Seemingly opportunistic management earnings guidance before stock option grants: does it misrepresent firms’ underlying performance?

Abstract

The exercise price of stock options is typically the closing stock price on the option grant dates, so managers can potentially benefit from low stock prices on those dates. Prior studies find that on average, managers issue more pessimistic guidance before than after grant dates. They interpret this asymmetric pattern as representing managers’ opportunistic behavior. Nevertheless, it is not clear whether this pattern reflects managers’ selective timing of bad news disclosures or a deliberate misrepresentation of underlying firm performance. This paper extends the literature by examining the effects of managerial stock option incentives on the accuracy and the information content of firms’ voluntary earnings guidance. We find that pre-grant guidance significantly improves existing consensus earnings forecasts, and is similar in bias and accuracy to post-grant guidance. Moreover, investors and analysts react similarly to pre-grant vs. post-grant guidance, suggesting that the main consumers of earnings guidance view these two types of guidance to be equally informative. Our results are consistent with the notion that managers may opportunistically select the type or the timing of disclosures, but when they opt to disclose, they do not misrepresent the underlying firm performance. On the contrary, the seemingly opportunistic guidance before option grants improves overall firms’ information environment, because it is at least as truthful and informative as post-grant guidance, and is issued more frequently than by managers who do not have incentives to report bad news before earnings announcements.

Keywords:

• stock options
• earnings guidance
• information content
• incentives

Notes

1. See, for example, Morck, Shleifer, and Vishny (1988); Hanlon, Rajgopal, and Shevlin (2003); and Desai and Dharmapala (2006).

2. Yermack (1997), Balsam, Chen, and Sankaraguruswamy (2003), Huddart and Lang (2003), Bartov and Mohanram (2004), Bergstresser and Philippon (2005), Cheng and Warfield (2005), and Coles, Hertzel, and Kalpathy (2006).

3. Burns and Kedia (2006) and Efendi, Srivastava, and Swanson (2007) find evidence that managers manipulate earnings to increase the value of their outstanding options, while Balsam, Chen, and Sankaraguruswamy (2003), Baker, Collins, and Reitenga (2003), and McAnally, Weaver, and Srivastava (2008) find evidence of managerial actions to reduce stock prices before option awards.

4. Prior studies examine stock price or abnormal accruals surrounding option plan events to assess management’s behavior (Bartov and Mohanram 2004; Chauvin and Shenoy 2001; Coles, Hertzel, and Kalpathy 2006; Yermack 1997). Voluntary disclosures offer a number of advantages over these approaches: (1) unlike stock price, voluntary disclosures are fully controlled by management and can be directly linked to managers’ compensation incentives to influence security prices; (2) it might be difficult to disentangle managers’ ex ante actions and ex post option back-dating when examining stock price behavior around stock option grants; (3) managers have greater latitude in choosing the timing and extent of guidance than they do with discretionary accruals; and (4) management guidance can be directly observed, but there is little agreement on how best to measure earnings management.

5. Execucomp provides the future option expiration date as obtained from proxy statements filed with the SEC. Consistent with prior research, we assume that the grant date in the current year is the same as the expiration month and day.

6. The mean calculations are done separately for quarterly and annual periods.

7. This requirement eliminated approximately 8% and 10% of the quarterly and annual observations, respectively. In all analyses throughout the paper, we note how our results change when we retain these observations. Overall, our conclusions are qualitatively unchanged by this data screen.

8. Note that the three subsamples presented in each panel in Table 1 do not sum up to the full sample. This occurs because some guidance observations can fall into more than one category when firms issue separate stock grants within an interval of less than sixty days. Guidance that is issued within this interval can be classified as “guidance after grant” for the first grant and “guidance before grant” for the next grant, and thus these observations are not included in either category.

9. Early research on management forecasts finds more frequent upward guidance than downward guidance (e.g. Patell 1976; Penman 1980), but the distribution of guidance types in our sample is consistent with more recent research (Skinner 1994; Soffer, Thiagarajan, and Walther 2000).

10. We repeat tests related to earnings announcements using 30-day and 5-day windows with qualitatively equivalent results to those reported.

11. We focus on guidance related to earnings announced after the stock option award. Grant-date stock prices for firms whose earnings are announced before the award are likely to be primarily determined by the announced earnings rather than by guidance, so managers have no reason to use guidance to increase option compensation in these cases.

12. For the full untruncated sample, we find a marginally significant difference in the annual earnings 30-day window (t-statistic = 1.80). However, the direction of the difference indicates more optimistic guidance in the pre-grant period, which is inconsistent with option incentives inducing managers to misrepresent their private information.

13. We find one exception for the full untruncated sample –90-day window for the quarterly data–where we find a marginally significant difference (t-statistic = 1.75).

14. We repeat our regression analyses for our full untruncated sample. All results reported in Table 3 are qualitatively similar for the full sample except that α₂ – α₃ is significantly negative for quarterly earnings when the dependent variable is Xret_Guide (t-statistic = –2.17). This is the only piece of evidence consistent with a lower return response by market participants to pre-grant guidance than to post-grant guidance. However, this result could be unduly influenced by large values of Guide_News.

15. The low-grant and high-grant firms’ coefficients for each grant quarter (e.g. a₅ – a₁) are all significantly different at better than the 1% level in both models.

16. In untabulated analysis, we find that analyst forecasts are significantly more accurate for both low-grant and high-grant firms than for no-grant firms. None of the coefficient differences for low vs. high stock option grant firms are statistically significant.