이익조정 구간과 수단이 기업신용등급 및 부채조달비용에 미치는 효과에 대한 비교분석

This paper examines whether there is a difference between credit rating agencies and lenders in detecting earnings management using a sample of KOSPI and KOSDAQ listed firms. Credit rating agencies indicate that earnings and cash flows are important financial factors for assessing a firm’s creditworthiness. In addition, lenders use earnings and other accounting information to assess firm’s financial solvency, stability, credibility, and feasibility. Lenders tend to focus on a firm’s ability to generate future cash flows to ensure payment of periodic interest and bond’s principal. Quality of accounting information affects credit rating agencies and lenders’ estimates of future cash flows. However, much empirical evidence suggests that earnings management is a common phenomenon. In particular, prior studies simply use credit rating estimated by credit rating agencies as a proxy of cost of debts because it is difficult to accurately measure cost of debt. However, it is expected that information provided by credit rating agencies and lenders would be different because credit rating agencies do not provide funds and lenders do not directly measure a credit rating of a firm. Thus, main purpose of this study is to examine how credit rating agencies and lenders catch each earnings management case based on 4 earnings management intervals and means categories. Burgstahler and Dichev (1997; hereafter B&D) document that firms manage their reported earnings to avoid losses or earnings decreases, thus increase incidence of producing small positive earnings and small earnings increases. B&D find that earnings management to avoid annual losses and earnings decreases is common. B&D show earnings management in the context of net income, a measure that includes the effects of discontinued operations, extraordinary items, and special items. Specifically, we consider whether credit rating agencies and lenders predict the earnings management to avoid annual losses and earnings decreases documented in B&D. Meanwhile, current-period reported earnings can be managed in two different ways. First, managers can manipulate reported earnings through discretionary accrual choices that are allowed under Generally Accepted Accounting Principles (GAAP). This within-GAAP accrual-based earnings management (hereafter AM) typically occurs at the end of an accounting period, after most real operating activities are completed. While it directly influences the amount of accounting accruals, AM has no direct effect on cash flows. Second, managers can also manipulate reported earnings by adjusting real activities. Specifically, they can alter timing and scale of real activities such as sales, production, investment, and financing throughout accounting period in such a way that a specific earnings target can be met (Roychowdhury 2006; Kim and Sohn 2009; Ge and Kim 2013 etc). Following Roychowdhury (2006), these real operation management activities that deviate from normal business practices with the primary objective of manipulating current- period earnings are referred to as real earnings management (hereafter RM). RM activities distort the quality of reported earnings, which can impact stakeholders’ estimates of future cash flows. Because RM can have direct negative consequences on the level of future net cash flows, so auditors, credit rating agencies and lenders, and analyst are likely to be concerned about and respond to RM activities. Earnings management distorts the quality of reported earnings, which can impact credit rating agencies and lenders’ estimates of future cash flows. Because fundamental risk and opportunistic earnings management behavior reduce information precision and therefore lead to poor earnings quality. This paper categorized two earnings management intervals (earnings decrease avoidance and loss avoidance intervals) and two earnings management means (discretionary accruals and real earnings management), and expected that credit rate is lower and cost of debt is higher if a firm manages its earnings to avoid earnings decreases and to avoid losses or a firm manages its earnings through AM and RM. Therefore, ities that deviate from normal business practices with the primary objective of manipulating current- period earnings are referred to as real earnings management (hereafter RM). RM activities distort the quality of reported earnings, which can impact stakeholders’ estimates of future cash flows. Because RM can have direct negative consequences on the level of future net cash flows, so auditors, credit rating agencies and lenders, and analyst are likely to be concerned about and respond to RM activities. Earnings management distorts the quality of reported earnings, which can impact credit rating agencies and lenders’ estimates of future cash flows. Because fundamental risk and opportunistic earnings management behavior reduce information precision and therefore lead to poor earnings quality. This paper categorized two earnings management intervals (earnings decrease avoidance and loss avoidance intervals) and two earnings management means (discretionary accruals and real earnings management), and expected that credit rate is lower and cost of debt is higher if a firm manages its earnings to avoid earnings decreases and to avoid losses or a firm manages its earnings through AM and RM. Therefore, we predict a positive association between earnings management intervals and two earnings management means and credit ratings, whereas a negative association between earnings management intervals and two earnings management means and cost of debt. To test this prediction, we take credit rating information from KISVALUE. Corporate credit ratings have ranged from the value 1 to 10. The value 1 represents a firm with the worst credit rank and 10 represents a firm with the best credit rank (Park and Nam 2010). Following the prior studies (e.g., Park and Kim 2013; Park et al. 2011a etc.), we employ the yield spread of interest rates as proxies for the cost of debt. As B&D show the cross-sectional discontinuity of earnings distribution around zero, we identify the regions with small positive earnings and small earnings increases. Also, we measure two proxies for earnings management means, which are accrual- based earnings management measured using performance-adjusted discretionary accruals in Kothari et al. (2005), and real earnings management measured using three proxies in Roychowdhury (2006); abnormal cash flows from operations, abnormal production cost, and abnormal discretionary expenditures, and specifically, we use a comprehensive measure of the three RM measures. We examine all listed firms in the Korean stock market with available corporate credit ratings and cost of debt. Tests are performed using a full sample of KOSPI and KOSDAQ listed firms and also using subsamples of each type of stock markets. Therefore, this study uses the total 12,012 (credit rating) and 10,630 (cost of debt) firm-year observations for the period from 2001 to 2010. As a result, this research found that credit rating agencies and lenders are different when they detect firm’s earnings management except for loss avoidance firm’s earnings management and these results are consistent in the analysis of KOSPI and KOSDAQ listed firms. It is found that credit rating agencies do not reflect earnings decrease avoidance firm’s earnings management and accrual-based earnings management when they rate those firms, while lenders do not detect accrual-based earnings management and real earnings management thus inefficiently allocate their resources in the capital market. It can be concluded that prior researches using credit rating as a proxy of cost of debts is limited and these findings will help for creditors and investors to understand the difference between credit rating and cost of debt. This research also contributes to provide useful information related to earnings management intervals and means to scholars, practitioners, accounting standard setters and regulators. In addition, researchers who are interested in this area can also apply the discussion in this paper for the related studies.