Evidence for effects of non-operating cash flows on working capital management policy and its econometric consequences for accrual models

In this paper, using a sample including 2642 observations of 2001-2016 annual data of firms listed in TSE, we provide evidence that working capital management policy bears a change following large inflows or outflows of non-operating cash that are correlated with corporate events like IPOs or financing and investing events and predicted to affect firm’s bearable risk of investment in working capital. The research findings show that, because of the ensuing increase/decrease in investment in working capital, this change is likely to cause a change in working capital accruals independent of shock to sales and, consequently, given that its effect are uncontrolled, accruals models tend to overstate or understate abnormal accruals, especially in earnings management detection tests around corporate events. The results of Monte Carlo simulations of type I error show that the examined accrual models understate (overstate) the level of estimated abnormal accruals when applied to random samples of observations with extreme negative (positive) changes in current non-operating cash flows. Since there are correlations between the non-operating cash flows and partitioning variables used in earnings management studies around corporate events, our findings suggests that it is likely that there are correlation between measurement error in the estimate of abnormal accruals proxy and portioning variables, and, as a consequence, the results of such tests may not be unbiased. The study highlights the need for a control of non-operating cash flows effects on working capital management, especially in empirical tests for detection of earnings manipulations around corporate events.