Printer Friendly

The effects of option incentives on backdating and earnings management.

I. INTRODUCTION

CEO stock options are typically granted at-the-money, where the exercise price of the option is set at the market price on the grant date. Lower share price on the grant date and price appreciation afterwards can help managers profit from their option compensation. Previous research suggests that managers can pursue a variety of opportunistic managerial behaviors to increase the value of their option grants, including backdating option grants to periods of lower share prices (Yermack, 1997; Aboody and Kasznik, 2000; Chauvin and Shenoy, 2001; Lie, 2005; Heron and Lie, 2007; and Narayanan and Seyhun, 2006), or decreasing earnings to suppress share price when options are granted (Baker, Collins, and Reitenga, 2003; Cheng and Warfield, 2005; Bergstresser and Philippon, 2006; Cornett, Marcus, and Tehranian, 2008; McAnally, Srivastava, and Weaver, 2008). In other words, newly granted options induce CEOs to engage in either option backdating or downwards earnings management. But CEOs' option portfolios do not include only newly granted options. In addition to new options granted in the current year, CEOs' option portfolios also have vested and unvested outstanding options that CEOs retain from their grants in previous years. These two elements can also motivate CEOs to pursue self-interest. Therefore, the purpose of this paper is to examine how various elements of CEOs' option portfolios create conflicting incentives for both earnings management and option backdating, and investigate the link between the two opportunistic behaviors in order to explain the effects.

We examine a sample of 9,568 CEO-year observations in Standard & Poor's Execucomp database from 1992 through 2005, among which 685 CEO years are labeled as the potential backdaters according to the methodology in Collins, Gong and Li (2009), and 8,883 CEO years are labeled as the benchmark firms. We first test the effects of various elements of CEOs' option portfolios on accrual-based earnings management and the occurrence of backdating separately. Given the conflicting effects of various option elements on both earnings management and option backdating, we then examine the link between these two opportunistic actions in order to explain the effects.

Our main findings are as the following. We find that higher incentives from newly granted options are associated with income-decreasing earnings management and higher incidence of backdating, whereas higher incentives from unvested options are related to income-increasing earnings management and lower incidence of backdating. However, incentives from vested options have an insignificant effect on both earnings management and backdating. We further show that negative total accruals and discretionary accruals are associated with higher likelihood of backdating, which suggests that CEOs engaging in option backdating are more likely to use income-decreasing earnings management to get more favorable price for their new option grants as well.

Due to the correlation between downward earnings management and option backdating, the need for managers to diversify the increased-risk associated with stock-based compensation (Ofek and Yermack, 2000; Cheng and Warfield, 2005) can clearly explain the effects of CEOs' outstanding options on both earnings management and the occurrence of backdating. Those managers with high incentives from unvested options in a given year are more likely to take income-increasing accruals to beat-up short-term stock price when they sell stock to diversify risk (Cheng and Warfield, 2005), and then have disincentives for backdating. However, we further show that the negative relation between unvested options and the occurrence of backdating is largely driven by managers with less persistent incentives from unvested options. Those managers with highly persistent incentives from unvested options are more willing to bear the risk involved in their compensation. As a result, they have lower incentives for upward earnings management for selling purpose and then have a less motivation to reduce the incidence of backdating. In addition, by the time that options become exercisable, vested options likely have a constant hedge ratio and generate no needs for further risk diversification (Cheng and Warfield, 2005), and produce unclear preferences for the choice of accrual-based earnings management (1). Due to the insignificant effect on earnings management, vested options do not generate a significant effect on backdating either.

Our paper provides new evidence on the conflicting effects of various option elements on both option backdating and earnings management. Especially, we study the occurrence of backdating in previously unexplored situations where managers face conflicting incentives from newly granted options and outstanding options. We are not aware of any previous studies that investigate the occurrence of backdating relate to such conflicting incentives. For example, Collins, Gong and Li (2009) only consider the effect of newly granted options on backdating. Minnick and Zhao (2009) use a variable of pay-for-performance sensitivity of total option holdings, which include, but do not distinguish, newly granted options and outstanding options. Therefore, our study is the first to examine how conflicting incentives associated with various option elements affect the occurrence of option backdating.

Furthermore, our paper extends the literature on backdating and earnings management. The prior studies separately examine backdating and option grant related earnings management. We combine and extend the two strands of literature by providing new evidence that self-serving CEOs are likely to take both opportunistic actions at the same time when they are granted options.

In addition, our findings have implications for corporate executive compensation policies. The evidence that managers take into account the conflicting incentives from their whole option portfolios in determining their opportunistic actions should be of interest to board of directors who contemplate compensation contracts for managers. The findings that indicate multiple opportunistic actions related to option grants are also relevant to the current public policy debate regarding options and executive pay in general.

The paper proceeds as follows. In Section II we briefly review the literature and develop our hypotheses. Section III describes our sample and data. Section IV presents empirical tests and results, and Section V concludes the paper.

II. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT

A. Literature Review

Option grants to CEOs are designed to align the interests of managers with those of shareholders (Jensen and Meckling, 1976; Jensen and Murphy, 1990). However, the literature also documents that option awards can induce opportunistic behaviors by management. For example, one line of research suggests that managers manipulate the timing of option award dates as a means of increasing the fair value of their awards (Yermack, 1997; Aboody and Kasznik, 2000; Chauvin and Shenoy, 2001; Lie, 2005; Heron and Lie, 2007; and Narayanan and Seyhun, 2006). Specifically, options are found to be awarded prior to increases in stock price, and after decreases in stock price. The other line of research suggests that option grants emerge as a particularly strong predictor of income-decreasing earnings management (Baker, Collins, and Reitenga, 2003; Cheng and Warfield, 2005; Bergstresser and Philippon, 2006; Cornett, Marcus, and Tehranian, 2008; McAnally, Srivastava, and Weaver, 2008), which can directly benefits managers via lower strike price of their option awards.

B. Hypotheses Development

The literature has shown that new option grants create incentives for CEOs to manipulate exercise prices, either through backdating or through downward earnings management. However, the option portfolios that CEOs hold not only include new options that are granted in the current year, but also retain unexercised vested and unvested options that are granted in previous years. These two elements can also motivate CEOs to pursue self-interest and affect the selection of opportunistic managerial actions such as earnings management and option backdating.

The incentives from outstanding options for earnings management can arise from either risk diversifications or option exercises. When a manager is granted options, the options usually are not exercisable until three or four years later. As unvested options cannot be exercised immediately, the effect of unvested options on earnings management through option exercises is unclear. However, managers with high incentives from unvested options have more needs to diversify increased risk associated with stock-based compensation and tend to sell shares. Because of the selling, managers have a motivation to increase the short-term stock price through income-increasing earnings management. Consistent with this argument, Cheng and Warfield (2005) find managers with high incentives from unvested options are more likely to report earnings that meet or just beat analysts' forecast.

Different from unvested options, vested options are exercisable. If managers choose to exercise their vested options, they have a motivation to take income-increasing accruals to beat up short-term stock price, as price appreciation can benefit their option exercises. However, managers can choose to hold their vested options instead and have incentives to manage earnings downward in the current period to reserve good earnings for the future when they decide to exercise. Therefore, the choice of earnings management associated with vested options through option exercises differs across firms and over time, depending on the relative value of the options to be exercised in the current period and the discounted value of options to be exercised in the future (Srivastava, 2005). As a result, the effect of vested options on earnings management through option exercises can be inconsistent. Similarly, the risk diversification argument also predicts an insignificant association between incentives from vested options and earnings management. It is because that by the time when options become exercisable, options likely have a constant hedge ratio and generate no needs for further risk diversification (Cheng and Warfield, 2005).

In sum, earnings management can not only benefit mangers' newly granted options, but also affect managers' profit from outstanding options. Therefore, we develop the following hypothesis:

Hypothesis 1: Accrual-based earnings management is negatively associated with incentives from newly granted options, and positively associated with incentives from unvested outstanding options. However, there is no significant association between accrual-based earnings management and incentives from vested options.

Next, we want to examine the effects of vested and unvested outstanding options on the occurrence of option backdating. The practice of using hindsight to select a date in the past as the option grant date is supposed to affect the new option grants only. However, if option backdating correlates with earnings management, incentives from outstanding options can also affect the incidence of option backdating. On the one hand, if managers can easily backdate their options and already secure gains from their option compensation, it is possible that they are less interested in managing earnings. On the other hand, it is also reasonable for self-serving CEOs to take both opportunistic actions at the same time to reap even greater profits from their newly granted options. Especially, those CEOs who engage in backdating are considered to have more influences over board of directors (Collins, Gong, and Li, 2009). Consequently, they can also be more likely to circumvent board monitoring to manipulate earnings for their own benefits. If managers engage in option backdating and downward earnings management at the same time, incentives driven by outstanding options for backdating will be different from those by newly granted options, since they have conflicting effects on earnings management. Specifically, as we test in Hypothesis 1, unvested outstanding options can create a preference for earnings management in a direction opposite to newly granted options, whereas vested outstanding options do not have a significant correlation with earnings management. Therefore, we test the following hypothesis:

Hypothesis 2: The incidence of option backdating is positively associated with incentives from newly granted options, and negatively associated with incentives from unvested outstanding options. However, there is no significant association between the incidence of option backdating and incentives from vested options.

III. SAMPLE AND DATA

To study the effects of CEO option incentives on earnings management and CEO option backdating, we start with a sample of CEOs from the Execucomp database by restricting our attention to nonfinancial and non-utilities firms with available data for CEO compensation, tenure, age, and option portfolio data in the years between 1992 and 2005.

We then collect CEO stock option grant data from Thomson Financial Insider Trading database for our initial sample, after eliminating multiple grants that occur on the same date. The Insider Trading database provides data on insider trading activities reported on SEC forms 3, 4, 5, and 144. Following Collins, Gong, and Li (2006), we label a firm in a given year as a potential backdater if the stock price on at least one grant date in that year falls within the lowest decile over 120 trading days before and 120 trading days after the grant date. A firm in a given year is labeled as a benchmark firm if the stock prices on all grant dates in that year do not fall in the bottom decile of the firm's 240-day stock price distribution.

In addition, we employ accounting data from Compustat, stock return data from the Center for Research in Security Prices (CRSP), board data from the Corporate Library and proxy statements, and institutional ownership data from CDA Spectrum. To be included in the final sample, a firm must have data available from all the above sources for a given year. The requirement results in a sample of 9,568 CEO-year observations, among which 685 CEO years are for the potential backdaters and 8,883 CEO years are for the benchmark firms.

A. Measures of Earnings Management

Following the literature (Dechow, Sloan, and Sweeney, 1995; Bartov, Gul, and Tsui, 2001; Bergstresser and Philippon, 2006; Cornett, Marcus, Tehranian, 2008), we construct total accruals and discretionary accruals as our measures of earnings management.

In order to construct the variable of total accruals, we first calculate earnings before extraordinary items and discontinued operations minus operating cash flows from continuing operations (Cornett, Marcus, Tehranian, 2008). We then divide the number by the previous year's assets to obtain the measure of total accruals (Ratio_ta).

After the calculation of total accruals, we use the modified Jones (1991) model to construct the variable of discretionary accruals. Discretionary accruals equal the difference between total accruals and "normal" accruals. The modified Jones model estimates "normal" accruals as a fraction of lagged assets from the following model:

[TA.sub.jt] / [Assets.sub.jt-1] = [[alpha].sub.o] 1 / [Asssts.sub.jt-1] + [[beta].sub.1] [DELTA][Sales.sub.jt] / [Assets.sub.jt-1] + [[beta].sub.2] [PPE.sub.jt] / [Assets.sub.jt-1] (1)

where [TA.sub.jt] denotes total accruals for firm j in year t, [Assets.sub.jt-1] denotes total assets for firm j in year t-1, [DELTA][Sales.sub.jt] denotes a change in sales for firm j in year t, and PP[E.sub.jt] denotes property, plant, equipment for firm j in year t. We estimate model (1) by using the firms in Compustat with the same two-digit SIC code as our sample firms in each year of the sample period.

Discretionary accruals then are defined as a fraction of assets as

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

where hats denote estimated values from model (1). The inclusion of [DELTA][Receivables.sub.jt] in equation (2) is the "modification" of the Jones (1991) model. This variable attempts to capture the extent to which a change in sales is due to aggressive recognition of questionable sales.

In addition to Ratio_ta and Ratio_da, we also use other variables derived from these two measures of earnings management, in order to capture changes in accruals between years, or to distinguish different directions of accruals. Please see the Appendix for the definition of these variables.

B. Measures of CEO Option Incentives

To examine the effects of the level of option incentives on earnings management and the likelihood of backdating, we use three measures of the pay-for-performance sensitivity: Pps_new, Pps_vest, and Pps_unvest. Pps_new is the pay performance sensitivity of newly granted options, which is the value change in CEO' s newly granted options in the current year for 1000-dollar change in the market value of equity. Pps_vest is the pay performance sensitivity of exercisable outstanding options that are granted from previous years and CEOs retain in their portfolios. Similarly, Pps_unvest is the pay performance sensitivity of unexercisable outstanding options.

To capture the value of option sensitivity, we begin by calculating the partial derivative of individual stock option with respect to one-dollar change in share price (the Black and Scholes (1973) hedge ratio with dividends, i.e., delta), times the proportion of shares represented by executive option award (see, Yermack, 1995). The risk-free rate is the interest rate on seven-year constant-maturity Treasury bond, obtained from the website of the Federal Reserve Bank of St. Louis, and the volatility (i.e., Volat in the Appendix) is the standard deviation of stock price over the prior sixty months. For previously granted options, we follow Core and Guay's methodology (2002) to estimate the average exercise prices.

C. Measures of Other Governance Variables, Firm Characteristics, and CEO Characteristics

In order to examine the effects of CEO option incentives on the incidence of backdating and earnings management, we also control for various firm characteristics, CEO characteristics, and other governance characteristics such as board characteristics, institutional ownership, and CEO ownership, by following the backdating and earnings management literature. The Appendix defines the above variables in details.

In the following analysis, we winsorize all the variables except Ceoown at the top and bottom 1% of the observations, in order to mitigate the inordinate influence of extreme values.

IV. EMPIRICAL ANALYSIS AND RESULTS

Table 1 presents the descriptive statistics of the key variables for our sample. The five panels provide information on earnings management, CEO option incentives, other governance variables, firm characteristics, and CEO characteristics.

In addition, Table 2 reports correlations between our key variables. Both the correlation matrix and variance influence factors (VIF) do not indicate problematic multicollinearity for our regression models in the following sections.

A. The Effects of Various Option Elements on Earnings Management

We run regressions of accrual-based earning management against variables for incentives from various option elements and the other variables defined in Section III. We estimate the following equation:

[Accruals.sub.it] = [[beta].sub.0] + [[beta].sub.1] [CEOOptionIncentives.sub.it] + [[beta].sub.2] [OtherGovernance.sub.it] + [[beta].sub.3] [FirmCharacteristics.sub.it] + [[beta].sub.4] [CEOCharacteristics.sub.it] + [[epsilon].sub.it] (3)

Variables for option incentives (i.e., Ppsnew, Ppsvest, and Ppsunvest) will enter one at a time before all of them are put together to test their conflicting effects.

Table 3 presents the results. (2) When only Pps_new is included, higher incentives from newly granted options are significantly related to decreasing total accruals and greater decreases in total accruals from the prior three-year average. When only Pps_vest is included, incentives from vested outstanding options are not significantly related to any accrual measures. When only Pps_unvest is included, on the one hand, higher incentives from unvested outstanding options are significantly correlated with increasing total accruals, greater increases in total accruals from the prior three-year average, and lower likelihood of negative total accruals. On the other hand, higher incentives from unvested options are also related to lower discretionary accruals and greater decreases in discretionary accruals from the prior three-year average. However, the latter relation is only marginally significant. When all the option incentives are included as in Table 3, Pps_new is significantly related to decreasing total accruals and greater decreases in total accruals from the prior three-year average. Pps_unvest is significantly related to increasing total accruals, greater increase in total accruals from the prior three-year average, and lower likelihood of negative total accruals. At the same time, incentives from vested outstanding options continue to have no significant relationship with any accrual measures. Therefore, the results suggest that on average high incentives from newly granted options decrease earnings and high incentives from unvested options increase earnings via accruals.

The results support our Hypothesis 1. When CEOs are granted new options, downward earnings management can help managers profit from their option compensation. In addition, when CEOs have higher incentives from unvested outstanding options, upward earnings management can help them to sell shares to diversify risk associated with their option portfolios.

B. The Effects of Various Option Elements on Option Backdating

1. Univariate Analysis between Potential Backdaters and the Benchmark Sample

Table 4 reports the univariate comparison for the key variables between potential backdaters (Backdate = 1) and the benchmark sample (Backdate = 0). As can be seen from Panel A of Table 4, potential backdaters generally have more negative values and greater decreases in accruals. For instance, potential backdaters have average (median) total accrual of -7.50% (-6.20%) of assets, which is significantly smaller than -6.60% (-5.80%) for the benchmark firms. Potential backdaters also have significantly higher value of the dummy variable for negative total accruals, which indicates more negative values. In addition, compared to the benchmark firms, potential backdaters have greater decreases in accruals from the previous three-year average. Specifically, in terms of discretionary accruals, the mean (median) deviation from the previous three-year average value is -8.9% (-0.8%) of assets, which is significantly smaller than 5.3% (-0.1%) of assets for the benchmark firms.

In addition, according to Panel B, on average, CEOs in potential backdaters have higher incentives from newly granted options (average value of $3.260 vs. $1.780 per $1,000 change in market value of equity, and median value of $1.740 vs. $0.780 per $1,000 change in market value of equity) and higher incentives from vested option holdings. Furthermore, according to Panel C and Panel D, potential backdaters have more institutional shareholders, higher institutional ownership, fewer directors on the boards, lower leverage, higher growth, and larger stock return volatility than the benchmark firms. Finally, Panel E shows that CEOs who are in potential backdaters are younger than those in the benchmark firms, but they are not significantly different from each other in terms of the length of time on tenure.

2. Logit Analysis of the Effects of Various Option Elements on Option Backdating

To further analyze the relation between incentives from various option elements and option backdating, we estimate a logit regression model, where the dependent variable (Backdater) equals one if the firm is a potential backdater, and zero if the firm is a non-backdater (i.e. the benchmark firm), and the independent variables are variables for option incentives and the other variables defined in Section III. We estimate the following equation:

P(Backdater) = [[beta].sub.0] + [[beta].suB.1][CEOOptionIncentives.sub.it] + [[beta].sub.2][OtherGovernance.sub.it] + [[beta].sub.3][FirmCharacteristics.sub.it] + [[beta].sub.4][CEOCharacteristics.sub.it] + [[epsilon].sub.it] (4)

Similar as Section IV.A, variables for option incentives (i.e. Ppsnew, Ppsvest, and Ppsunvest) will enter one at a time before all of them are put together to test their conflicting effects.

Table 5 presents the results. It shows that the sensitivity of newly granted option values to changes in stock price is significantly and positively related to the probability of firms' option backdating. The result suggests that as the option-based compensation becomes more important, and greater benefit can be generated from option backdating, CEOs have stronger incentives to engage in option backdating. This is consistent with Minnick and Zhao (2009) and Collins, Gong, and Li (2009). In addition, higher incentives from unvested options are related to lower probability of backdating, whereas incentives from vested options do not have a significant effect. The results are consistent with our Hypothesis 2.

According to Section IV.A, managers with high incentives from unvested options tend to manage earnings upward, which is in the direction opposite to the preference created by newly granted options. Therefore, it is possible that backdating is correlated with income-decreasing earnings management, which creates disincentives for backdating by those managers with high incentives from unvested outstanding options. But managers with high incentives from vested outstanding options do not have a similar disincentive for backdating, since vested options are shown to be insignificantly correlated with earnings management. We are going to test the implied correlation between backdating and downward earnings management in Section IV.C and IV.D.

Furthermore, among the governance variables, the number of institutional investors is positively related to the likelihood of backdating. It is possible that more institutional investors become interested in purchasing the stock in the year of backdating when stock price goes down due to income-decreasing earnings management. In addition, higher CEO stock ownership is related to a lower likelihood of backdating. This is consistent with the notion that higher ownership improves the incentive alignment between CEOs and shareholders, making it less likely that backdating will occur (Collins, Gong, and Li, 2009). However, the proposed correlation between backdating and income-decreasing earnings management can also possibly explain the negative relation between CEO stock ownership and the occurrence of backdating, as the value of CEO stock holdings will reduce when stock price goes down due to downward earnings management. Among the variables of firm characteristics, we find that greater stock return volatility increases the likelihood of backdating. The result is consistent with Collins, Gong, and Li (2009) and Minnick and Zhao (2009).

C. The Effects of Earnings Management on Option Backdating

In order to confirm the correlation between backdating and income-decreasing earnings management, in this section, we run logit regressions of the likelihood of option backdating against both variables for earnings management and variables for option incentives, as well as other control variables. We estimate the following equation:

P(Backdater) = [[beta].sub.0] + [[beta].sub.1][EarningsManagement.sub.it] + [[beta].sub.2][CEOOptionIncentives.sub.it] + [[beta].sub.3][OtherGovernance.sub.it] + [[beta].sub.4][FirmCharacteristics.sub.it] + [[beta].sub.5][CEOCharacteristics.sub.it] + [[epsilon].sub.it] (5)

Table 6 reports the results. The incidence of option backdating is shown to be positively related to income-decreasing earnings management. As can be seen from regression (3) and (4), backdaters have greater decreases in both discretionary accruals and total accruals from the prior three-year average than the benchmark firms. Higher likelihood of backdating is also shown to be associated with more negative total accruals, according to regression (5). The results suggest that managers engaging in option backdating are also likely to use income-decreasing earnings management to get more favorable price for their new stock option grants as well.

In addition, the effects of various option elements continue to be consistent with our Hypothesis 2. That is, higher probability of backdating is related to greater incentives from newly granted options and smaller incentives from unvested options, whereas incentives from vested options do not have a significant effect on backdating.

D. The Effect of Persistent Unvested Option Incentives vs. Non-Persistent Unvested Option Incentives on Option Backdating

The evidence that managers with high incentives from unvested options are less likely to engage in backdating is consistent with their incentives to manage earnings upward (Stein, 1989; Cheng and Warfield, 2005). In particular, if CEOs have unusually high incentives from their unvested options in a given year, they are likely to sell stocks to diversify risk. As a result, they have incentives to manage earnings upward to beat up short-term stock price for selling purpose. However, if CEOs have highly persistent incentives from their unvested options, it indicates that they are willing to bear the risk involved in their compensation. As a result, they have lower incentives for upward earnings management because they have a less motivation to sell stocks. Consequently, if option backdating is associated with income-decreasing earnings management, the managers with highly persistent incentives from unvested options should be less likely to avoid backdating, compared to those with non-persistent incentives from unvested options. Therefore, in this section, we distinguish between persistent and non-persistent unvested option incentives and investigate their different effects on backdating.

To distinguish between persistent and non-persistent unvested option incentives, we first rank the pay-for-performance sensitivity of unvested options (Pps_unvest) in each year into percentiles. We then calculate the average ranking for each firm over the sample period. To increase the reliability of the classification, we only consider those firms that have at least five years of data for Pps_unvest over the sample period. If a firm's average ranking is above the top 25% or below the bottom 25% of the average ranking distribution, then that firm is regarded as having highly persistent incentives from unvested options (consistently high or consistently low Pps_unvest, respectively). Under this approach, about 60% of the firm-years exhibit highly persistent Pps_unvest. We use a dummy variable, Recurring, to denote these firm-years. (3) We then add an interaction of this dummy with Pps_unvest to the logit regressions for the likelihood of backdating to capture the incremental effect for the firms with highly persistent incentives from unvested options.

Table 7 reports the results. As in previous sections, pay-for-performance sensitivity of unvested options continues to be related to a lower probability of backdating. However, the coefficient on the interaction term does not have any significant effects on the incidence of backdating. The results suggest that the negative relation between Pps_unvest and the likelihood of backdating is largely driven by the managers with less persistent incentives from their unvested options. These managers have incentives to manage earnings upward to beat up short-term stock price when they sell stocks to diversify risk. As backdating is shown to be associated with income-decreasing earnings management, they have a disincentive for backdating. Differently, those managers with highly persistent incentives from their unvested options have lower incentives for upward earnings management because they are less likely to sell stocks for risk diversification purpose. As a result, they have a less motivation to reduce the incidence of backdating as well.

V. CONCLUSION

In this paper, we study the conflicting incentives from the various elements of CEO option portfolios on both earning management and the occurrence of backdating, and investigate the link between these two opportunistic actions to explain the conflicting effects. We find that higher incentives from newly granted options are associated with income-decreasing earnings management and higher incidence of backdating, whereas higher incentives from unvested options are related to income-increasing earnings management and lower incidence of backdating. However, incentives from vested options have an insignificant effect on both earnings management and backdating. We further show that the correlation between downward earnings management and option backdating, and the need for managers to diversify the increased-risk associated with stock-based compensation can clearly explain the effects of CEOs' outstanding options on both earnings management and the occurrence of backdating.

Our study is the first to examine how conflicting incentives associated with newly granted options and outstanding options affect the occurrence of option backdating. Our paper also combines and extends the literature on backdating and option grant related earnings management, by providing new evidence that self-serving CEOs are likely to take both opportunistic actions at the same time when they are granted options. Our findings have implications for corporate executive compensation policies, and are relevant to the current public policy debate regarding options and executive pay in general.
Appendix
Definition of the key variables

 Variable Definition

 Panel A: Earnings management

Ratio_ta Total accruals/the previous year's assets
Ratio_da Discretionary accruals/the previous year's assets
Dev_rta Deviation of Ratiota from the prior 3-year average
Dev_rda Deviation of Ratioda from the prior 3-year average
Neg_ta Dummy equal to unity if total accruals are negative
Neg_da Dummy equal to unity if discretionary accruals
 are negative

 Panel B: CEO compensation

Pps_new The value change in CEO's newly granted options
 per 1000-dollar change in the market value of equity

Pps_vest The value change in CEO's outstanding exercisable
 options per 1000-dollar change in the market
 value of equity
Pps_unvest The value change in CEO's outstanding unexercisable
 options per 1000-dollar change in the
 market value of equity
Recurring Dummy equal to unity if average ranking of Pps_unvest
 over the sample period is in top or bottom 25%
 of the average ranking distribution

 Panel C: Other governance variables

Ceoown CEOs' holdings of common shares/total shares outstanding
Ln_ninst log(number of institutional investors)
Instown Total shares held by institutional investors/total
 shares outstanding
Ln_bdsize Log(the number of directors on the board)
Pctbdind The proportion of outsiders on the board
Duality Dummy equal to unity if the CEO is also the
 chairman of the board

 Panel D: Firm characteristics

Mve Market value of equity
Lev Book value of debt/(book value of debt + market
 value of equity)
Nisd The standard deviation of net income during the
 three-year period from two
 years before to the current year
Q Market value of assets/book value of assets
Ln_ta Log(total assets in million dollars)
Volat Standard deviation of daily stock returns over the
 prior 5 years

 Panel E: CEO characteristic

Age CEO's age
Ceotenure The length of time in whole years since the CEO
 was on the position

 Panel F: Option backdating

Backdate Dummy equal to unity if the firm is a potential
 backdater in that year


REFERENCES

Aboody, D., and R. Kasznik, 2000, "CEO Stock Option Awards and the Timing of Corporate Voluntary Disclosures, "Journal of Accounting and Economics, 29, 73-100.

Arthur, N., 2001, "Board Composition as the Outcome of an Internal Bargaining Process: Empirical Evidence," Journal of Corporate Finance, 7, 307-340.

Baker, T., D. Collins, and A. Reitenga, 2003, "Stock Option Compensation and Earnings Management Incentives," Journal of Accounting, Auditing and Finance, 18 (Fall), 557-582.

Bartov, E., F. Gul, and J. Tsui, 2001, "Discretionary-Accrual Models and Audit Qualifications," Journal of Accounting and Economics, 30, 421-452.

Bergstresser, D., and T. Philippon, 2006, "CEO Incentives and Earnings Management," Journal of Financial Economics, 80, 511-529.

Black, F., and M. Scholes, 1973, "The Pricing of Options and Corporate Liabilities," Journal of Political Economy, 81, 637-654.

Boone, A., L. Field, J. Karpoff, and C. Raheja, 2007, "The Determinants of Board Size and Composition: an Empirical Analysis," Journal of Financial Economics, 85(1), 66-101.

Chauvin, K., and C. Shenoy, 2001, "Stock Price Decreases Prior to Executive Stock Option Grants," Journal of Corporate Finance, 7, 53-76.

Cheng, Q., and T. Warfield, 2005, "Equity Incentives and Earnings Management," The Accounting Review, 80 (2), 441-476.

Collins, D., G. Gong, and H. Li, 2006, "The Effect of the Sarbanes-Oxley Act on the Timing Manipulation of CEO Option Awards," Working Paper, University of Iowa.

Collins, D., G. Gong, and H. Li, 2009, "Corporate Governance and Backdating of Executive Stock Options," Contemporary Accounting Research, 26(2), 403-445.

Core, J. E., and W. Guay, 2002, "Estimating the Value of Employee Stock Option Portfolios and Their Sensitivities To Price and Volatility," Journal of Accounting Research, 40, 613-630.

Cornett, M., A. Marcus, and H. Tehranian, 2008, "Corporate Governance and Pay-For-Performance: The Impact of Earnings Management," Journal of Financial Economics, 87, 357-373.

Dechow, P., R. Sloan, and A. Sweeney, 1995, "Detecting Earnings Management," Accounting Review, 70, 193-226.

Del Guercio, D., and J. Hawkins, 1999, "The Motivation and Impact of Pension Fund Activism," Journal of Financial Economics, 52, 293-340.

Finkelstein, S., and D. C. Hambrick, 1988, "Chief Executive Compensation: A Synthesis and Reconsideration," Strategic Management Journal, 9, 543-558.

Gibbons, R. and K. Murphy, 1992, "Optimal Incentive Contracts in The Presence of Career Concerns: Theory and Evidence," Journal of Political Economy, 100(3), 468-505.

Grinstein Y., and P. Hribar, 2004, "CEO Compensation and Incentives: Evidence from M&A Bonuses," Journal of Financial Economics, 73, 119-143.

Heron, R., and E. Lie, 2007, "Does Backdating Explain the Stock Price Pattern Around Executive Option Grants?" Journal of Financial Economics, 83, 271-295.

Jensen, M. C., 1993, "The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems," Journal of Finance, 48, 831-880.

Jensen, M., and W. Meckling, 1976, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Journal of Financial Economics, 3, 305-360.

Jensen, M., and K. Murphy, 1990, "Performance Pay and Top-Management Incentives," Journal of Political Economy, 98, 225-264.

Jones, J. 1991, "Earnings Management during Import Relief Investigations," Journal of Accounting Research, 29, 193-228.

Lie, E., 2005, "On the Timing of CEO Stock Option Awards," Management Science, 51, 802-812.

McConnell, J. J., and H. Servaes, 1990, "Additional Evidence on Equity Ownership and Corporate Value," Journal of Financial Economics, 27, 595-612.

McAnally, M. L., A. Srivastava, and C. Weaver, 2008, "Executive Stock Options, Missed Earnings Targets and Earnings Management," The Accounting Review, 83 (1), 185-216.

Minnick K., and M. Zhao, 2009, "Backdating and Director Incentives: Money or Reputation," Journal of Financial Research, 32(4), 449-477.

Morck, R., A. Shleifer, and R.W. Vishny, 1988, "Management Ownership and Market Valuation," Journal of Financial Economics, 20, 293-315.

Narayanan, M. P., and H.N. Seyhun, 2006, "Effect of Sarbanes-Oxley Act on the Influence of Executive Compensation," Working paper, University of Michigan.

Nesbitt, S. L., 1994, "Long-Term Rewards from Shareholder Activism: A Study of the 'CalPERS Effect'," Journal of Applied Corporate Finance, 6, 75-80.

Ofek, E., and D. Yermack, 2000, "Taking Stock: Equity-Based Compensation and the Evolution of Managerial Ownership," Journal of Finance, 55(3), 1367-1384.

Smith, M., 1996, "Shareholder Activism by Institutional Investors: Evidence from CalPERS," Journal of Finance, 51, 227-252.

Srivastava, A., 2005, "Earning Management Choices when Stock Option Incentives Conflicts," Working paper, Texas A&M University

Stein, J. C., 1989, "Efficient Capital Markets, Inefficient Firms: A Model of Myopic Corporation Behavior," Quarterly Journal of Economics, 104 (4), 655-669.

Sweeney, A. P., 1994, "Debt-Covenant Violations and Managers' Accounting Response," Journal of Accounting and Economics, 17(3), 281-308.

Warfield, T. D., J. J. Wild, and K.L. Wild, 1995, "Managerial Ownership, Accounting Choices, and Informativeness of Earnings," Journal of Accounting and Economics, 20(1), 61-91.

Watts, R. L. and J. L. Zimmerman, 1986, Positive Accounting Theory, Englewood Cliffs, NJ: Prentice Hall.

Yermack, D., 1995, "Do Corporations Award CEO Stock Options Effectively?" Journal of Financial Economics, 39, 237-269.

Yermack, D., 1997, "Good Timing: CEO Stock Option Awards and Company News Announcements," Journal of Finance, 52, 449-476.

ENDNOTES

(1.) The effect of vested options on earnings management can also occur through option exercises. However, the choice of earnings management associated with vested options through option exercises differ across firms and over time and can be inconsistent. Please refer to Section II.B. Hypothesis Development for details.

(2.) Table 3 only shows the results when all the option incentive variables are included. The regression results for Pps_new, Pps_vest, and Pps_unvest separately are available upon request.

(3.) Our empirical results do not essentially change if we only include those firms with consistently high Ppsunvest for Recurring = 1.

Hatice Uzun (a), Yudan Zheng (b) *

(a) School of Business, Public Administration and Information Sciences Long Island University, H700, 1 University Plaza, Brooklyn, NY 11201 hatice.uzun@liu.edu

(b) School of Business, Public Administration and Information Sciences Long Island University, H700, 1 University Plaza, Brooklyn, NY 11201 yudan.zheng@liu.edu

* The authors thank K.C. Chen (the editor), an anonymous referee, and the participants at the FMA 2010 annual meeting in New York, Northeast Business & Economics Association 37th annual meeting in Morristown, NJ, and the Wharton School 2010 Global Conference on Corporate Governance and the Global Financial Crisis, Pennsylvania.
Table 1
Summary statistics

The table reports the descriptive statistics for the
key variables for the sample over 1992 through
2005. The variables are defined in the Appendix. The
major variables are winsorized at the top
and bottom 1% of the observations. Mve and Nisd are
in million dollars.

Variable Obs. Mean Median Std. Dev.

 Panel A: Earnings management

Ratio_ta 9568 -0.067 -0.058 0.085
Ratio_da 9568 0.149 0.011 1.777
Dev_rta 9381 -0.005 0.000 0.089
Dev_rda 9288 0.043 -0.001 2.148
Neg_ta 9568 0.865 1.000 0.342
Neg_da 9568 0.448 0.000 0.497

 Panel B: CEO compensation

Pps_new 9568 1.8800 0.8400 3.216
Pps_vest 9568 6.2500 3.5500 8.014
Pps unvest 9568 2.2100 0.9500 3.416

 Panel C: Other governance variables

Ceoown 9568 2.344% 0.000% 6.172%
Ln_ninst 9568 4.978 4.949 0.786
Instown 9568 63.844% 65.853% 18.081%
Ln_bdsize 9568 2.185 2.197 0.279
Pctbdind 9568 63.767% 66.667% 17.458%
Duality 9568 0.657 1.000 0.475

 Panel D: Firm characteristics

Mve 9568 6370.577 1328.494 16,156,293,600
Lev 9568 0.200 0.148 0.197
Nisd 9568 135.874 30.832 319,823,421
Q 9568 2.126 1.649 2.126
Ln_ta 9568 7.140 6.978 1.503
Volat 9568 0.428 0.382 0.183

 Panel E: CEO characteristics

Age 9568 55.613 56.000 7.292
Ceotenure 9568 7.411 5.000 7.402

Variable 10% 90%

 Panel A: Earnings management

Ratio_ta -0.154 0.012
Ratio_da -0.189 0.508
Dev_rta -0.095 0.078
Dev_rda -0.691 0.527
Neg_ta 0.000 1.000
Neg_da 0.000 1.000

 Panel B: CEO compensation

Pps_new 0.000 4.711
Pps_vest 0.000 15.811
Pps unvest 0.000 6.019

 Panel C: Other governance variables

Ceoown 0.000% 6.800%
Ln_ninst 4.025 5.994
Instown 38.331% 85.965%
Ln_bdsize 1.792 2.565
Pctbdind 40.000% 85.714%
Duality 0.000 1.000

 Panel D: Firm characteristics

Mve 204,947,000 13,837,100,000
Lev 0.000 0.487
Nisd 4,899,060 326,743,000
Q 1.035 3.661
Ln_ta 5.340 9.260
Volat 0.200 0.700

 Panel E: CEO characteristics

Age 46.000 64.000
Ceotenure 1.000 18.000

Table 2
Correlation matrix

The table reports the correlation matrix for the key
variables for the sample over 1992 through 2005.
The variables are defined in the
Appendix.

 Ratio_ta Ratio_da Pps_new Pps_vest

Ratio ta 1.00
Ratio da 0.04 1.00
Pps new -0.08 0.01 1.00
Pps vest -0.01 0.00 -0.01 1.00
Pps unvest -0.01 0.00 -0.01 0.99
Ceoown 0.04 -0.01 -0.15 0.00
Ln ninst 0.03 0.00 0.15 0.01
Instown 0.05 0.01 0.10 0.00
Ln bdsize 0.07 -0.01 -0.05 0.00
Pctbdind -0.02 0.00 0.08 0.00
Duality 0.04 0.00 -0.04 0.00
Mve -0.01 0.00 0.08 0.00
Lev -0.03 0.00 -0.12 -0.01
Nisd -0.09 0.01 0.06 0.00
Q -0.11 -0.01 0.18 0.00
Ln_ta 0.03 0.00 0.04 0.01
Volat -0.18 0.06 0.21 0.02
Age 0.09 0.00 -0.17 -0.01
Ceotenure 0.06 -0.01 -0.11 0.00

 Pps_unvest Ceoown Ln_ninst Instown

Ratio ta
Ratio da
Pps new
Pps vest
Pps unvest 1.00
Ceoown 0.00 1.00
Ln ninst 0.01 -0.21 1.00
Instown 0.00 -0.26 0.43 1.00
Ln bdsize 0.00 -0.18 0.40 -0.05
Pctbdind 0.00 -0.24 0.20 0.24
Duality 0.00 0.13 0.15 0.04
Mve 0.00 -0.05 0.56 -0.02
Lev 0.00 -0.05 -0.15 -0.08
Nisd 0.00 -0.08 0.43 -0.02
Q 0.00 0.05 0.20 0.01
Ln_ta 0.01 -0.16 0.79 0.21
Volat 0.02 0.03 -0.29 -0.05
Age -0.01 0.13 0.06 -0.01
Ceotenure 0.00 0.39 -0.04 -0.05

 Ln_bdsize Pctbdind Duality Mve

Ratio ta
Ratio da
Pps new
Pps vest
Pps unvest
Ceoown
Ln ninst
Instown
Ln bdsize 1.00
Pctbdind 0.09 1.00
Duality 0.11 0.11 1.00
Mve 0.29 0.07 0.09 1.00
Lev 0.19 0.03 0.07 -0.10
Nisd 0.25 0.10 0.07 0.56
Q -0.12 -0.03 -0.03 0.20
Ln_ta 0.55 0.17 0.20 0.55
Volat -0.43 -0.03 -0.17 -0.19
Age 0.13 -0.01 0.28 0.05
Ceotenure -0.08 -0.17 0.25 -0.02

 Lev Nisd Q Ln_ta Volat Age

Ratio ta
Ratio da
Pps new
Pps vest
Pps unvest
Ceoown
Ln ninst
Instown
Ln bdsize
Pctbdind
Duality
Mve
Lev 1.00
Nisd 0.12 1.00
Q -0.40 -0.03 1.00
Ln_ta 0.28 0.55 -0.19 1.00
Volat -0.01 -0.02 0.10 -0.38 1.00
Age 0.07 0.02 -0.12 0.15 -0.23 1.00
Ceotenure -0.07 -0.08 0.02 -0.06 -0.04 0.40

 Ceotenure

Ratio ta
Ratio da
Pps new
Pps vest
Pps unvest
Ceoown
Ln ninst
Instown
Ln bdsize
Pctbdind
Duality
Mve
Lev
Nisd
Q
Ln_ta
Volat
Age
Ceotenure 1.00

Table 3
OLS regression of the effects of various
option elements on earnings management

The table shows the coefficients from OLS
regressions of incentives from various option
elements on earnings management. The dependent
variables and the independent variables are
defined in the Appendix. The major variables
are winsorized at the top and bottom 1% of the
observations. The variance inflation factors (VIF)
do not indicate problematic multicollinearity.
For each coefficient, p-values for t-tests are
provided in parentheses. *, **, *** denote
significance at the 0.10, 0.05, and 0.01,
levels respectively.

 Dep. Variables

Indep. Ratio_ta Ratio_da
Variables

 Coeff. P-value Coeff. P-value

 CEO option incentives

Pps_new -0.002 *** (0.000) 0.002 (0.692)
Pps_vest 0.000 (0.282) -0.002 (0.451)
Pps_unvest 0.001** (0.022) -0.007 (0.132)

 Other governance variables

Ceoown 0.041 ** (0.016) 0.190 (0.534)
Ln_ninst 0.004 * (0.057) -0.041 (0.269)
Instown 0.030 *** (0.000) 0.338 *** (0.007)
Ln_bdsize 0.021 *** (0.000) -0.140 * (0.064)
Pctbdind -0.018 *** (0.001) 0.288 *** (0.004)
Duality 0.002 (0.384) -0.047 (0.239)

 Firm characteristics

Mve 0.443 *** (0.000) 0.075 (0.966)
Lev -0.001 (0.846) -0.034 (0.745)
Nisd -39.500 *** (0.000) 63.400 (0.462)
Q -0.005 *** (0.000) -0.013 (0.277)

 CEO characteristics

Age 0.001 *** (0.000) -0.004 (0.151)
Ceotenure 0.000 (0.489) 0.002 (0.444)
Intercept -0.162 (0.000) 0.531 (0.020)

# of obs. 9904 9568
F statistics 13.07 2.06
Prob. 0.000 0.009

 Dep. Variables

Indep. Dev_rta Dev_rda
Variables

 Coeff. P-value Coeff. P-value

 CEO option incentives

Pps_new -0.001 *** (0.012) 0.004 (0.584)
Pps_vest 0.000 (0.189) 0.000 (0.968)
Pps_unvest 0.001 ** (0.046) -0.009 (0.107)

 Other governance variables

Ceoown 0.031 * (0.076) 0.424 (0.340)
Ln_ninst 0.006 *** (0.002) -0.057 (0.201)
Instown 0.018 *** (0.014) 0.254 (0.105)
Ln_bdsize 0.013 *** (0.006) -0.043 (0.643)
Pctbdind 0.003 (0.649) 0.238 * (0.059)
Duality -0.001 (0.562) -0.07 (0.172)

 Firm characteristics

Mve -0.113 (0.213) 1.170 (0.539)
Lev -0.010* (0.084) 0.129 (0.301)
Nisd 2.89 (0.635) 14.700 (0.876)
Q 0.001 (0.204) -0.009 (0.571)

 CEO characteristics

Age 0.000 (0.657) -0.002 (0.488
Ceotenure 0.000 * (0.055) 0.003 (0.408
Intercept -0.078 (0.000) 0.242 (0.389

# of obs. 9713 9288
F statistics 6.060 1.04
Prob. 0.000 0.409

 Dep. Variables

Indep. Neg_ta Neg_da
Variables

 Coeff. P-value Coeff. P-value

 CEO option incentives

Pps_new 0.001 (0.543) 0.002 (0.196)
Pps_vest 0.000 (0.737) -0.001 (0.367)
Pps_unvest -0.003 *** (0.004) 0.001 (0.545)

 Other governance variables

Ceoown -0.033 (0.635) 0.150 (0.120)
Ln_ninst 0.053 *** (0.000) 0.006 (0.579)
Instown -0.056 ** (0.016) -0.040 (0.255)
Ln_bdsize 0.010 (0.514) -0.031 (0.169)
Pctbdind 0.091 *** (0.000) 0.022 (0.476)
Duality -0.012 (0.140) -0.013 (0.258)

 Firm characteristics

Mve -0.115 (0.675) 0.238 (0.611)
Lev 0.054 *** (0.008) -0.015 (0.625)
Nisd -28.200 ** (0.023) 21.300 (0.307)
Q -0.012 *** (0.000) -0.010 *** (0.010)

 CEO characteristics

Age 0.000 (0.818) -0.002 ** (0.047)
Ceotenure -0.001 (0.214) 0.000 (0.939)
Intercept 0.583 (0.000) 0.608 (0.000)

# of obs. 9904 9568
F statistics 11.96 1.69
Prob. 0.000 0.046

Table 4
Univariate comparison of backdaters
vs. non-backdaters

The table compares the key variables between
the potential backdaters and the benchmark firms
over 1992 through 2005. The variables are defined
in the Appendix. The major variables are winsorized
at the top and bottom 1% of the observations. Mve
and Nisd are in million dollars. We use t-test and
Wilcoxon rank-sum test to determine mean and median
difference between the two sub-samples respectively.
*, **, *** denote significance at the 0.10, 0.05,
and 0.01 levels.

 Variable Potential Backdaters Benchmark

 Obs. Mean Median Obs. Mean Median

 Panel A: Earnings management

Ratio_ta 685 -0.075 -0.062 8883 -0.066 -0.058
Ratio_da 685 0.095 0.011 8883 0.153 0.011
Dev_rta 677 -0.014 -0.003 8704 -0.004 0
Dev_rda 665 -0.089 -0.008 8623 0.053 -0.001
Neg_ta 685 0.904 1 8883 0.862 1
Neg_da 685 0.439 0 8883 0.449 0

 Panel B: CEO option incentives

Pps_new 685 3.26 1.74 8883 1.78 0.78
Pps_vest 685 6.72 4.13 8883 6.21 3.53
Pps_unvest 685 2.12 0.99 8883 2.22 0.95

 Panel C: Other governance variables

Ceoown 685 1.88% 0.02% 8883 2.38% 0.00%
Ln_ninst 685 5.029 4.977 8883 4.975 4.949

Instown 685 65.98% 68.53% 8883 63.68% 65.61%
Ln_bdsize 685 2.142 2.197 8883 2.188 2.197
Pctbdind 685 63.91% 66.67% 8883 63.76% 66.67%
Duality 685 0.653 1 8883 0.657 1

 Panel D: Firm characteristics

Mve 685 6046.795 1397.663 8883 6395.545 1324.14
Lev 685 0.18 0.132 8883 0.202 0.15
Nisd 685 136.955 33.215 8883 135.79 30.607
Q 685 2.279 1.735 8883 2.114 1.643
Ln_ta 685 7.114 6.975 8883 7.142 6.98
Volat 685 0.483 0.445 8883 0.424 0.379

 Panel E: CEO characteristics

Age 685 54.809 55 8883 55.675 56
Ceotenure 685 7.045 5 8883 7.439 5

 Variable p-values for
 difference tests

 Mean Median

 Panel A: Earnings management

Ratio_ta 0.008 *** 0.010 ***
Ratio_da 0.411 0.635
Dev_rta 0.006 *** 0.033 **
Dev_rda 0.101 * 0.048 **
Neg_ta 0.000 *** 0.002 ***
Neg_da 0.641 0.641

 Panel B: CEO option incentives

Pps_new 0.000 *** <.0001 ***
Pps_vest 0.108 0.035 **
Pps_unvest 0.47 0.776

Panel C: Other governance variables

Ceoown 0.015 ** 0.814
Ln_ninst 0.064 * 0.055 *

Instown 0.001 *** 0.001 ***
Ln_bdsize 0.000 *** <.0001 ***
Pctbdind 0.82 0.773
Duality 0.814 0.814

 Panel D: Firm characteristics

Mve 0.586 0.203
Lev 0.005 *** 0.003 ***
Nisd 0.927 0.356
Q 0.008 *** 0.005 ***
Ln_ta 0.639 0.721
Volat 0.000 *** <.0001 ***

 Panel E: CEO characteristics

Age 0.003 *** 0.015 **
Ceotenure 0.141 0.612

Table 5
Logit analysis of the effect of various
option elements on the likelihood of backdating

The table shows the coefficients from a logit
regression of incentives from various option
elements on the likelihood of backdating. The
dependent variable equals one if the firm is a
potential backdater, and zero if the firm is a
non-backdater (i.e. the benchmark firm). The
independent variables are defined in the Appendix.
The major variables are winsorized at the top
and bottom 1% of the observations. The models
are estimated by maximum likelihood method.
The variance inflation factors (VIF) do not
indicate problematic multicollinearity. For
each coefficient, p-values for t-tests are
provided in parentheses. *, **, *** denote
significance at the 0.10, 0.05, and 0.01
levels, respectively.

Indep 1 2
Variables

 Coeff. P-value Coeff. P-value

 CEO option incentives

Pps_new 0.083 *** (0.000)
Pps_vest 0.000 (0.981)
Pps_unvest

 Other governance variables

Ceoown -0.017 ** (0.051) -0.019 ** (0.034)
Ln_ninst 0.285 *** (0.002) 0.218 ** (0.019)
Instown 0.300 (0.223) 0.326 (0.186)
Ln_bdsize -0.280 (0.126) -0.363 ** (0.045)
Pctbdind -0.183 (0.444) -0.263 (0.267)
Duality 0.125 (0.165) 0.149 * (0.096)

 Firm characteristics

Ln_ta 0.000 (0.998) -0.012 (0.816)
Volat 1.241 *** (0.000) 1.531 *** (0.000)

 CEO characteristics

Age -0.004 (0.530) -0.007 (0.277)
Ceotenure 0.004 (0.527) 0.001 (0.897)
Intercept -4.070 (0.000) -3.206 (0.000)

# of obs 9944 9943
Chi-square 169.920 94.480
Prob 0.000 0.000

Indep 3 4
Variables

 Coeff. P-value Coeff. P-value

 CEO option incentives

Pps_new 0.083 *** (0.000)
Pps_vest -0.002 (0.712)
Pps_unvest -0.033 *** (0.011) -0.028 ** (0.029)

 Other governance variables

Ceoown -0.020 ** (0.025) -0.018 ** (0.043)
Ln_ninst 0.215 ** (0.020) 0.282 *** (0.003)
Instown 0.362 (0.140) 0.347 (0.163)
Ln_bdsize -0.391 ** (0.031) -0.307 * (0.094)
Pctbdind -0.252 (0.287) -0.172 (0.472)
Duality 0.155 * (0.082) 0.134 (0.139)

 Firm characteristics

Ln_ta -0.024 (0.632) -0.012 (0.810)
Volat 1.575 *** (0.000) 1.281 *** (0.000)

 CEO characteristics

Age -0.008 (0.198) -0.005 (0.422)
Ceotenure 0.001 (0.859) 0.005 (0.460)
Intercept -2.955 (0.000) -3.839 (0.000)

# of obs 9942 9942
Chi-square 101.620 175.710
Prob 0.000 0.000

Table 6
Logit analysis of the effects
of earnings management on the
likelihood of backdating (*, **,
*** significance at the 0.10,
0.05, and 0.01 level)

 1 2

Indep. Coeff. P-value Coeff. P-value
Variables

 Earnings management

Ratio_ta 0.049 (0.910)
Ratio_da -0.030 (0.181)
Dev_rta
Dev_rda
Neg_ta
Neg da

 CEO compensation

Pps_new 0.083 *** (0.000) 0.084 *** (0.000)
Pps_vest -0.002 (0.711) -0.002 (0.655)
Pps_unvest -0.028 ** (0.029) -0.026 ** (0.046)

 Other governance variables

Ceoown -0.018 ** (0.043) -0.013 (0.134)
Ln_ninst 0.281 *** (0.003) 0.271 *** (0.004)
Instown 0.349 (0.163) 0.349 (0.169)
Ln_bdsize -0.307 * (0.094) -0.313 * (0.094)
Pctbdind -0.173 (0.471) -0.147 (0.548)
Duality 0.134 (0.137) 0.117 (0.203)

 Firm characteristics

Ln_ta -0.011 (0.824) -0.003 (0.960)
Volat 1.285 *** (0.000) 1.310 *** (0.000)

 CEO characteristics

Age -0.005 (0.417) -0.005 (0.428)
Ceotenure 0.005 (0.464) 0.001 (0.862)
Intercept -3.834 (0.000) -3.848 (0.000)

# of obs. 9933 9596
Chi-square 175.570 172.830
Prob. 0.000 0.000

 3 4

Indep. Coeff. P-value Coeff. P-value
Variables

 Earnings management

Ratio_ta
Ratio_da
Dev_rta -0.811 ** (0.044)
Dev_rda -0.033 * (0.090)
Neg_ta
Neg da

 CEO compensation

Pps_new 0.083 *** (0.000) 0.083 *** (0.000)
Pps_vest -0.002 (0.696) -0.003 (0.526)
Pps_unvest -0.027 ** (0.042) -0.028 ** (0.042)

 Other governance variables

Ceoown -0.021 ** (-0.021) -0.017 * (0.070)
Ln_ninst 0.295 (0.295) 0.294 *** (0.003)
Instown 0.296 (0.296) 0.228 (0.376)
Ln_bdsize -0.395 (-0.395) -0.427 ** (0.025)
Pctbdind -0.209 (-0.209) -0.149 (0.549)
Duality 0.138 (0.138) 0.113 (0.225)

 Firm characteristics

Ln_ta -0.012 (0.815) -0.011 (0.834)
Volat 1.207 *** (0.000) 1.292 *** (0.000)

 CEO characteristics

Age -0.004 (0.518) -0.003 (0.611)
Ceotenure 0.003 (0.683) -0.003 (0.719)
Intercept -3.654 (0.000) -3.613 (0.000)

# of obs. 9741 9315
Chi-square 174.420 166.630
Prob. 0.000 0.000

 5 6

Indep. Coeff. P-value Coeff. P-value
Variables

 Earnings management

Ratio_ta
Ratio_da
Dev_rta
Dev_rda
Neg_ta 0.311 ** (0.015)
Neg da -0.051 (0.531)

 CEO compensation

Pps_new 0.083 *** (0.000) 0.085 *** (0.000)
Pps_vest -0.002 (0.721) -0.002 (0.651)
Pps_unvest -0.027 ** (0.037) -0.026 ** (0.048)

Other governance variables

Ceoown -0.018 ** (0.046) -0.013 (0.135)
Ln_ninst 0.283 *** (0.002) 0.272 *** (0.004)
Instown 0.356 (0.153) 0.339 (0.182)
Ln_bdsize -0.311 * (0.090) -0.315 * (0.092)
Pctbdind -0.200 (0.404) -0.153 (0.533)
Duality 0.137 (0.130) 0.118 (0.199)

Firm characteristics

Ln_ta -0.018 (0.714) -0.003 (0.960)
Volat 1.255 *** (0.000) 1.289 *** (0.000)

CEO characteristics

Age -0.005 (0.413) -0.005 (0.421)
Ceotenure 0.005 (0.454) 0.001 (0.865)
Intercept -4.035 (0.000) -3.807 (0.000)

# of obs. 9933 9596
Chi-square 181.870 171.380
Prob. 0.000 0.000

Table 7
Logit analysis of the effect of persistent
unvested option incentives vs. non-persistent
unvested option incentives on the likelihood
of backdating

The table shows the coefficients from a logit
regression of the effect of incentives from
persistent vs. non-persistent unvested options
on the likelihood of backdating. The dependent
variable equals one if the firm is a potential
backdater, and zero if the firm is a non-backdater
(i.e. the benchmark firm). The independent variables
are defined in the Appendix. The major variables
are winsorized at the top and bottom 1% of the
observations. The models are estimated by maximum
likelihood method. The variance inflation factors
(VIF) do not indicate problematic multicollinearity.
For each coefficient, p-values for t-tests are
provided in parentheses. *, **, *** denote
significance at the 0.10, 0.05, and 0.01
levels respectively.

 1 2

Indep. Coeff. P-value Coeff. P-value
Variables

 Earnings management

Ratio_da 0.065 (0.884) -0.027 (0.229)
Dev_rta
Dev_rda
Neg_ta
Neg_da

 CEO compensation

Pps_new 0.085 *** (0.000) 0.086 *** (0.000)
Pps_vest -0.001 (0.791) -0.002 (0.712)
Pps_unvest -0.032 ** (0.044) -0.029 * (0.077)
Recurring *
Pps_unvest 0.001 (0.913) -0.001 (0.952)

 Other governance variables

Ceoown -0.021 ** (0.025) -0.016 * (0.085)
Ln_ninst 0.268 *** (0.005) 0.267 *** (0.006)
Instown 0.342 (0.182) 0.324 (0.212)
Ln_bdsize -0.309* (0.101) -0.313 * (0.102)
Pctbdind -0.171 (0.486) -0.122 (0.627)
Duality 0.150 (0.106) 0.134 (0.155)

 Firm characteristics

Ln_ta -0.011 (0.838) 0.007 (0.902)
Volat 1.295 *** (0.000) 1.324 *** (0.000)

 CEO characteristics

Age -0.006 (0.324) -0.006 (0.335)
Ceotenure 0.005 (0.495) 0.001 (0.860)
Intercept -3.698 (0.000) -3.738 (0.000)

# of obs. 9479.000 0.335
Chi-square 171.890 0.860
Prob. 0.000 0.000

 3 4

Indep. Coeff. P-value Coeff. P-value
Variables

 Earnings management

Ratio_da
Dev_rta -0.693 * (0.094)
Dev_rda -0.033 * (0.097)
Neg_ta
Neg_da

 CEO compensation

Pps_new 0.085 *** (0.000) 0.085 *** (0.000)
Pps_vest -0.002 (0.762) -0.003 (0.548)
Pps_unvest -0.030 * (0.060) -0.030 * (0.073)
Recurring *
Pps_unvest 0.001 (0.940) -0.001 (0.951)

 Other governance variables

Ceoown -0.021 ** (0.026) -0.017 * (0.077)
Ln_ninst 0.285 *** (0.003) 0.290 *** (0.003)
Instown 0.296 (0.249) 0.225 (0.394)
Ln_bdsize -0.389 ** (0.040) -0.416 ** (0.033)
Pctbdind -0.198 (0.422) -0.113 (0.659)
Duality 0.150 (0.108) 0.124 (0.192)

 Firm characteristics

Ln_ta -0.013 (0.801) -0.016 (0.770)
Volat 1.223 *** (0.000) 1.307 *** (0.000)

 CEO characteristics

Age -0.005 (0.413) -0.004 (0.493)
Ceotenure 0.002 (0.725) -0.003 (0.723)
Intercept -3.558 (0.000) -3.547 (0.000)

# of obs. 9325.000 8926.000
Chi-square 171.820 164.790
Prob. 0.000 0.000

 5 6

Indep. Coeff. P-value Coeff. P-value
Variables

 Earnings management

Ratio_da
Dev_rta
Dev_rda
Neg_ta 0.327 *** (0.013)
Neg_da -0.062 (0.455)

 CEO compensation

Pps_new 0.084 *** (0.000) 0.086 *** (0.000)
Pps_vest -0.001 (0.794) -0.002 (0.715)
Pps_unvest -0.030 ** (0.054) -0.028 * (0.078)
Recurring *
Pps_unvest 0.001 (0.910) -0.001 (0.953)

 Other governance variables

Ceoown -0.021 ** (0.026) -0.016 * (0.085)
Ln_ninst 0.272 *** (0.004) 0.267 *** (0.006)
Instown 0.349 (0.171) 0.313 (0.229)
Ln_bdsize -0.313 * (0.096) -0.316 * (0.100)
Pctbdind -0.201 (0.414) -0.128 (0.611)
Duality 0.153 * (0.099) 0.135 (0.151)

 Firm characteristics

Ln_ta -0.019 (0.714) -0.006 (0.906)
Volat 1.264 *** (0.000) 1.304 *** (0.000)

 CEO characteristics

Age -0.006 (0.325) -0.006 (0.327)
Ceotenure 0.005 (0.481) 0.001 (0.862)
Intercept -3.918 (0.000) -3.690 (0.000)

# of obs. 9479.000 9168.000
Chi-square 178.460 168.190
Prob. 0.000 0.000
COPYRIGHT 2012 Premier Publishing, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Uzun, Hatice; Zheng, Yudan
Publication:International Journal of Business
Date:Jan 1, 2012
Words:10352
Previous Article:Hard and soft controls: mind the gap!
Next Article:Optimal portfolio of corporate investment and consumption under market closure.
Topics:

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters