (i.e., those granted by the central government) in China dur- ing the period 1978 to 1998, when the country’s special economic zones were established. This method is appropriate because preferential policies have existed in these areas for many years and it reflects China’s stable institutional environment. Government interference in the market is another main factor influencing bank loan behaviour. As a precondition, a judicial department that is supposed to have indepen- dent judicial power should not be attached to the government and should not judge cases following government policies and official opinions. However, China’s judicial authorities are restricted by local governments and their judicial systems lack indepen- dence. Government departments or officials often intervene in the judicial departments, making regulations difficult to implement (Allen, Qian, & Qian, 2005). If a local gov- ernment’s intervention in a market is relatively limited, a bank will rely more on market information when making lending decisions. Hence, the more conservative a company’s accounting information, the more likely it can obtain the trust of banks. In summary, banks are more likely to make lending decisions based on the account- ing information of companies in areas with better legal environments, more preferential policies, and less government intervention. In such situations, accounting conservatism can play a more important role. We interact the legal protection (Fan et al., 2009), gov- ernment preferential policies (Demruger et al., 2002) and government-market relation indices (Fan et al., 2009) with accounting conservatism to examine the effects of accounting conservatism on loans in different institutional environments. Table 9 presents the regression results of the interaction between institutional envi- ronments and accounting conservatism. Columns 1 to 3 present the results of cross- multiplying accounting conservatism (Cscore) with the legal protection index (Legal), that is, Cscore_legal. Columns 4 to 6 present the results of cross-multiplying account- ing conservatism (Cscore) with the preferential index (Deregulation), that is, Cscore_deregulation. Columns 7 to 9 present the results of cross-multiplying account- ing conservatism (Cscore) with the government and market index (Government and Market), that is, Cscore_Gov_Market. First, the interaction coefficients for Cscore_legal are significantly positive in Columns 1 and 2. In Column 3, the coefficient is signifi- cantly negative, indicating that in regions with stronger legal protection, higher levels of accounting conservatism are associated with higher loan amounts, longer loan matu- rities, and lower loan rates. Second, the coefficients for Cscore_deregulation in Col- umns 4 and 5 are significantly positive and the coefficient in Column 3 is significantly negative, indicating that companies that exhibit stronger accounting conservatism tend to obtain higher loan amounts with longer maturities and lower interest rates in regions with more preferential policies. Finally, the coefficients for Cscore_Gov_Market in 220 Zhao et al. Table 9. The regression of loan contract terms on the interaction between institutional environments and conservatism. (1) (2) (3) (4) (5) (6) (7) (8) (9) Lnloans Lnterm Adjust Rate Lnloans Lnterm Adjust Rate Lnloans Lnterm Adjust Rate Conservatism and interaction Cscore*Institutions 1.03*** 0.75*** –0.16* 1.11*** 1.09*** –0.59*** 0.70*** 0.36** –0.13** (2.70) (2.65) (–1.80) (2.03) (2.73) (–4.54) (3.09) (2.26) (–2.31) Cscore –1.39 0.65 0.67 2.89* 3.52*** 0.38 –1.50 1.76 0.74 (–0.55) (0.35) (1.12) (1.82) (3.03) (0.90) (–0.63) (1.09) (1.18) Institutions Legal Protection –0.01 –0.03*** 0.01* (–0.80) (–2.90) (1.68) Deregulation –0.04* –0.03* 0.00 (–1.94) (–1.95) (0.47) Government and Market –0.02** –0.02*** –0.01*** (–2.56) (–2.80) (–2.89) Loan characteristics Loan Balance –0.02 –0.08*** 0.01*** –0.02 –0.08*** 0.02*** –0.01 –0.08*** 0.02*** (–0.77) (–5.94) (3.22) (–0.72) (–5.95) (3.62) (–0.64) (–5.84) (3.35) Lnloans –0.04*** –0.04*** –0.04*** (–13.84) (–13.35) (–13.46) Lnterm 0.05*** 0.05*** 0.05*** (13.61) (13.70) (13.41) Firm characteristics Size 0.42*** 0.18*** –0.04*** 0.42*** 0.17*** –0.04*** 0.43*** 0.18*** –0.04*** (13.43) (7.95) (–4.03) (13.44) (7.84) (–4.10) (13.47) (7.94) (–4.01) Leverage –1.03*** –0.64*** 0.23*** –1.05*** –0.65*** 0.24*** –1.02*** –0.63*** 0.23*** (–5.24) (–4.41) (4.76) (–5.35) (–4.45) (4.98) (–5.17) (–4.36) (4.80) ROE 0.39*** 0.35*** –0.15*** 0.40*** 0.35*** –0.16*** 0.40*** 0.36*** –0.15*** (3.98) (5.54) (–6.11) (4.08) (5.57) (–6.16) (4.06) (5.70) (–5.90) Tangibility 0.16** 0.59*** –0.01 0.13** 0.61*** –0.03** 0.12* 0.58*** –0.02* (2.43) (13.22) (–0.42) (2.03) (13.48) (–1.99) (1.89) (13.32) (–1.70) SOE –0.16*** 0.07*** –0.04*** –0.14*** 0.07*** –0.04*** –0.14*** 0.07*** –0.03*** (–5.48) (3.72) (–5.62) (–5.06) (3.64) (–5.54) (–5.07) (3.82) (–4.75) China Journal of Accounting Studies 221 MTB 0.01* 0.01** –0.00** 0.01* 0.01** –0.01*** 0.01 0.01** –0.00** (1.71) (2.50) (–2.11) (1.76) (2.56) (–2.71) (1.44) (2.19) (–2.03) Shareholding 0.46*** –0.16*** –0.03** 0.46*** –0.17*** –0.03* 0.46*** –0.16*** –0.03* (6.25) (–3.24) (–1.98) (6.27) (–3.34) (–1.91) (6.28) (–3.31) (–1.92) Bank4 0.02 –0.03** –0.01** 0.02 –0.03** –0.01** 0.02 –0.03** –0.01*** (0.93) (–2.09) (–2.16) (0.81) (–2.00) (–2.36) (0.96) (–1.98) (–2.62) Constant 8.07*** 4.73*** 1.35*** 8.04*** 4.64*** 1.34*** 8.10*** 4.68*** 1.40*** (13.86) (10.19) (7.87) (13.91) (10.09) (7.82) (14.00) (10.17) (8.17) Year Dummy Yes Yes Yes Yes Yes Yes Yes Yes Yes Industry Dummy Yes Yes Yes Yes Yes Yes Yes Yes Yes Observations 10,180 9,815 7,788 10,180 9,815 7,788 10,180 9,815 7,788 Adj-R 0.19 0.17 0.27 0.19 0.17 0.28 0.19 0.16 0.28 Note: The main variables are defined as follows. Lnloans, namely loan amounts, is the logarithm value of loan amounts. Lnterm equals the logarithm of the loan maturity in days. Adjust Rate, adjusted loan interest rates, is equal to the difference between the interest rate for every loan and the basic interest rate of the current year, divided by the basic inter- est rate. Cscore is the proxy for accounting conservatism based on Khan and Watts (2009). Loan Balance is the logarithm value of the loan balance from the previous year. Size is the logarithm value of the total assets. Leverage equals to the total liabilities divided by the total assets. ROE equals the net income divided by the net assets. Tangibility, namely tangible assets ratio, is equal to the tangible assets divided by the total assets. SOE is a dummy variable, which is equal to 1 if the firm is a state-owned company and 0 otherwise. MTB refers to growth and is the market-to-book ratio at the end of the year. Shareholding is the shareholding of the biggest stockholder. Bank4 is a dummy variable, which is equal to 1 if the lender bank is a Big Four state-owned bank and 0 otherwise. Legal is the legal protection index, taken from a study by Fan et al. (2009). Deregulation is the government preferential policy index, which is taken from a study by Demruger et al. (2002). Government and Market is the market index, taken form a study by Fan et al. (2009). ***, **, *denote significance at 1%, 5%, 10% levels respectively. 222 Zhao et al. Table 10. Distribution and correlation of annual accounting conservatism. Panel A: Distribution of Cscore by year 2000 2001 2002 2003 2004 2005 2006 2007 Mean 0.02 0.02 0.03 0.03 0.04 0.04 0.04 0.04 Median 0.02 0.02 0.03 0.03 0.04 0.04 0.04 0.04 Standard deviation 0.02 0.02 0.02 0.02 0.02 0.03 0.02 0.02 Panel B: Correlation of current- and previous-year Cscore values Cscore Cscore_lag Cscore 1.000 Cscore_lag 0.871*** 1.000 Note: Cscore is the proxy for accounting conservatism based on Khan and Watts (2009); Cscore_lag is the proxy for accounting conservatism of previous year based on Khan and Watts (2009); ***, **, *denote signif- icance at 1%, 5%, 10% levels respectively. Columns 7 and 8 are significantly positive, indicating a similar relation to that men- tioned previously, in which government intervention in the market is limited. These results show that in regions with stronger legal protection, more government preferen- tial policies, and less government intervention, companies that exhibit high accounting conservatism can obtain more preferential loans. The results that control for the other variables are consistent with the basic regression results. 6.7. The sustainability of conservatism A firm increases its accounting conservatism to obtain more preferential loans. When a firm obtains such a loan, does it cease to increase its accounting conservatism? Regard- less of whether the loan has a long or short maturity, firms and banks engage in multi- ple games. This implies that if a firm breaks its promise with a bank, the firm may not be able to obtain a preferential loan or any new loans in the future. Thus, in terms of market efficiency, rational managers avoid such decisions. Instead, they may maintain high accounting conservatism over a long period. As shown in Table 10, we analyse the annual accounting conservatism distribution and correlation and find the accounting conservatism of each year to be relatively stable, with no significant changes. Further- more, the current and previous year’s Cscore values exhibit a strong correlation, with a coefficient of 0.871. These results indicate that corporate accounting conservatism is relatively stable in our sample period. 7. Conclusion and limitations Accounting conservatism plays an important role in contracts. In bank loan covenants, an appropriate level of accounting conservatism can decrease credit risk. Correspond- ingly, banks may grant firms preferential treatment. However, few studies to date have examined this issue in China. Those studies that have examined this issue, have focused on only one particular aspect of loan contracts, with a lack of comprehensive studies. We rely on the hand-collected bank loan information of A-share listed companies in China from 2000 to 2007 to conduct a more comprehensive empirical study. Using the Cscore method proposed by Khan and Watts (2009) to measure accounting China Journal of Accounting Studies 223 conservatism, we examine the influence of accounting conservatism on loan amounts, maturities, and interest rates and discover that when a firm exhibits stronger accounting conservatism, it obtains higher loan amounts, longer maturities, and lower interest rates. We also analyse the effects of institutional environments. The results show that better institutional environments enhance the relation between accounting conservatism and bank loans. This paper is significant for the following reasons. First, it complements the litera- ture on the effect of accounting conservatism on loan amounts and maturity. Second, our analysis of different institutional environments helps us understand how accounting conservatism affects loan contracts under different constraints. In addition, these unique data improve the accuracy of the findings. This paper also has some policy implica- tions. The findings show that accounting conservatism can affect interest rates and that this effect is different within different institutional environments. This is especially true in the current Chinese context, in which lending rates are regulated. As interest rates are increasingly deregulated, the level of accounting conservatism becomes more important and its role within different institutional environments may be further widened. The effect of financial reform is likely to be weak in poor institutional envi- ronments. Therefore, it is especially important for companies to improve the quality of their accounting information. This paper has the following limitations. First, accounting conservatism encompasses both conditional and unconditional conservatism. Beaver and Ryan (2005) provide the specificdefinitions of unconditional and conditional conservatism. Unconditional conservatism is ex ante or news independent, which means that aspects of the accounting process determined at the inception of assets and liabilities yield expected unrecorded goodwill. Examples of unconditional conservatism include imme- diate expensing of most internally developed intangibles, depreciation of property, plant, and equipment that is more accelerated than economic depreciation, and histori- cal cost accounting for positive net present value projects (Beaver & Ryan, 2005). Con- ditional conservatism is ex post or news dependent, which means that book value is written down under sufficiently adverse circumstances but not written up under favour- able circumstances. Basu (1997) gives a similar definition of conditional conservatism and defines it as the asymmetric verification threshold for gains versus losses: the veri- fication threshold for gains is higher. However, we focus only on conditional conserva- tism. We mainly use Cscore to measure accounting conservatism and, until now, no study has explored the applicability of this measurement in China. On the topic of mea- suring conservatism, Basu (1997) points out that this method is affected by the validity of a company’s information. Piotroski and Wong (2011) have shown that, although Chinese companies have improved the validity of their information, they lag far behind companies in the United States and other developed countries. In addition, we do not pay significant attention to industry effects when performing our cross-sectional regres- sions and this may introduce bias to Cscore. Second, although we test the robustness of Cscore for loan covenants, this does not entirely ensure a causal relation between conservatism and loans. In this regard, other methods should be used to test the rela- tion. Finally, we do not collect information on loan types and cannot control for such a factor in the model. It would be meaningful to explore the effect of accounting conser- vatism on bank loan covenants in different macroeconomic contexts, such as monetary policy, fiscal policy, and changes in refinancing policy. We plan to address these limitations in future research. 224 Zhao et al. Acknowledgements This research is supported by grants from the National Natural Science Foundation of China (No. 71102124), the Beijing Municipal Commission of Education ‘Joint Construction Project’, and the Beijing Municipal Commission of Education ‘Pilot Reform of Accounting Discipline Clustering’. Notes 1. Graham et al. (2008) show that, over the past decade, about $780 billion in new debt securi- ties were issued in the US market, compared with only $2 billion in new equity securities. About 54% of the debt issues were bank loans. 2. Ahmed et al. (2002) and Zhang (2008) provide early evidence of this issue; however, they focus on the influence of accounting conservatism on only loan interest rates and do not ver- ify its influence on loan amounts or loan maturity. 3. In China, lending rates are regulated by People’s Bank of China, which offers a benchmark lending rate and upper and lower floating percentages (e.g., 10%). The lending rates of com- mercial bank loans should be in this range. 4. Their paper provides the measurement for conditional conservatism, like Basu (1997). However, the only difference between Khan and Watts (2009) and Basu (1997) is that the former’s measurement is for firm-year level, which is necessary for this paper. 5. The results of this paper show that accounting conservatism also has a significant impact on lending rates, emphasising the importance of accounting conservatism. 6. See Berger et al. (2009), Lin and Zhang (2009), and Luo, Zhang, and Zhu (2011) for addi- tional information on the institutional background of the Chinese banking system and bank loans. 7. Regression results that use the raw interest rate are similar. 8. China’s five major economic zones were established during this period. Shenzhen, Zhuhai, Shantou, and Xiamen were established as special economic zones in 1980. Hainan was estab- lished as a special economic zone in 1988. Demruger et al. (2002) provide further details and consider the county-level city as a benchmark for measuring special economic zones. References Ahmed, A., Billings, B., Morton, R., & Harris, M. (2002). The role of accounting conservatism in mitigating bondholder-shareholder conflict over dividend policy and in reducing debt cost. The Accounting Review, 77, 867–890. Ahmed, A.S., Morton, R.M., & Schaefer, T.F. (2000). Accounting conservatism and the valuation of accounting numbers: Evidence on the Feltham-Ohlson (1996) model. Journal of Account- ing, Auditing & Finance, 15, 271–292. Alicia, G., Sergio, G., & Daniel, S. (2006). China’s banking reform: An assessment of its evolu- tion and possible impact. CESifo Economics Studies, 52, 304–363. Allen, F., Qian, J., & Qian, M.J. (2005). Law, finance, and economic growth in China. Journal of Financial Economics, 77(1), 57–116. Allen, F., & Carletti, E. (2008). The roles of banks in financial systems. Oxford Handbook of Banking. Allen, F., Qian, J.Q.J., Qian, M., & Zhao, M. (2009). A review of China’s financial system and initiatives for the future. China’s Emerging Financial Markets, 3, 72. Ball, R., Kothari, S.P., & Robin, A. (2000). The effect of international institutional factors on properties of accounting earnings. Journal of Accounting and Economics, 29(1), 1–51. Ball, R., Robin, A., & Wu, J.S. (2003). Incentives versus standards: properties of accounting income in four East Asian countries. Journal of Accounting & Economics, 36(1-3), 235–270. Ball, R., & Shivakumar, L. (2005). Earnings quality in UK private firms: Comparative loss recog- nition timeliness. Journal of Accounting and Economics, 39(1), 83–128. Ball, R., Robin, A., & Sadka, G. (2008). Is financial reporting shaped by equity markets or by debt markets? An international study of timeliness and conservatism. Review of Accounting Studies, 13, 168–205. Basu, S. (1997). The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics, 24(1), 3–37. China Journal of Accounting Studies 225 Beaver, W., & Ryan, S. (2000). Biases and lags in book value and their effects on the ability of the book-to-market ratio to predict book return on equity. Journal of Accounting Research, 38, 127–148. Beaver, W.H., & Ryan, S.G. (2005). Conditional and unconditional conservatism: Concepts and modeling. Review of Accounting Studies, 10, 269–309. Berger, A.N., Hasan, I., & Zhou, M. (2009). Bank ownership and efficiency in China: What will happen in the world’s largest nation? Journal of Banking & Finance, 33(1), 113–130. Bharath, S.T., Dahiya, S., Saunders, A., & Srinivasan, A. (2011). Lending relationships and loan contract terms. Review of Financial Studies, 24, 1141–1203. Bushman, R., Smith, A., & Piotroski, J. (2011). Capital allocation and timely accounting recogni- tion of economic losses. Journal of Business Finance & Accounting, 38(1–2), 1–33. 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, X., & Huang, D. (2008). Corporate life cycle, accrual characteristics and accounting con- servatism. China Accounting and Finance Review, (1), 1–33. Dechow, P., Hutton, A., & Sloan, R. (1999). An empirical assessment of residual income valua- tion model. Journal of Accounting and Economics, 26(1), 1–34. Demruger, S., Sachs, J.D., Woo, W.T., Bao, S., Chang, G., & Mellinger, A. (2002). Geography, economic policy and regional development in China. Asian Economic Papers, 1(1), 146–197. Dennis, S., Nandy, D., & Sharpe, I.G. (2000). The determinants of contract terms in bank revol- ving credit agreements. Journal of Financial and Quantitative Analysis, 35(1), 87–110. Djankov, S., McLiesh, C., & Shleifer, A. (2007). Private credit in 129 countries. Journal of Financial Economics, 84, 299–329. Doyle, J., Ge, W., & McVay, S. (2007). Determinants of weaknesses in internal control over financial reporting. Journal of Accounting & Economics, 44(1-2), 193–223. Elgers, P.T., & Lo, M.H. (1994). Reduction in analysts’ annual earnings forecast errors using infor- mation in prior earnings and security returns. Journal of Accounting Research, 32, 290–303. Fama, E.F., & MacBeth, J.D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81, 607–636. Fan, G., Wang, X., & Zhu, H. (2009). NERI INDEX of marketization of China’s provinces 2009 Report. Fang, X., Sun, A., & Cao, Y. (2010). The study on financial support and economic growth based on space model – Evidence from Chinese provincial city of 1998-2008. Journal of Financial Research, (10), 68–82. (In Chinese.). Givoly, D., & Hayn, C. (2000). The changing time-series properties of earnings, cash flows and accruals: Has financial reporting become more conservative? Journal of Accounting and Economics, 29, 287–320. Graham, J.R., Li, S., & Qiu, J.P. (2008). Corporate misreporting and bank loan contracting. Journal of Financial Economics, 89(1), 44–61. Guo, F., Ma, Y., Bai, P., & Huo, X. 2014. The analysis and forecast report of investment in Chinese small and medium-bank market 2014-2018. China Investment Advisory Network. Holthausen, R., & Watts, R. (2001). The relevance of the value-relevance literature for financial accounting standard setting. Journal of Accounting and Economics, 31(1-3), 3–75. Jappelli, T., Pagano, M., & Bianco, M. (2005). Courts and banks: Effects of judicial enforcement on credit markets. Journal of Money, Credit and Banking, 37, 223–244. Khan, M., & Watts, R. (2009). Estimation and empirical properties of a firm-year measure of accounting conservatism. Journal of Accounting and Economics, 48, 132–150. LaFond, R., & Watts, R.L. (2008). The information role of conservatism. The Accounting Review, 83, 447–478. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R.W. (1998). Law and finance. Journal of Political Economy, 106,1113–1155. Lara, J., Osma, B., & Penalva, F. (2007). Cost of equity and accounting conservatism. Working paper, University of Navarra. Li, J., Ran, G., & Wen, T. (2007). The internal mechanism of finance affect the economic growth —the analysis based on theory of labor division. Journal of Financial Research, (6), 80–99. (In Chinese.). 226 Zhao et al. Lin, T. (2005). China banking reform: Retrospect and prospect. China Finance, 3,16–18. (In Chinese.). Liu, Y., Wu, X., & Jiang, T. (2010). Ownership, debt financing and accounting conservatism: Empirical evidence from Chinese listed companies. Accounting Research, (1), 43–50. (In Chinese.). Lin, X., & Zhang, Y. (2009). Bank ownership reform and bank performance in China. Journal of Banking & Finance, 33(1), 20–29. Luo, W., Zhang, Y., & Zhu, N. (2011). Bank ownership and executive perquisites: New evidence from an emerging market. Journal of Corporate Finance, 17, 352–370. Moerman, R. (2008). The role of information asymmetry and financial reporting quality in debt trading: Evidence from the secondary loan market. Journal of Accounting and Economics, 46, 240–260. Myers, J. (1999). Implementing residual income valuation with linear information dynamics. The Accounting Review, 74(1), 1–28. Piotroski, J.D., & Wong, T. (2011). Institutions and information environment of Chinese listed firms. Chicago, IL: University of Chicago Press. Pope, P.F., & Walker, M. (1999). International differences in the timeliness, conservatism, and classification of earnings. Journal of Accounting Research, 37(Supplement), 53–87. Rao, P., & Jiang, G. (2011). Monetary policy change, debt contracts and accounting conserva- tism. Journal of Financial Research, (3), 51–71. (In Chinese.). Qian, X., & Cao, C. (2013). Will credit environment affect bank loans? Journal of Financial Research, (4), 57–70. (In Chinese.). Tan, X. (1999). An empirical study of Chinese financial development and economic growth. Eco- nomic Research Journal, (10), 53–61. (In Chinese.). Wang, G. (2001). Interest rate control and market-oriented interest rate in China. Economic Research Journal, (6), 13–20. (In Chinese.). Wang, D., & Zhang, X. (2007). Interest rate liberalization, corporate finance and credit behavior of financial institutions. The Journal of World Economy, (2), 50–59. (In Chinese.). Watts, R.L. (1993). A proposal for research on conservatism. Working paper, University of Rochester. Watts, R. (2003a). Conservatism in accounting. Part I: Explanations and implications. Accounting Horizons, 17, 207–221. Watts, R. (2003b). Conservatism in accounting. Part II: Evidence and research opportunities. Accounting Horizons, 17, 287–301. Watts, R., & Zimmerman, J. (1986). Positive accounting theory. Englewood Cliffs, NJ: Prentice- Hall. Xu, H., & Sun, Z. (2006). An empirical study on the relationship between stock market cycles and conservation of enterprise accounting policy in China. Journal of Finance and Econom- ics, 12,53–67. (In Chinese.). Xu, X., & Shen, H. (2010). The effect of supervision of bank loans and earnings conservatism. Journal of Financial Research, (2), 102–111. (In Chinese.). Yang, D., Wang, N., & Ye, J. (2011). Accounting conservatism and investment behavior of listed firms. Accounting Research, (3), 27–33. (In Chinese.). Yin, J., Wang, Q., Guo, J., Li, Y., & Wang, G. 2012. Annual report of China’s financial develop- ment (2012). Blue book of finance. (In Chinese.). Zhang, J. (2008). The contracting benefits of accounting conservatism to lenders and borrowers. Journal of Accounting and Economics, 45,27–54. Zhang, J., & Wang, P. (2012). Bank risk taking, bank loans, and legal protection. Economic Research Journal, (5), 18–29. (In Chinese.). Zhao, D., Zeng, L., & Tan, L. (2008). Supervision of independent directors and earnings conser- vatism: An empirical study based on Chinese listed firms. Accounting Research, (9), 55–63. (In Chinese). Zhu, S., Xia, D., & Chen, C. (2010). Auditing tenure and accounting conservatism. Auditing Research, (3), 89–95. (In Chinese.). China Journal of Accounting Studies 227 Appendix. The calculation of Cscore Based on Basu (1997) and Khan and Watts (2009), create a new measurement known as Cscore to proxy for accounting conservatism, based on the following model: EPS =P ¼ a þ a DR þ a R þ a R DR þ e (A1) jt j;t1 0 1 jt 2t jt 3t jt jt jt where EPS is the earnings per share for firm j in year t, P is the closed stock price at the jt j;t1 end of year t–1 for firm j, R is the stock return for firm j in year t, and DR is a dummy variable jt jt equal to one when R <0 and zero otherwise. In this model, good and bad news are determined jt according to whether R is larger or smaller than zero, respectively (R ≥0 indicates good news jt jt and R <0 indicates bad news). This model reflects the correlation between accounting earnings jt and stock returns. The sum of coefficients α +α represents the timeliness of bad news and the 2 3 coefficient α represents the timeliness of good news. The coefficient α refers to accounting con- 2 3 servatism. We further suppose that Gscore ¼ a ¼ l þ l MV þ l MB þ l Lev (A2) 2 1t 2t jt 3t jt 4t jt Gscore ¼ k ¼ k þ k MV þ k MB þ k Lev (A3) 3 1t 2t jt 3t jt 4t jt where MV is the logarithm of the market value, MB is the market-to-book ratio of stock equity, and Lev is the debt ratio. We add equations (A2) and (A3) to equation (A1) and obtain the following: EPS =P ¼ a þ a DR þ R ðl þ l MV þ l MB þ l LevÞ jt j;t1 0 1 jt jt jt jt 1t 2t 3t 4t þ R DR ðk þ k MV þ k MB þ k Lev Þ jt jt 1t 2t jt 3t jt 4t jt þðd MV þ d MB þ d DR Lev þ d DR MV þ d DR MB þ d DR LevÞþ e 1t 2t 3t jt 1t jt 2t jt 3t jt jt (A4) We perform a yearly cross-sectional regression on equation (A4) (Fama & MacBeth, 1973) and obtain the coefficients λ , λ , λ and λ . We then add these coefficients to equation (A3) and 1t 2t 3t, 4t calculate Cscore.
http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png
China Journal of Accounting Studies
Taylor & Francis
http://www.deepdyve.com/lp/taylor-francis/accounting-conservatism-and-bank-loan-contracts-evidence-from-china-iiHEdsS0sj
