채권투자자 관점에서의 비상장기업에 대한 이익, 영업현금흐름 및 이익구성요소의 상대적 그리고 추가적 정보력에 관한 연구

The purpose of this paper is to address two questions related to unlisted firms. The first, this paper examines the effect of accounting earnings and cash flows from operating on the cost of debt, and which information is relatively better than others. The second, this paper examines the relative and incremental information effect of each component of accounting earnings on the cost of debt. Specifically, we decompose accounting earnings into cash flows from operating and total accruals, further accounting earnings into cash flow from operating, discretionary accruals, and non-discretionary accruals. Then we examine the effect of each component on the cost of debt and estimate the relative and incremental information content of each component in terms of debt holders’ perspectives. Previous studies have heavily investigated a usefulness of accrual-based earnings since Ball and Brown (1968). They document that accounting earnings provide the incremental information compared to cash flows from operating on stock returns. However, the argument on the incremental information of cash flows from operating compared to accounting earnings is still controversial. Bowen et al. (1987) indicates that cash flows from operating has incremental information compared to accounting earnings while Livnat and Zarowin (1990) argue that cash flow information doesn't have any incremental contents. Other studies suggest that usefulness of accruals information should be considered in the model and find that the predictability of accruals is greater than that of cash flows from operating. This finding suggests controversial arguments on the role of discretionary accruals. One is a earnings management method for managers' opportunistic behavior (Warfield et al. 1995; Dechow 1994). Another is a private signal from informed managers (Healy and Palepu 1993; Dechow 1994). However, these findings are restricted to the listed companies. Some studies suggest that accounting environments of unlisted companies are quite different from that of listed companies. Especially, they argue that for unlisted companies, information about cash flows from operating is more useful than accounting earnings (Penno and Simon 1986; Mills and Newberry 2001; Beatty et al. 2002; Chaney et al. 2004; Ball and Shivakumar 2005; Burgstahler et al. 2006). And unlisted companies make a major contribution to economic growth and proportion and importance in the market have increased (Burgstahler et al. 2006). This leads to increased concerns from varied interested parties. Therefore, investigation on the effect of accounting earnings produced by unlisted firms is worth exploring. To test the relative and incremental information content of each component of earnings on the cost of debt, 39,185 firm-year observations from 2004 to 2009 in unlisted firms are used for analyses. We use the unlisted companies that are subject to a external audit for tests. Because unlisted companies that are not under a external audit have low level of reliability in financial statements. For a dependent variable, we use two kinds of proxies for the cost of debt. First, we calculate three variables: COD1, COD2, and COD3, respectively. COD1 represents a ratio that total finance cost divide by average interest cost of debt. COD2 presents a ratio that interest cost divide by average interest cost of debt. Finally COD3 presents a average rate for borrowing. Then we estimate the average value of three variables and use this value as a proxy for the cost of debt (COD). Second, we use yield spreads as another proxy for the cost of debt (CODYS). These variables are transformed as fractional rankings variables to reduce the impact of outliers (Francis et al. 2004; Kim and Sohn 2013). The empirical findings of this paper are following. The first, when we test incremental information effect of accounting earnings and cash flows from operating in the model, we find that both accounting earnings and cash flows from operating are negatively associated with the cost of debt, respectively. When it comes to the incremental information of each component on the cost of debt, both information provide different contents. In addition, the test on relative information effect indicates that accounting earnings have a greater relative information than that of cash flows from operating. This suggests that accrual-based earnings provide useful information to creditors even in unlisted firms. This finding implies that the conventional wisdom that accounting earnings are a good indicator of a firm's performance than cash flows from operating can be applied for unlisted firms. The second, when accounting earnings are decomposed into cash flows from operating and total accruals, we find that both informations have incremental contents and the relative information of cash flows from operating is greater than that of total accruals. This result is consistent with prior studies using listed companies (Sloan 1996; Subramanyam 1996; Guay et al. 1996). The third, when accounting earnings are decomposed into cash flows from operating, discretionary accruals, and non-discretionary accruals, each component of earnings is negatively associated with the cost of debt. Furthermore, each three component also has incremental information on the cost of debt. When it comes to a relative information effect, discretionary accruals, cash flows from operating, and non-discretionary accruals are classed in the order of relative information contents. These results imply that discretionary accruals rather than total accruals provides more useful information to creditors. In other words, discretionary accruals information in unlisted firms is considered as a vehicle to deliver private information related to a manager’s prospect, not a opportunistic earnings management (Healy and Palepu 1993; Dechow 1994; Subramanyam 1996; Guay et al. 1996). These results are identically significant when we apply Newey and West t-statistics that can address cross-sectional and time-series problems in the regression. The findings of this paper have following contributions. The results of this paper support the conventional wisdom that accounting earnings based on accrual-based accounting is still useful for unlisted firms. And the findings of this paper indicate that even unlisted firms, accounting earnings is very important information and has the relative and incremental information contents to creditors. Therefore, these findings are very useful and provide a lot of important implications to regulators, investors and creditors who are interested in unlisted firms. Academics can also apply the discussion in this paper for related researches.