Accruals and real earnings management around debt covenant violations
This paper examines real earnings management and accrual-based earnings management in the quarters around debt covenant violations. Earlier studies have examined whether firms engage in one or the other of these activities. The present study examines how firms engage in real earnings management around covenant violations and how real and accrual-based earnings management activities vary around the violation period.
To measure accrual-based earnings management we use the cross-sectional model developed by Jones (1991) to estimate abnormal levels of total accruals and working capital accruals. We use working capital accruals in addition to total accruals as the former are more susceptible to management manipulation (DeFond and Jiambalvo, 1994). To capture real earnings management, we follow Roychowdhury (2006) and Cohen and Zarowin (2010) to estimate abnormal levels of discretionary expenses, production costs and cash flow from operations.
This paper makes four contributions to the literature. First, it is the first study that uses the covenant violation quarterly data created by Sufi to provide evidence of earnings management around violation quarters. Second, it provides direct evidence that managers manipulate accruals to avoid debt covenant violations. Previous studies have used a proxy for the existence and tightness of accounting-based covenants. The most frequently used proxy is the debt-equity ratio, but, as noted by Watts and Zimmerman (1986), researchers in effect have tested a debt- equity hypothesis: that is, the larger a firm’s debt-equity ratio, the more likely the firm is to make accounting choices that shift reported earning from future periods to the current period. This hypothesis is different from the debt covenant hypothesis, which states that managers will choose to shift reported earnings from the future to the current period when a firm is close to violating a debt covenant. The present study is the first to provide evidence on the covenant hypothesis.
Third, we provide evidence on real earnings management around debt covenant violations. The prior literature on the covenant hypothesis has focused mainly on the accounting choices available to managers to avoid covenant violations. The present study appears to be the first to focus on real activities manipulation to test for the debt covenant hypothesis.
Finally, this study details the differences in accrual-based and real earnings management activities in the quarter of and quarters surrounding the violation. While managers engage in accruals management in the quarters surrounding the violation, the opportunities for manipulating real activities may be limited. We find that managers decrease discretionary expenses in the quarter prior to, quarter of and quarter following the violation. However, it is not practicable to manipulate production costs as doing so will affect inventory values. We observe a reversal in abnormal production costs in the quarter following the violation. Likewise, operating cash flows rise as managers withdraw limited time discounts and tighten credit terms. In contrast, abnormal total and working capital accruals are positive in the quarter of and quarters surrounding the violation.