A review of Roychowdhury, S., 2006, “Earnings Management through Real Activities Manipulation,” Journal of Accounting and Economics 42, 335-370.
I. Research question and its importance
Continuing the seminal works of earning management by Hayn (1995) and Burgstahler and Dichev (1997) whose primary focus was discontinuities in distribution around zero threshold, the author advances the stream by developing an empirical method to detect real earning management. Specifically, he documents evidence that earning management meets or beats analyst forecast. A series of robustness checks strengthen his claim, which complements extant prior studies in this literature about accrual as well as the scant research on R&D expense deduction. This paper also responds to survey research pertaining management’s willingness to manipulate operational activities in order to meet earning target. The author contributes by providing direct evidence such as reduced discretionary expenditure as real earning management activity.
II. Method and findings
The observation covers annual data of all public firms from Compustat database from 1987 to 2001. Final sample with complete information consists of 3,672 firms. For each firm-year, the author constructs cash flow from operation, CFO, in accordance to Burgstahler and Dichev (1997). Measures of production cost and discretionary expense are developed and estimated using the model proposed by Dechow et al. (1998) to capture abnormality. In addition, the author examines separately for matching behaviours toward analyst forecast and prior performance. Fama-Macbeth regression is adopted. For robustness, the author advances the work of Thomas and Zhang (2002) by ruling out response to prevailing economic circumstance. The author documents (1) loss avoidance where price was decreased to boost revenue, (2) overproduction which was intended to decrease COGS, and (3) a reduction of discretionary expense to increase margin. From the regression results, he also finds that sophisticated investor is negatively related to earning management through real activities. Nonetheless, several other factors are also found to have power in explaining earing management: industry membership (manufacturer or not), inventory, receivable, growth (as measured by M/B ratio), and debt level. In brief, the findings are consistent with literature.
III. Future research
More attention has been devoted to the antecedent of real earning management. Two notable examples from top tier journal (link 1; link 2) attribute real EM to TMT factors such as “the number of years to retirement to capture key subordinate executives’ horizon incentives,” “compensation relative to CEO compensation to capture their influence” and “compensation recovery policies that authorize the board to recoup compensation paid to executives based on misstated financial reports.” Thus, future research might connect with TMT literature.