88 / 322
營運資金管理與裁決性應計數之估計
88
changes in performance reinforced the conclusion that increasing the investment in working
capital results in a large amount of discretionary accruals.
4. Implications
Both earnings management and working capital management caused changes in the
total accruals, and the estimated discretionary accruals were affected by both earnings
manipulation and normal operating strategy. Excluding normal accruals from estimated
discretionary accruals is difficult. This study indicated that Jones-type models fail to control
the effect of working capital management on discretionary accruals and may cause a biased
estimation in discretionary accruals. Therefore, people who use Jones-type models should
understand such limitations and cautiously estimate discretionary accruals.
5. Contribution
This study contributes to the literature in three aspects. First of all, this is the first study
to examine the limitations of Jones-type models. We argue that Jones-type models cause
estimation bias in discretionary accruals because revenue growth cannot completely capture
the effect working capital management on accruals. Secondly, this study involves
categorizing sample companies into those that increase capital and those that reduce capital.
The results of these particular samples support our hypotheses, which indicate that Jones-
type models cannot be used to estimate discretionary accruals effectively and thus are not
credible for testing hypotheses related to earnings management. Finally, given the
bidirectional implications for working capital management, future research should consider
working capital management in determining the level of earnings management. This would
help prevent potential biased estimates caused by Jones-type models.