당기 이익에 영향을 미치는 이익구성요소의 특성

This study compares attributes between traditional and time-specific earnings components. Sloan (1996) shows that operating cash flows are more persistent than accruals. However, Francis and Smith (2005) document that Sloan's (1996) result for the persistence difference between traditionally defined operating cash flows and accruals is weakened when both components are measured in a “time-specific” manner. Francis and Smith's (2005) time-specific earnings components exclude such items that are not included in current-period earnings but included in either previous-period or next-period earnings. This study examines the differential attributes of time-specific earnings components compared to traditional earnings components. Specifically, the attributes such as the association with earnings, first-order autocorrelation, volatility, and persistence are compared between two sets of earnings components. Adopting Francis and Smith's (2005) assumptions on time-series properties, variance and covariance of earnings components, I expect (1) that the first- order autocorrelation of traditional operating cash flows is larger than that of traditional accruals only if the first-order autocorrelation exceeds 0.333 while the first-order autocorrelations of time-specific earnings components are indistinguishable, (2) that traditional operating cash flows are more volatile than traditional accruals only if the first-order autocorrelation exceeds 0.250 while time-specific accruals are more volatile that time-specific operating cash flows,and (3) that traditional operating cash flows are more persistent than traditional accruals unless the absolute value of the first-order autocorrelation exceeds 0.707 while time-specific earnings components are identically persistent. The final sample consists of 9,200 non-financial firm-years with December fiscal year that are traded over Korea Exchange for 1995-2011. Empirical findings are summarized as follows. First, when the cross-sectional distribution of two sets of earnings components, traditional accruals are more highly associated with earnings than operating cash flows but tims-specific accruals are less highly associated with earnings than operating cash flows. Second, the firstorder autocorrelations of time-specific earnings components are larger than those of traditional earnings components. Third, traditional earnings components are more volatile than timespecific earnings components. Fourth, time-specific operating cash flows are less persistent than traditional operating cash flows while time-specific accruals are more persistent than traditional accruals. And as a consequence, the difference of persistence between operating cash flows and accruals becomes smaller for time-specific earnings components than for traditional earnings components as evidenced by Francis and Smith (2005). The results in this study deepen our understanding of the accrual basis of accounting by documenting a distinguishable nature of accounts with the current-period earnings effect from those without the current-period earnings effect. The results in this study suggest that timespecific earnings components are less volatile than traditional earnings components and that the persistence difference between operating cash flows and accruals becomes smaller for timespecific earnings components than for traditional earnings components. Higher volatility and larger difference of persistence of traditional earnings components are of course caused by accrual-entry and deferral-entry accounts that do not affect current-period earnings but affect previous-period or next-period earnings. In other words, although the accounts without currentperiod earnings effect do not affect current-period earnings, such accounts are more likely to deteriorate the quality of operating cash flows or accruals. Information users should pay more attention to a distinctive nature between traditional and time-specific earnings components for their economic decisions.