The effect of SOX on the predictability of future cash flows in litigious and non-litigious industries

Abstract

We investigate whether the role of discretionary accruals in predicting future operating cash flows changes after the passage of SOX. We also examine the information content of discretionary accruals in litigious industries. We find that discretionary accruals are positively associated with future operating cash flows and that discretionary accruals become even more important to predict future cash flows during the post-SOX period. Findings also indicate that litigious industry firms impart greater information content relative to those in nonlitigious industries prior to SOX being issued and that the SOX effect on discretionary accruals is weaker for such firms as a result.

Keywords:

• discretionary accruals
• operating cash flows
• SOX
• litigious industries.

JEL Descriptors:

• C12
• C21
• M41

Notes

*We consider a firm to operate in a litigious environment, if it conducts its business in an industry classified under one of the following SIC Codes: 2833–2836, 3570–3577, 3600–3674, 5200–5961, and 7370.

*The correlation is significant at the 0.05 level or better.

1. Section 906 of SOX states: (1) “Whoever certifies any statement … knowing that the periodic report accompanying the statement does not comport with all the requirements set forth in this section shall be fined not more than $1,000,000 or imprisoned not more than 10 years, or both;” and (2) “Whoever willfully certifies any statement … knowing that the periodic report accompanying the statement does not comport with all the requirements set forth in this section shall be fined not more than $5,000,000 or imprisoned not more than 20 years, or both.”

2. Audit committees have expanded authority over the appointment and compensation of the external auditor. Audit committees also have increased oversight of the auditor’s work for the purpose of preparing or issuing an audit report or related work. For example, the audit committee must pre-approve all services provided by the external auditor, after first determining that the services do not pose a conflict with the auditor’s independent role. Moreover, each audit committee must comprise independent directors. The company must disclose whether at least one member of the audit committee meets the specified criteria of an “audit committee financial expert” and, if not, the reasons why.

3. Although the litigious nature and the corporate governance preferences among certain industries may impact the informativeness of discretionary accruals with respect to future cash flows differently, that is not the focus of our paper. We instead directly investigate the impact of SOX on the predictive ability of discretionary accruals as potentially moderated by the higher likelihood of litigation in certain industries. Therefore, we assume that the higher threat of legal liability in a particular industry increases the likelihood of firms in that industry adopting a more proactive corporate governance policy, which may in part, then influence how informative its discretionary accruals will be for predicting future cash flows. Hence, we expect that whether or not a firm is in a litigious industry should, at least weakly, proxy for corporate governance effectiveness. The literature does present evidence to support this assumption (Baker and Griffith 2007, Krishnan and Lee 2009). To that extent, we also use an alternative measure of board independence as a proxy for corporate governance effectiveness to moderate the impact of SOX in place of the litigious industry indicator variable in our robustness tests.

4. Though we are able to replicate Subramanyam’s (1996) results and generate consistent inferences overall using OLS, we present our results based on fixed effects models holding both industry and year effects constant. By controlling for average differences across industries and years, we are able to significantly reduce the threat of omitted variables in our models.

5. While we have no theoretical expectation regarding an association between LIT and future cash flows, including the LIT dummy as a standalone variable in the model (not tabulated) does not significantly alter the results or our inferences based on them.

6. The SIC Codes represent firms in the biotechnology, computer equipment, electronics, retailing, and computer services industries, respectively.

7. Results using the Jones (1991) model to estimate discretionary and nondiscretionary accruals are consistent with those we present in the paper.

8. Results are qualitatively similar; whether we follow Subramanyam (1996, Hribar and Collins (2002), or Kraft et al. (2007) in computing total accruals.

9. Although the correlation coefficients are generally highly significant, Eigenvalues, variance inflation factors, condition indices, and tolerance tests for the reported results are all within acceptable levels.

10. Untabulated F-tests for coefficient estimate differences are all highly significant.

11. We present results obtained using the cash flows from operations variable definition endorsed by Subramanyam (1996). Results using other variable definitions, which are not tabulated, are consistent with our reported results.

12. Hribar and Collins (2002) use a cash flow from continuing operations measure, defined as Compustat annual data items #308 minus #124. Kraft et al. (2007) define cash flows from operations as the difference between operating income (Compustat annual data item #178) and total accruals (Compustat annual data items ((#4 – Lag(#4)) – (#1 – Lag(#1)) – (#5 – Lag(#5)) + (#34 – Lag(#34)) - #14)).

13. We follow Subramanyam and Venkatachalam (2007) to generate the ex post intrinsic value of equity by discounting future dividends, over a three-year horizon, and a terminal dividend, the stock price at the end of three years (see pages 463–464).

14. Due to data limitations, we are only able to use a 1999–2007 sample period in the determination of ex post intrinsic values over a three-year horizon period. We also consider a two-year horizon period value which allows us to extend our usable sample period to 2008. Results are robust to such changes in the ex post intrinsic value horizon period.

15. We calculate board overlap by dividing the total number of directors on the audit committee into the total number of directors serving on both the audit and compensation committees.