Internal corporate governance and the profitability of insider trading

Lili Dai, Renhui Fu, Jun-Koo Kang, and Inmoo Lee*

January 2015

Abstract This paper examines how internal governance systems limit insiders’ ability to profit from their information advantage. We find that compared with insiders of poorer-governed firms, insiders of better-governed firms earn significantly smaller abnormal profits from their sales transactions but not from their purchase transactions. This governance effect is more pronounced in firms with greater litigation risk. Moreover, better-governed firms are more likely to put in place ex-ante preventive measures, such as voluntary insider trading restriction policies, and are more active in taking ex-post disciplinary actions against CEOs who earn large abnormal profits from their sales transactions.

* Lili Dai is from the College of Business and Economics, Australian National University, Australia (lili.dai@anu.edu.au, +61-2-612-59341), Renhui Fu is from Antai College of Economics and Management, Shanghai Jiaotong University, PR China (renhuifu@sjtu.edu.cn, +86-21-52301575), Jun-Koo Kang is from Nanyang Business School, Nanyang Technological University, Singapore, (jkkang@ntu.edu.sg, +65-6790-5662), and Inmoo Lee is from KAIST College of Business, KAIST, Seoul, Korea (inmool@kaist.ac.kr,+82-2-958-3441) and is currently visiting McCombs School of Business, University of Texas at Austin, Austin, USA (+1-512-232-6860). We thank Tim Loughran, David Veenman, and seminar participants at the Australian National University, University of Amsterdam, Erasmus University, KAIST, National Taiwan University, Seoul National University, Queen’s University Belfast, and University of Texas at Austin for their helpful comments. All errors are ours.

Internal corporate governance and the profitability of insider trading

Abstract This paper examines how internal governance systems limit insiders’ ability to profit from their information advantage. We find that compared with insiders of poorer-governed firms, insiders of better-governed firms earn significantly smaller abnormal profits from their sales transactions but not from their purchase transactions. This governance effect is more pronounced in firms with greater litigation risk. Moreover, better-governed firms are more likely to put in place ex-ante preventive measures, such as voluntary insider trading restriction policies, and are more active in taking ex-post disciplinary actions against CEOs who earn large abnormal profits from their sales transactions. JEL Classification: G34; K22 Keywords: Internal corporate governance; Insider purchases; Insider sales; Profitability of

insider trading; Legal risk

1. Introduction

Although insider trading could play a positive role by making the market more efficient

and by helping the firms compensate managers for their successful entrepreneurship (e.g., Manne,

1966; Roulstone, 2003; Piotroski and Roulstone, 2005), a number of previous studies also

emphasize that informed insider trading allows insiders to exploit their information advantage

over other market participants and thus enables them to extract private benefits (e.g., Seyhun

1986; Fishman and Hagerty, 1992; Bettis et al., 2000; Jagolinzer et al., 2011). Consistent with

this view, policymakers have placed various restrictions on insider trading, such as Rule 10b-5 of

the Securities Exchange Act of 1934, the Insider Trading and Securities Fraud Enforcement Act

(ITSFEA), and the Stock Enforcement Remedies and Penny Stock Reform Act (SERPSRA).1

Previous studies also document that firms try to restrict insider trading by adopting various

insider trading policies (Bettis et al., 2000; Jagolinzer et al., 2011). While these studies help us

understand how corporate policies on insider trading restrict insiders’ use of private information,

there is little evidence on how other mechanisms prevent informative insider trading within a

firm. In particular, we have a very limited understanding about the role of internal governance

mechanisms in preventing informative insider trading and whether this role differs depending on

the nature of inside information (i.e., positive versus negative information).

In this paper, we extend the literature on insider trading by examining whether internal

governance systems limit insiders’ ability to profit from their information advantage. Specifically,

we focus on whether the restrictive effect of internal governance on the informativeness of

insider trading differs between sales and purchase transactions. We also examine whether

internal governance systems exert influence on the firms’ adoption of ex-ante mechanisms that

1 See Bhattacharya (2013) for a summary of various views on insider trading. He uses the artifice of a hypothetical trial to present the cases for and against insider trading.

1

restrict future informed insider trading such as insider trading policies (ITPs) and policies that

require general council’s (GC’s) pre-approval (Bettis et al., 2000; Jagolinzer et al., 2011) and

whether they play an ex-post disciplinary role in removing top executives who engage in

informed insider trading.2

There are at least two reasons why well-governed firms are expected to discourage their

insiders from exploiting private information. First, if managers engage in insider trading for their

own benefit at the expense of shareholder interests and if internal governance systems are

designed to align the interests of shareholders with those of the managers, these systems should

effectively limit managers’ incentives and abilities to profit from such trading. Second, due to the

ITSFEA, firms are responsible for employees’ illegal transactions. Previous studies show that

weak internal governance is associated with more securities fraud class actions (Helland and

Sykuta, 2005) and more accounting enforcement actions by the Securities and Exchange

Commission (SEC) (Beasley, 1996; Dechow et al., 1996). Therefore, well-governed firms are

expected to adopt internal governance mechanisms to minimize legal risk arising from informed

insider transactions.

We focus on internal governance systems rather than external governance systems because

they are more likely to be directly responsible for preventing informed insider trading that makes

the firms susceptible to high litigation risk. Walsh and Seward (1990) argue that when internal

governance mechanisms are well designed, external control mechanisms are less likely to be

necessary. Daily et al. (2003) further argue that external mechanisms are typically activated

when internal mechanisms for controlling managerial opportunism have failed. Consistent with

this view, Acharya et al. (2011) emphasize the importance of internal corporate governance in

2 Other examples of disciplinary actions include oral or written warning, suspension, removal of job duties and responsibilities, and reduction in compensation.

2

mitigating shareholder-manager conflicts and show that it reduces firms’ agency problems even

without any external governance mechanisms. Previous studies also show that firms with good

internal governance, such as those with outside-dominated boards (Hermalin and Weisbach,

1998), those with high CEO pay-performance sensitivity (Li and Srinivasan, 2011), and those

with independent large shareholders (Shleifer and Vishny, 1986) are effective at monitoring top

management, suggesting that to protect shareholder wealth, these firms are expected to engage in

various actions that discourage informed insider trading.3

However, the effect of internal governance on insider trading profitability might differ

depending on the transaction types: purchase versus sales. It is shown that sales transactions tend

to be less informative than purchase transactions because insiders have various reasons to sell

their shares that are not related to private information (e.g., Lakonishok and Lee. 2001; Rogers

2008). Moreover, some studies document that insiders do not experience abnormal losses (i.e.,

do not earn abnormal profits) from their sales transactions in recent periods, although they

continue to earn abnormal profits from their purchase transactions (e.g., Jeng et al., 2003;

Jagolinzer et al., 2011). These findings indicate that shareholders’ concerns are lower in insider

sales than in insider purchases, limiting the role of governance mechanisms in restricting insider

sales.

3 In a related study, Ravina and Sapienza (2010) show that market-adjusted abnormal returns earned by executives and independent directors are closely related to firms’ GIM index. Although their study provides evidence on how the quality of a firm’s governance affects the profitability of insider trading, to the extent that the GIM index is constructed using the firm’s antitakeover provisions and thus is not able to explicitly capture the effectiveness of internal governance mechanisms, it does not provide a direct answer to whether internal governance plays a role in preventing insiders from exploiting private information. Moreover, their main research question is not to examine the relation between corporate governance and profits earned by insiders, but to compare the trading performance of independent directors with that of other executives. Recently, using Dutch data, Cziraki et al. (2014) show that insiders are more likely to exploit private information when they cannot reap large private benefits of control due to stronger external corporate governance standards related to anti-shareholder mechanisms. This result suggests that insider trading profits and private benefits of control are substitutes, raising the question of whether external corporate governance mechanisms are able to limit informed insider transactions.

3

However, there are at least three reasons why good internal governance discourages

insiders from exploiting negative private information (i.e., selling). First, previous studies show

that insiders exploit negative private information and earn abnormal profits by selling their

shares prior to the disclosure of such information (e.g., Jagolinzer, 2009; Muller et al., 2012).4

Second, evidence suggests that legal risk associated with insider sales is greater than legal risk

associated with insider purchases (e.g., Cheng and Lo, 2006; Johnson et al., 2007; Rogers 2008)

and that such risk is particularly high if they occur prior to the release of negative earnings news

with no warning about earnings disappointments (Billings and Cedergren, 2015). 5 Finally,

allowing insider sales cannot be viewed as an optimal way to compensate managers for their

success because insider sales are more likely to reflect managerial failure rather than

entrepreneurial success. Therefore, even well-governed firms with optimal compensation

structures are likely to devote efforts to discouraging informed insider sales.

Overall, these arguments suggest that although insiders on average do not earn abnormal

profits from their sales, firms have incentives to design internal governance mechanisms that can

discourage informed insider sales. Additionally, even though insiders earn significantly positive

abnormal profits from their purchase transactions, shareholders may be less concerned about

4 Jagonlinzer (2009) and Muller et al. (2012) point out that even when insiders use pre-planned selling within the Rule 10b5-1, which was promulgated by the SEC in October 2000, they can still take advantage of their private information by timing their disclosures to facilitate the pre-planned sales. 5 Cheng and Lo (2006) show that managers strategically choose disclosure polices and time their purchase transactions but not sales transactions, suggesting that they are aware of potential legal risk involved with exploiting negative private information. Moreover, Rogers (2008) points out that private litigants focus almost exclusively on insider selling cases and shows that managers optimally choose disclosure policies to reduce legal risk before insider sales, but not before insider purchases. In addition, Johnson et al. (2007) show that litigation significantly increases after abnormal insider selling, especially after the Private Securities Litigation Reform Act of 1995, and Billings and Cedergren (2015) find that the probability of being sued significantly increases when insiders engage in insider sales prior to the release of negative earnings news, especially when they do not provide prior warnings or they engage in opportunistic trading. Jagolinzer (2009) and Henderson et al. (2012) further show that insiders exploit their negative private information in a way that minimizes legal risk. They show that to reduce the legal risk of insider sales, insiders strategically use voluntary disclosure of planned trades according to Rule 10b5-1 of the Securities and Exchange Act of 1934. They further show that insiders earn high abnormal returns from their sales after creating legal cover by articulating specific plan details based on Rule 10b5-1.

4

restricting these transactions because they involve lower legal risk, and abnormal profits can be

considered as compensation for managerial success.

There are several mechanisms through which internal governance prevents insiders from

exploiting private information. One possible mechanism is influencing firms to adopt voluntary

corporate policies such as ITPs to restrict insider trading ex ante. Bettis et al. (2000) and

Jagolinzer et al. (2011) show that firms voluntarily adopt ITPs to restrict insider trading and that

these policies significantly affect the profitability of insider trading. Given that ITPs reduce firms’

legal risk, firms with better internal governance are more likely to adopt such policies ex-ante.6

Another mechanism is to limit managerial incentives to engage in insider trading through ex-post

disciplinary actions against insiders who engage in informed transactions. This ex-post

mechanism can restrict future informed insider trading by signaling a firm’s commitment to

discipline top executives engaged in transactions that exploit private information.7 We test the

role of internal governance in preventing informed insider trading by examining whether the

quality of a firm’s internal governance is positively associated with the likelihood of adopting

ITPs and policies that require GC’s pre-approval and whether the likelihood of forced CEO

turnover is high when CEOs are engaged in informed insider transactions.

Using a large sample of insider trading of officers and directors of firms listed on the

NYSE, AMEX, or NASDAQ from 1998 to 2011, we find that better internal governance

significantly reduces the six-month profitability of insider sales but not that of insider purchases.

6 In their recent paper, Lee et al. (2014) show that firms with ITPs tend to have more anti-takeover provisions as measured by a governance index constructed by Gompers et al. (2003; hereafter, GIM index), possibly because these provisions, by allowing mangers to avoid short-term market pressures, enable them to pursue corporate policies that increase long-term shareholder value. In a related study, using a relatively small sample of 260 firms with ITPs, Jagolinzer et al. (2011) show that internal corporate governance measures are not generally related to the likelihood of adopting a voluntary corporate policy that requires general council’s approval prior to insider trading. 7 Niehaus and Roth (1999) show that the likelihood of CEO turnover among the defendant firms under securities class actions is significantly positively associated with the extent of insider selling during the period when firms are accused of not releasing material information to the public.

5

This asymmetric effect of internal governance on the profitability of insider trading is consistent

with both the compensation-based explanation suggested by Roulstone (2003)8 and the legal

risk-based explanation by Cheng and Lo (2006) and Rogers (2008). The results are robust to

using a variety of internal corporate governance measures, regression model specifications, and

insider trading profitability measures, and controlling for endogeneity bias.

We also find that the impact of a firm’s internal governance on the profitability of insider

sales is more pronounced for firms with greater ex-ante litigation risk and for opportunistic

transactions that are more likely to attract the attention of the SEC (Cohen et al., 2012). These

results indicate that good internal governance plays an important role in reducing the profitability

of insider sales transactions that are likely to be motivated by negative private information and

thus increase legal risk.

Finally, we find that firms with better internal governance are more likely to adopt

voluntary restriction policies, such as ITPs and policies that require GC pre-approval. More

importantly, the negative relation between the quality of internal governance and the profitability

of insider sales holds even after controlling for these ex-ante voluntary restriction policies,

suggesting that in addition to these ex-ante preventive policies, internal governance uses other

mechanisms to discourage insiders from exploiting private information. Consistent with this

interpretation, we find that better-governed firms are more likely to take active ex-post

disciplinary action (i.e., forced CEO turnover) against CEOs who engage in informed insider

sales transactions. Thus, internal governance affects the profitability of insider trading by

8 Since allowing profitable insider purchases and prohibiting opportunistic insider sales are potential ways to reward and discipline managers for their success and failure, respectively, an optimal compensation/governance structure may restrict only insider sales, suggesting that good internal governance asymmetrically affects the profitability of insider trading.

6

influencing firms to adopt ex-ante preventive measures such as ITPs and policies that require GC

pre-approval, as well as to take ex-post disciplinary actions such as forced CEO turnover.

Our paper contributes to the literature in at least two important ways. First, it documents

the role of internal governance in preventing insiders from exploiting private information.

Previous studies show that corporate governance affects a firm’s idiosyncratic risk, stock

liquidity, executive compensation, and valuation (Ferreira and Laux, 2007; Aggarwal et al., 2009;

Chung et al., 2010; Armstrong et al. 2012). We add to these studies by showing that better

governance benefits shareholders by discouraging insiders from exploiting negative private

information. We also identify two channels, ex-ante preventive measures (ITPs and the policy

that requires GC pre-approval) and ex-post disciplinary actions (removing CEOs who engage in

highly informed insider sales), through which internal governance restricts insider trading.

Second, our paper contributes to the literature on insider trading by showing that

shareholders pay more attention to legal risk associated with types of insider transactions (i.e.,

sales and purchases) rather than average abnormal profits earned by insiders from those

transactions. We find that internal governance is effective in restricting informed insider sales

but not informed insider purchases mainly due to greater legal risk associated with insider sales,

even though insiders on average earn positive abnormal profits only from their purchases. The

restrictive effect of internal governance on the profitability of insider sales is particularly strong

when firms have high ex-ante litigation risk or when their officers and directors are engaged in

opportunistic transactions.

The remainder of the paper is organized as follows. In Section 2, we discuss the

construction of our key variables of interest (the measures of the quality of a firm’s internal

governance and the profitability of insider trading) and describe the data and sample

7

characteristics. Section 3 outlines our empirical methodology and presents the main empirical

results. In Section 4 we examine ex-ante and ex-post mechanisms through which internal

governance discourages insiders from taking advantage of negative private information. We

present the summary and concluding remarks in Section 5.

2. Variable construction and sample description

2.1. Measures of internal governance quality

To measure the quality of a firm’s internal governance, we focus on three important types

of governance mechanisms: board independence, compensation structure, and institutional

ownership. Previous studies emphasize the importance of board independence in internal

governance. We use the percentage of outside directors on the board and the percentage of

outside directors in the compensation committee as the measures of board independence.

Weisbach (1988), Rosenstein and Wyatt (1990), Byrd and Hickman (1992), and Hermalin and

Weisbach (1998) show that outside directors protect the interests of shareholders when their

interests diverge from those of managers, suggesting that the percentage of outside directors on

the board can serve as an important measure of board independence. Roulstone (2003) shows

that the executives of firms with restriction policies on insider trading tend to receive a premium

in their total compensation, suggesting that insider trading policies are closely related to

executive compensation. Given the importance of the compensation committee in determining

executive compensation structures, we use the percentage of outside directors in the

compensation committee as another measure of board independence.

Compensation structures can also play an instrumental role in internal governance.

Previous studies show that agency problems between managers and shareholders can be reduced

8

by optimally designing compensation structures (Jensen and Murphy, 1990) and that firms with

better governance tend to have a higher pay-performance sensitivity (Li and Srinivasan, 2011).

When executives and directors have compensation schemes that are tied to firm performance,

their interests are more likely to be aligned with those of shareholders, suggesting that

compensation structures play an important role in mitigating shareholder-manager conflicts. We

measure the effectiveness of compensation structures as an internal governance mechanism using

two variables (Hoechle et al., 2012): CEO pay-performance sensitivity and an indicator for firms

in which any non-executive directors receive stock or options as a part of their compensation.

Previous studies also show that large shareholders perform an important monitoring

function (Shleifer and Vishny, 1986). For example, Hartzell and Starks (2003) find that

institutional investors have strong incentives to closely monitor managers when they have a large

stake in the firms. We use two measures to capture the governance role of large shareholders: the

percentage of ownership held by institutional investors and the percentage of shares held by the

top five independent, long-term, and dedicated/quasi-indexer institutional investors as defined in

Chen et al. (2007).

To measure the overall quality of a firm’s internal governance, we aggregate the above six

measures of three main internal governance attributes into one composite measure using an

approach similar to the one used in Larcker et al. (2007) and Armstrong et al. (2012).

Specifically, we first standardize each variable to have zero mean and unit variance, and then

sum these standardized values of all individual governance variables to obtain our main

composite internal governance score, ICG.

We collect information on directors, executive compensation, and institutional ownership

from the RiskMetrics, S&P’s ExecComp, and Thomson Reuters’ institutional holdings (13F)

9

databases, respectively. RiskMetrics provides information on director ownership and

compensation committee membership starting from 1998, while ExecComp provides information

on executive compensation for about 2,800 firms that are or were members of the Standard and

Poor’s (S&P) 500 index, the S&P MidCap 400 index, or the S&P SmallCap 600 index. This data

availability allows us to measure the quality of internal governance for those S&P 1500 firms

covered by ExecComp from 1998.

To check the robustness of our results, we experiment with two alternative measures of the

quality of a firm’s internal governance. First, we use a more comprehensive list of governance

variables including board size, the percentages of old directors, an indicator for busy directors,

an indicator for at least one independent director who is a large shareholder, the fraction of

directors whose tenure predates the CEO, an indicator for CEO/Chairman duality, and the

ownership held by insiders as in Armstrong et al. (2012), Hazarika et al. (2012) and Hoechle et al.

(2012). We first standardize these variables and then add their standardized values to ICG to

obtain a more comprehensive governance measure, ICG1.9 Appendix 1 summarizes the detailed

description on the construction of each governance variable used in the analysis.

Second, we use the self-constructed governance score based on the Corporate Governance

Quotient (CGQ) attributes provided by Institutional Shareholder Services (ISS). This measure

has been used in several papers including Aggarwal et al. (2009), Bernile and Jarrell (2009),

Chung et al. (2010) and Chung and Zhang (2011), and it mostly captures the quality of a firm’s

internal governance. ISS covers a large set of U.S. firms included in the S&P 500 index, the S&P

9 When a higher value of the standardized variable means weak governance, we multiply it by -1 in order to interpret ICG1 in a consistent manner.

10

SmallCap 600 index, or the Russell 3000 index starting from 2001.10 Following these previous

studies, we define the ISS governance score as the number of minimally acceptable governance

attributes met by a firm out of 64 governance attributes along eight major dimensions, with

higher ISS scores indicating better governance.

2.2. Measure of insider trading profitability

We measure the profitability of insider trading by estimating abnormal returns over the

180 calendar days following the transaction date.11 We use two approaches to calculate abnormal

returns. First, similar to Ravina and Sapienza (2010), we use the six-month market-adjusted

abnormal return defined as the difference between a firm’s buy-and-hold return over the 180

calendar days following the transaction date and the corresponding buy-and-hold return for the

market (BHAR6m), where we use the CRSP value-weighted index as a proxy for the market

portfolio.

Second, as an alternative measure of insider trading abnormal returns, we follow

Jagolinzer et al. (2011) and use an intercept from the Carhart (1997) four-factor model estimated

over the 180 calendar days subsequent to the transaction date (Alpha6m). To save space, we

present most of our results based on the market-adjusted abnormal returns since the results based

10 ISS also covers the firms that are required to file various documents and forms with the SEC through the Electronic Data Gathering, Analysis, and Retrieval system (EDGAR). For a detailed description of CGQ for U.S. firms covered by ISS, see Aggarwal and Williamson (2006) and Aggarwal et al. (2009). 11 We focus on the window of the 180 days because of the “short swing” rule (Section 16(b) of the Securities Exchange Act of 1934) that prohibits insiders from earning profits in a round-trip transaction within a six-month interval, which is likely to force insiders not to reverse their position for at least six months. We also examine how short-term (three days around the filing date of insider trading) abnormal returns are affected by the effectiveness of internal governance measured and find that the market interprets the announcements of insider sales in better-governed firms as less informative news than those in poorer-governed firms. Finally, we use the log of total dollar profits earned from insider trading as an alternative measure of the informativeness of insider trading and find similar results. Since total dollar profits are confounded by size and individual wealth effects, we focus on the measures based on abnormal returns as in previous studies (Ravina and Sapienza, 2010; Jagolinzer et al., 2011).

11

on the four-factor model are qualitatively similar.12 Throughout the paper, we multiply post-

transaction abnormal returns of sales by -1 to indicate that the numbers reported in the paper are

abnormal profits earned by insiders.

2.3. Sample selection and summary statistics

Our initial sample includes insider transactions of firms listed on the NYSE, AMEX, or

NASDAQ covered in Thomson Financial Insiders Data Feed (IDF) from 1998 to 2011. The

Thomson Financial IDF contains trade information on directors, officers, and large stockholders

with holdings greater than 10% of a firm’s stock, all subject to disclosure requirements as

defined in Section 16 of the Securities Exchange Act of 1934. Since our main hypothesis about

the role of governance in preventing insiders from exploiting private information is relevant only

for officers and directors, we exclude transactions made by large shareholders from the sample

and use only transactions made by officers and directors in most of our analyses. We focus only

on valid open market or private purchase and sales transactions of common shares.13

To be included in our sample, we also require that firms be covered by RiskMetrics,

ExecuComp, and Thomson Reuters’ institutional holdings (13F) data, and that their stock return

and financial data be available in CRSP and Compustat, respectively. Following previous studies,

we further limit the sample by requiring that share codes in CRSP be 10 or 11, and we exclude

the following transactions from the sample: (1) transactions with less than 100 shares or those

with trading prices less than $2; (2) transactions with traded prices outside the range between the

daily low and high prices reported in CRSP; (3) transactions with the number of shares

12 In Panel C of Table 3, we use the monthly time-series portfolio approach to estimate average abnormal returns (alphas) based on the four-factor model as in Loughran and Ritter (1995) and Cohen et al. (2012). 13 A valid transaction is one without a cleanse code of “A” or “S” in Thomson Financial Insiders Data Feed database. Any shares acquired as a part of compensation are not included in the analyses, even though the sales of shares received as a part of compensation are included in the analyses, since we cannot separate open market or private sales of shares that are acquired as a part of compensation from other open market or private sales.

12

exceeding the total number of shares outstanding in CRSP; (4) transactions with the number of

shares traded exceeding the total daily trading volume in CRSP; and (5) regulated firms in the

financial or utilities industries (firms with SIC codes between 6000 and 6999 or between 4900

and 4999). These restrictions result in a final sample of 11,310 firm-year observations and

463,527 insider transactions.14 We obtain data on analyst forecasts from I/B/E/S and data on

litigation from Stanford Securities Class Action Clearinghouse. Because the data requirements

differ across tests, the sample size for each test varies depending on the availability of the data in

the analysis.

Table 1 reports summary statistics for the sample of 463,527 insider transactions. Detailed

descriptions of each variable are available in Appendix 2. Panel A reports the summary statistics

of variables used in our regression analyses and Panel B presents the mean and median

governance measures for two subperiods, before and after the 2002 enactment of the Sarbanes-

Oxley Act (SOX). The numbers reported in the “Mean” column can be interpreted as the number

of insider transactions-weighted average numbers across sample firms since the weight of firms

with a large number of insider transactions will be greater than the weight of firms with a smaller

number of insider transactions. The mean six-month buy-and-hold abnormal return (BHAR6m)

and daily abnormal return over the six-month period (Alpha6m) are -0.31% and -1.15 basis

points, respectively, indicating that on average, insiders do not earn positive abnormal returns

during our sample period. However, as shown in Table 3, these low and insignificant abnormal

returns are primarily due to negative abnormal profits earned by insiders following insider sales.

14 There are a total of 1,429,547 transactions made by insiders of NYSE, AMEX, and NASDAQ firms that are available in the Thomson Financial’s IDF database during our sample period. Due to stock returns (CRSP) and financial data (Compustat) requirements, 40,659 and 45,589 observations are excluded, respectively. In addition, 212,710 observations are excluded due to the unavailability of analyst information or due to the requirement of a minimum of 3 analysts following the firm in I/B/E/S. Finally, 667,062 transactions are excluded from the sample due to the fact that the data is unavailable in RiskMetrics, ExecComp, or Thomson Reuters’ insider holdings (13F) database. Our results are similar when we exclude transactions made on a date when both insider purchases and sales are made in the same company.

13

The mean market-adjusted abnormal return over the 180 calendar days prior to the insider

trading date (Prior6m) is -14.54%, suggesting that insiders tend to trade their shares following

poor past stock performance.

We also find that the mean (median) ICG, ICG1, and ISS scores are 0.00 (0.48), 0.00

(0.32), and 36.78 (38), and the mean (median) annual changes in ICG, ICG1, and ISS scores are

0.385 (0.261), 0.166 (0.109), and 1.790 (1.00), respectively, suggesting that the quality of firm’s

internal governance generally improved over time during our sample period. The mean and

median market values of equity (Size) are $18.51 billion and $3.26 billion in 1998 purchasing

power, and the mean and median market-to-book equity ratios (MB) are 4.25 and 2.99,

respectively.

The mean ratio of the absolute value of the net number of shares purchased (i.e., number

of shares purchased minus number of shares sold) by all insiders of a firm on the transaction date

to the total number of shares outstanding (TradeSize) and the mean ratio of the sum of absolute

values of the daily net numbers of shares purchased by all insiders of the same firm during the

ten days prior to the transaction date to the total number of shares outstanding (RecentTrades)

are 0.27% and 0.58%, respectively. We also find that 25.5% and 7.6% of our sample transactions

are made by insiders of firms reporting non-zero R&D expenditures (RND) and by insiders of

firms with negative net income before extraordinary items during the most recent fiscal year

(Loss), respectively. The mean (median) standard deviation of financial analysts’ earnings per

share (EPS) forecasts over the average forecasted EPS (Dispersion) is 10.0% (3.4%). Finally, the

mean fraction of transactions made by insiders of firms with ITPs (a policy that requires GC pre-

approval) is 48.1% (8.9%).15

15 Similar to Jagolinzer et al. (2011), we collect data on ITPs and the policy that requires GC pre-approval by manually searching firms’ websites in early 2012. We set the indicator variable to one only if we find evidence of

14

In Panel B, we report the mean and median governance measures for the two subperiods—

before (1998-2002) and after (2003-2011) the enactment of the SOX—to test whether there are

significant structural changes in internal governance following the SOX. As expected, we find

that the quality of firms’ internal governance has significantly improved following the SOX as

measured by ICG, ICG1, and ISS.

Table 2 reports pairwise Pearson and Spearman correlation coefficients of key variables

used in our analyses. Since both correlation measures are similar in most cases, we only discuss

the results of Pearson correlation coefficients that are reported above the diagonal. The

correlations between our two abnormal return measures, BHAR6m and Alpha6m, and between

two internal governance measures, ICG and ICG1, are 0.77 and 0.69, respectively, while the

correlation between ICG (ICG1) and log (ISS) is relatively low at 0.40 (0.08). More importantly,

we find that abnormal returns are generally negatively correlated with internal governance

measures, consistent with our hypothesis that insiders of firms with better internal governance

are discouraged to exploit private information. We also find that our governance measures are

significantly positively correlated with the indicator for firms with ITPs and the indicator for

firms with a policy that requires GC pre-approval.

2.4. Abnormal returns

In Table 3, we report average abnormal returns by transaction type (i.e., purchases and

sales) and the quality of internal governance. In each calendar year, we divide sample firms into

such policies on websites. Otherwise, we set the variable to zero. One limitation of this approach is that we do not know when a firm has adopted the policy; therefore, we set the indicator to one for the whole sample period once we find evidence of such policies on its website. To check the robustness of our results, we experiment with an alternative approach similar to the one used in Roulstone (2003) by defining firms with 75% or more of insider trades within one month after earnings announcements as those with insider trading restriction policies and find similar results.

15

three groups, high (top 30%), medium (middle 40%), and low (bottom 30%), based on ICG

scores. Mean and median abnormal returns measured by BHAR6m (%) and Alpha6m (basis point)

are reported in Panels A and B, respectively. In addition, we calculate the average monthly

abnormal returns of each group based on the Carhart’s (1997) four-factor model. We follow

Loughran and Ritter (1995) and Cohen et al. (2012) and use the monthly time-series portfolio

approach to calculate the abnormal returns, Alphamret (%).16 The results are presented in Panel

C.

The results in Panel A show that during our sample period, insiders earn significantly

positive abnormal returns over six months following their purchase transactions but not

following their sales transactions. For example, the average (median) six-month market-adjusted

abnormal return for purchases is a significant 3.7% (2.89%), while the average (median) for sales

is a significant -0.64% (0.37%). We further find that the passage of SOX does not significantly

affect abnormal returns earned by insiders. Before SOX, the average (median) BHAR6m for

purchases is 3.91% (2.48%) and the average (median) BHAR6m for sales is -2.12% (-0.39%),

while after SOX, the average (median) BHAR6m for purchases is 3.59% (3.12%) and the average

(median) BHAR6m for sales is -0.46% (0.45%).

More importantly, we find that internal governance is not closely associated with

abnormal returns earned from insider purchases but it is negatively and significantly associated

with abnormal returns earned from insider sales. For instance, the average abnormal return from

insider sales (purchases) in High ICG firms is 3.95% (0.71%) lower (higher) than that in Low

ICG firms.

16 Specifically, we assign a sample firm into a net purchase (net sale) group based on the net number of shares purchased by insiders of the firm during the month. Then, in each month, we form net purchase (sale) portfolios using all firms classified as net purchase (sale) firms at least once over the past six-month period and calculate value-weighted portfolio returns. For aggregate portfolios, we include all firms in net purchase and sale portfolios but multiply the returns of the firms in net sale portfolios by -1. For each portfolio, we run monthly time-series regressions based on the Carhart’s (1997) four-factor model to estimate Alphamret.

16

The results are similar in Panels B and C. Compared to insiders of poorer-governed firms,

those of better-governed firms earn significantly smaller abnormal profits from their sales

transactions but not from their purchase transactions, suggesting that internal governance limits

insiders’ ability to profit from their negative private information, thereby reducing firms’ legal

risk.

In the following analyses, we investigate whether these results hold even after we control

for various factors that affect the profitability of insider trading.

3. Empirical methodology and main results

3.1. Empirical methodology

To examine whether internal governance plays a role in limiting the profitability of insider

trading, we estimate the following regression:

BHAR6m (Alpha6m) = α + β1Gov + β2log(Size) + β3MB + β4Prior6m + β5TradeSize

+β6RecentTrades + β7RND + β8Loss + β9Dispersion + β10ITP

+ β11GC + β12 Fixed Effects + ε, (1)

where

BHAR6m (Alpha6m) = Market-adjusted buy-and-hold abnormal returns over the 180 calendar

days following the transaction date (intercept from the Carhart (1997) four-factor model

estimated over the 180 calendar days subsequent to the transaction date),

Gov = Various governance measures including board independence, compensation structure,

institutional ownership, ICG, ICG1, and log(ISS) measures,

Size = Inflation-adjusted market capitalization at the end of the most recent fiscal quarter (in

1998 dollars),

17

MB = Ratio of the market value of equity to the book value of equity at the end of the most

recent fiscal quarter,

Prior6m = Market-adjusted buy-and-hold abnormal returns over the 180 calendar days prior to

the insider trading date,

RND = Indicator that takes the value of one if the firm reports non-zero research and

development expenditures in the most recent fiscal year and zero otherwise,

TradeSize = Absolute value of the net number of shares purchased by all insiders of a firm on the

transaction date divided by the total number of shares outstanding of the firm,

RecentTrades = Sum of absolute values of the daily net numbers of shares purchased by all

insiders of the same firm during the ten days prior to the transaction date, scaled by the total

number of shares outstanding,

Loss = Indicator that takes the value of one if the firm reports negative net income before

extraordinary items for the most recent fiscal year and zero otherwise,

Dispersion = Standard deviation of earnings per share (EPS) forecasts for the current fiscal year

divided by the average of EPS forecasts made during the month of transaction date,

ITP = Indicator that takes the value of one if the firm has insider trading policies and zero

otherwise, and

GC = Indicator that takes the value of one if the firm has a general counsel pre-approval

requirement and zero otherwise.

Following Lakonishok and Lee (2001), we include Size and MB to control for size and

book-to-market effects (Fama and French, 1993). In addition, following Brochet (2010), we

control for Prior6m, RND, TradeSize, RecentTrades, and Loss. We include Prior6m to control

for insiders’ contrarian behavior. RND and Dispersion are included since insider sales and

18

purchases are likely to be more informative in firms with higher R&D intensity or those with

greater analysts forecast dispersion, in which information asymmetry problems are perceived to

be greater (Aboody and Lev, 2000; Healy and Palepu, 2001). We further include TradeSize to

control for the possible link between the importance of private information and trade size, and

RecentTrades to control for either preemptions of a trade’s information content or reinforcements

of prior signals. We also include Loss to control for the potential reversal of poor accounting

performance. Finally, we include ITP and GC in the regression to control for previously

documented effects of these insider trading policies on the profitability of insider trading (Bettis

et al., 2000; Jagolinzer et al., 2011).

Our key variable of interest is Gov. If the internal governance system reduces the

profitability of insider trading, we expect the coefficient estimate on Gov to be negative. We

mitigate the potential bias caused by omitted unobservable industry characteristics by including

industry fixed effects in the regressions using Fama-French (1997) 48 industry indicators. We

also include year fixed effects to control for potential time trend effects. To incorporate the

guidance suggested by Petersen (2009) about the use of panel data sets, we also use clustered

standard errors at the individual firm level to calculate t-statistics. As robustness tests, we

calculate t-statistics using clustered standard errors at the firm and year levels, at the individual

insider level or at the transaction date level and find qualitatively similar results.

3.2. Internal governance and profitability of insider trading

The results using BHAR6m as the dependent variable are reported in Table 4. Given the

differential effects of internal governance on insider purchases and sales (as indicated in Table 3),

we run the regressions separately for these two types of insider trades. We also run the

19

regressions separately for six alternative governance measures: board independence,

compensation structure, institutional ownership, ICG, ICG1, and log (ISS). Consistent with

univariate results in Table 3, we find that the coefficient estimates on all six governance

measures are insignificant for purchase transactions but significantly negative at least at the 5%

level for sales transactions, suggesting that better internal governance significantly reduces the

profitability of insider sales but not insider purchases. 17 Thus, better internal governance

discourages insiders from exploiting negative private information, but not from exploiting

positive private information.18, 19 We also find that the effect of internal governance on the

profitability of insider sales is economically large and significant. For example, when ICG is

used as the governance measure, its coefficient estimate is -0.585 for sales transactions,

suggesting that one standard deviation increase in ICG (3.061) is associated with 1.79% lower

abnormal returns earned by insiders from their sales transactions. Among control variables, we

find that the coefficient estimates on log (Size) are significantly negative only for purchase

transactions. In contrast, the coefficient estimates on MB and Loss are significantly positive for

sales transactions, while they are significantly negative for purchase transactions.

Turning to the effect of voluntary insider trading restriction policies on the profitability of

insider trading, we find that the coefficient estimates on ITP and GC are negative but

insignificant for purchase transactions, whereas they are significantly negative for sales

transactions, indicating that these policies are effective in preventing insiders from exploiting

17 Fidrmuc et al. (2006) find that the market reaction to insider trading filings is positively associated with institutional ownership in the U.K., opposite to our findings, which might be due to different institutional environment between the U.S. and the U.K. 18 We check whether our results are robust to controlling for external governance measured by GIM index by estimating the similar regression specifications as those used in Ravina and Sapienza (2010) and find quantitatively similar evidence. 19 In untabulated tests, we use a similar approach as that used by Piotroski and Roulstone (2005), and examine the role of internal governance in restricting a specific type of information. We find that internal governance discourages insiders from exploiting private information but not from making contrarian trades based on public information.

20

negative private information only. More importantly, our finding that the coefficient estimates on

internal governance variables are significant for insider sales even after controlling for these

insider trading policies suggests that internal governance has a disciplinary effect on insider sales

that is above and beyond the effects of ITP and GC on insider trading.20

3.3. Sensitivity analysis

To check the robustness of the above results, we perform several additional tests in Table

5. We first examine whether our results are sensitive to alternative model specifications. The

results are reported in Panel A. For brevity, we report only the coefficient estimates on ICG. We

find that our key results in Table 4 are robust to (1) using the alpha estimate from the Carhart’s

(1997) four-factor model (Alpha6m) as the dependent variable; (2) replacing Dispersion with the

alternative measures of transparency, the number of analysts following (log (Following)) and

earnings quality;21 (3) reestimating regressions using the subsamples of insider trades that took

place before and after the 2002 enactment of the SOX;22 (4) aggregating the trades made on the

same date by all insiders of the same firm;23 (5) including firm and insider fixed effects to

address potential endogeneity bias caused by omitted unobservable firm and insider

characteristics, respectively; and (6) using the changes in ICG, ICG1, and ISS governance scores

instead of their levels.

20 Abnormal performance of insider sales is likely to be underestimated if firms are dropped from the sample due to stock-price decreases following insider sales. This sample selection bias problem is not a major concern for our study since our sample includes both active and inactive S&P 1500 companies. 21 Earnings quality is measured as the absolute value of discretionary accruals estimated using the modified Jones model (Dechow et al., 1995). 22 We find that even though the profitability of insider sales decreases after the enactment of the SOX, the influence of a firm’s internal governance on the profitability of sales remains significant after the SOX. 23 This aggregate approach can address the potential concern about cross-sectional dependence that arises when different insiders of a firm simultaneously purchase (sell) stocks several times on the same trading date and these trades are counted as separate observations in the regression, which leads to biased estimation of t-statistics. If the number of shares purchased (sold) during the day in a given firm is greater than the number of shares sold (purchased) during the same day, we define the aggregate transaction as an insider purchase (sale). Our results are also quantitatively similar when insider trade transactions are aggregated into monthly level.

21

In Panel B, we test whether the role of governance in discouraging insiders from

exploiting private information is different between firms that adopt ITP and GC and those that do

not by including the interaction terms between ICG and ITP and between ICG and GC. We find

that for sales transactions, the coefficient estimates on ICG and its interactions with ITP and GC

are significantly negative at least at the 5% level. The significantly negative coefficient on the

interaction term suggests that the role of internal governance in discouraging insiders from

exploiting negative private information becomes stronger, when better governance is

accompanied by insider trading restriction policies. The significant negative coefficient on ICG

suggests that internal governance affects the profitability of insider sales not only through the ex-

ante preventive measures (e.g., ITP and GC) but also through mechanisms other than ex-ante

preventive measures.

In Panel C, we perform a falsification test by examining whether internal governance

affects the profitability of insider trading made by large shareholders. Given that internal

governance is not designed to control large shareholders’ transactions, we expect that ICG does

not affect the profitability of insider trading made by them. Specifically, using all transactions

made by officers, directors, and large shareholders, we reestimate the previous regressions by

adding the following five variables as the independent variables: ICG × Officer, ICG × Director,

ICG × Large shareholder, Officer, and Director, where Officer, Director, and Large shareholder

are indicators for transactions made by officers, directors, and large shareholders, respectively.

We use the same control variables as those used in the Table 4 regressions. As expected, we find

that ICG matters only for sales transactions made by officers and directors but not for those made

by large shareholders.

22

Finally, in Panel D, we address the concern that ICG, ITP, and GC are endogenously and

simultaneously determined, which could result in biased and inconsistent coefficient estimates on

ICG. Specifically, we run three-stage least squares (3SLS) regressions using industry-median

ICG, industry-median ITP, and industry-median GC as the instrumental variables (IV) for ICG,

ITP, and GC, respectively. To the extent that firms in the same industry follow similar industry

practices in establishing internal governance systems, insider trading policies, and general

counsel pre-approval requirement policies, we expect these industry IV variables to be positively

related to ICG, ITP, and GC, respectively, thus satisfying the relevance requirement of an IV.

However, these variables are not likely to have any direct effects on the profitability of an

individual firm’s insider trading, other than affecting the profitability of insider trading only

through their correlation with the endogenous variables (i.e., ICG, ITP, and GC), satisfying the

exclusion restriction of an IV.

The first three rows report the coefficient estimates on IVs in the first-stage regressions

and the fourth row reports the coefficient estimate on ICG in the main regressions, in which

BHAR6m is the dependent variable and ICG, ITP, and GC are instrumented variables. We find

that all three instruments are significantly and positively related to endogenous variables for both

purchase and sales transactions, suggesting that our IVs satisfy the relevance requirement. The

results in the main regressions show that after controlling for the endogeneity of ICG, ITP, and

GC, the coefficient estimate on ICG is still significantly negative in sales transactions. Thus, our

finding regarding the effectiveness of internal governance in preventing insiders from exploiting

only negative private information appears to be robust when controlling for endogeneity

concerns.

23

Overall, the results in Table 5 suggest that our key results in the previous section are

robust to using alternative measures of trading profitability, various model specifications, and

subsample periods, and controlling for endogeneity concerns.

3.4. Internal governance and firm’s legal risk

In this subsection we examine whether the role of internal governance in limiting the

profitability of insider trading is particularly strong for firms and transactions with potentially

high litigation risk. As discussed earlier, if shareholders are concerned about legal risk arising

from insider trading, they should have strong incentives to reduce such risk by designing and

implementing good internal governance systems.

We focus on two cases in which legal risk is likely to be high. In these cases, shareholders

have strong incentives to limit insiders’ ability to profit from their transactions in order to reduce

such risk—that is, a case in which firms face high ex-ante litigation risk and a case in which

insiders engage in informative opportunistic trades. To the extent that firms with higher ex-ante

litigation risk and firms whose insiders engage in opportunistic trades face high legal risk that

arises from informed insider transactions, we expect the impact of a firm’s governance on the

informativeness of insider trading to be more pronounced for these firms and transactions. We

measure a firm’s ex-ante litigation risk using an indicator that takes the value of one if the

estimated probability of a firm’s litigation risk is above the 90th percentile and zero otherwise.24

24 We estimate the probability of a firm’s ex-ante litigation risk by using a probit model as in Francis et al. (1994), Johnson et al. (2001) and Rogers and Stoken (2005), for a sample of 457,999 firm-quarter observations over the 1998 to 2011 period. Specifically, we run a regression in which the dependent variable is an indicator that takes the value of one if a Rule 10b-5 lawsuit is filed against a firm in a given quarter and zero otherwise. The independent variables include firm size, beta, daily turnover, cumulative quarterly return, standard deviation of daily returns, minimum of daily returns, and skewness of daily returns and indicators for high-risk industries.

24

We identify opportunistic trades using an indicator that takes the value of one if the transactions

are classified as opportunistic trades and zero otherwise.25

We then add these two indicators and their interaction terms with ICG to equation (1), and

reestimate the models in Table 4. The results are reported in Table 6. We find that for insider

sales, the coefficient estimates on the interaction terms between ICG and two indicators for high

litigation risk are negative and significant, while they are insignificant for insider purchases.

These results suggest that internal governance plays a particularly important role in limiting the

profitability of sales transactions made by insiders of firms that are exposed to high legal risk and

opportunistic insider sales that are likely to attract greater legal problems.

4. Channels through which internal governance affects profitability of insider trading

In this section, we investigate the potential channels through which internal corporate

governance affects the profitability of insider trading.

4.1. Ex-ante preventive measures

One way through which firms with better internal governance can prevent insiders from

exploiting private information is to use ex-ante preventive measures such as adopting ITPs or a

policy that requires GC pre-approval. To examine whether firms with better internal governance

are indeed more likely to use these ex-ante preventive measures, we estimate a probit model in

which the dependent variable is either ITP or GC. Following Roulstone (2003), we control for

firm size (log(Size)), stock return volatility (log(Volatility)), insider ownership (InsOwn), insider

25 Similar to Cohen et al. (2012), in each calendar year, we first define routine traders as those insiders who trade stocks consistently in the same calendar month for at least three years in a row in the past. All others are defined as opportunistic traders. All trades made by routine (opportunistic) traders are defined as routine (opportunistic) transactions.

25

trading volume (Ins_Trade), analyst following (log(Following)), institutional ownership, and

litigation probability estimated using the Rogers and Stocken (2005) approach. In addition, we

control for industry and year fixed effects.

The results are reported in Table 7. The sample consists of 10,339 firm-year observations

between 1998 and 2011. 26 We find that the coefficient estimates on ICG are positive and

significant in both regressions that use ITP and GC as the dependent variables, respectively,

suggesting that firms with better internal governance are more likely to adopt ex-ante preventive

measures to prevent insiders from engaging in insider trading. These results, together with those

reported in Panel B of Table 5, suggest that ITPs and policies that require GC pre-approval are

more effective in reducing abnormal returns earned from sales transactions made by insiders of

firms with better internal governance, suggesting that better-governed firms use ex-ante

measures to prevent insiders from exploiting negative private information. 27 The coefficient

estimates on control variables are generally consistent with those reported in Roulstone (2003).

4.2. Ex-post disciplinary actions

In addition to adopting ex-ante preventive measures of insider trading, better-governed

firms may use ex-post disciplinary measures to inflict a sanction on top executives who engaged

in informed insider trading, thereby committing to discourage future informed insider trading by

other top executives. Of several potential ex-post disciplinary actions, we focus on forced CEO

turnover since it is considered to be one of the most aggressive corporate governance actions

(e.g., Kaplan and Minton, 2012).

26 In Tables 7 and 8, we conduct the analyses by aggregating transaction-level data used in previous sections at the firm level. 27 Our results are quantitatively similar when we define the insider trading policy following the approach used by Roulstone (2003).

26

We identify forced CEO turnover following the procedures used by Parrino (1997),

Bushman et al. (2010) and Li and Srinivasan (2011).28 We then estimate a probit regression in

which the dependent variable is an indicator that takes the value of one if forced CEO turnover

takes place during the fiscal year following the year when CEOs trade the shares of their firms

and zero otherwise. As key variables of interest, we include Avg_BHAR6m (value-weighted

average market-adjusted abnormal return over 180 calendar days subsequent to a CEO

transaction made during a fiscal year), ICG, and the interaction term between these two variables.

The coefficient estimate on the interaction term measures whether the likelihood of forced CEO

turnover is higher for better-governed firms whose insiders earn higher market-adjusted returns

from their transactions. Following Hazarika et al. (2012), we control for several factors that

influence the likelihood of CEO turnover including industry-adjusted return (Ret), return on

assets (ROA), industry median annual stock return (RetInd), annual sales growth rate (SG),

volatility of daily stock returns of the firm during a fiscal year (RetVol), market capitalization

(log(Size)), and leverage (Lev). We also control for industry and year fixed effects.

The results are presented in Table 8. The sample consists of 9,954 firm-year observations

where at least one insider transaction is made by CEOs between 1992 and 2010.29 In columns (1)

and (2), we examine whether the likelihood of forced CEO turnover is associated with the

average abnormal returns earned from insider transactions made by CEOs during the previous

year. The results show that the likelihood of forced CEO turnover is significantly higher when

28 Specifically, in each fiscal year, we identify CEO turnover by comparing the names of CEOs in current and following fiscal years using the ExecuComp database. We then search the Factiva news database to determine whether the turnover is routine or forced. Turnover is classified as forced if the articles report that the CEO is fired, demoted, or resigns under questionable circumstances (e.g., policy differences, pressure, lawsuits, or suspected earnings management). Among routine turnover events, we further classify them as forced turnover if the CEO retires at age below 60 or if the news article does not report the reason being death, poor health, or the acceptance of another position. We exclude CEO turnover due to death, interim appointments, mergers, or spinoffs from the sample. 29 In untabulated tests, we include firm-year observations with no CEO transactions in the analysis by setting the values of Avg_BHAR6m to be zero. We find that the results are qualitatively similar.

27

CEOs realize higher abnormal returns from their sales transactions (column 2). However, we do

not find such a significant relation for purchase transactions (column 1).

In columns (3) and (4), we use a subsample of firm-year observations from 1998 to 2010,

during which ICG data are available. We find that the coefficient estimate on the interaction term

between ICG and Avg_BHAR6m is significantly positive in column (4), suggesting that the

likelihood of forced CEO turnover is higher when the CEOs of better-governed firms earn higher

abnormal returns from their sales transactions. However, the coefficient estimate on the

interaction term between ICG and Avg_BHAR6m from purchase transactions is insignificant

(column 3). These results suggest that firms with better internal governance indeed take more

disciplinary actions against CEOs’ informed transactions that involve greater legal risk (i.e.,

sales transactions).30

5. Summary and conclusion

Previous studies show that insiders exploit their information advantage to extract private

benefits and that well-designed internal governance is effective for aligning the interests of

shareholders with those of top managers. To the extent that an increase in litigation risk caused

by informed insider transactions harms the interests of shareholders, internal governance is

expected to play an instrumental role in discouraging insiders from engaging in such transactions.

Since legal risk tends to be higher in informed insider sales than in insider purchases (Cheng and

Lo, 2006; Rogers, 2008), and abnormal profits earned from insider sales cannot be considered an

30 In untabulated tests, to address the potential endogeneity concerns in Avg_BHAR6m and ICG, we use a 3SLS regression approach, in which we use industry median ITP and GC as IVs for Avg_BHAR6m, and industry median ICG as an IV for ICG. To the extent that industry median ITP and GC are significantly related to Avg_BHAR6m but unlikely to directly affect firms’ forced CEO turnover, these two variables are likely to meet both the relevance and exclusion requirements of instrumental variables. The results do not change.

28

optimal way to compensate managerial success, we hypothesize that internal governance is

particularly effective in reducing the profitability of insider sales.

Consistent with our hypothesis, we find that the average abnormal returns earned over the

180 calendar days following the sales transaction date are significantly lower for insiders of

better-governed firms than for insiders of poorer-governed firms. We do not find such results for

purchase transactions. The effect of internal governance on the profitability of sales transactions

is economically large: one standard deviation increase in the governance score, ICG, lowers six-

month abnormal returns earned by insiders from their sales transactions by 1.79%. The results

continue to hold even after controlling for firms’ insider trading policies and the policy that

requires general counsel pre-approval and are robust to using a variety of model specifications,

alternative methods to estimate abnormal returns, and alternative measures of the quality of

internal governance.

We further find that the restriction effects of internal governance on the profitability of

insider trading are particularly strong for insider sales that involve higher litigation risk, such as

sales made by insiders of firms with higher ex-ante litigation risk and opportunistic insider sales.

We also find that better-governed firms are more likely to place ex-ante preventive measures of

informed insider trading, such as adopting voluntary insider trading restriction policies, and to

take active ex-post disciplinary actions, such as forced CEO turnover against CEOs involved in

highly informed insider trading.

Overall, our results show that firms with good internal governance use both ex-ante and

ex-post mechanisms to prevent insiders from engaging in informed insider transactions that are

likely to increase legal risk, and that these internal governance mechanisms are effective in

discouraging insiders from exploiting negative private information. Our findings have important

29

implications for investors, regulators, and corporate managers, and suggest that future insider

trading research should consider the channels through which internal governance mechanisms

affect informed insider trading and the role of internal corporate governance in reducing legal

risk in the analysis.

30

References

Aboody, D., Lev, B., 2000. Information asymmetry, R&D and insider gains. Journal of Finance 55, 2747–2766.

Acharya, V., Myers, S., Rajan, R., 2011. The internal governance of firms. Journal of Finance 66,

689–720. Aggarwal, R., Erel, I., Stulz, R., Williamson, R., 2009. Differences in governance practices

between U.S. and foreign firms: Measurement, causes, and consequences. Review of Financial Studies 22, 3131–3169.

Aggarwal, R., Williamson, R., 2006. Did new regulations target the relevant corporate

governance attributes? Working Paper, Georgetown University. Armstrong, C., Ittner, C., Larcker, D., 2012. Economic characteristics, corporate governance,

and the influence of compensation consultants on executive pay levels. Review of Accounting Studies 17, 322–351.

Beasley, M.S., 1996. An empirical analysis of the relation between the board of director

composition and financial statement fraud. The Accounting Review 71, 443–465. Bernile, G., Jarrell, G., 2009. The impact of the options backdating scandal on shareholders.

Journal of Accounting and Economics 47, 2–26. Bettis, C., Coles, J., Lemmon, M., 2000. Corporate policies restricting trades by insiders. Journal

of Financial Economics 57, 191–220. Bhattacharya, U., 2013. Insider trading controversies: A literature review. Working paper,

Indiana University. Billings, M.B., Cedergren, M.C., 2015. Strategic silence, insider selling and litigation risk.

Journal of Accounting and Economics, Forthcoming. Brochet, F., 2010. Information content of insider trades before and after the Sarbanes-Oxley Act.

The Accounting Review 85, 419–446. Bushman, R., Dai, Z., Wang, X., 2010. Risk and CEO turnover. Journal of Financial Economics

96, 381–398. Byrd, J., Hickman, K., 1992. Do outside directors monitor managers? Evidence from tender offer

bids. Journal of Financial Economics 32, 195–221. Carhart, M., 1997. On persistency in mutual fund performance. Journal of Finance 52, 57–82.

31

Chhaochhaia, V., Kumar, A., Niessen-Ruenzi, A., 2012. Local investors and corporate governance. Journal of Accounting and Economics 54, 42–67.

Chen, X., Harford, J., Li, K., 2007. Monitoring: Which institutions matter? Journal of Financial

Economics 86, 279–305. Chen, C., Lu, H., Sougiannis, T., 2012. Agency problem, corporate governance, and the

asymmetrical behavior of selling, general, and administrative costs. Contemporary Accounting Research 29, 252–282.

Cheng, Q., Lo, K., 2006. Insider trading and voluntary disclosures. Journal of Accounting

Research 44, 815–848. Cheng, Q., Warfield, T.D., 2005. Equity incentives and earnings management. The Accounting

Review 80, 441–476. Chung, K., Elder, J., Kim, J., 2010. Corporate governance and liquidity. Journal of Financial and

Quantitative Analysis 45, 265–291. Chung, K., Zhang, H., 2011. Corporate governance and institutional ownership. Journal of

Financial and Quantitative Analysis 46, 247–273. Cohen, L., Malloy, C., Pomorski, L., 2012. Decoding inside information. Journal of Finance 67,

1009–1043. Cziraki, P., De Goeij, P., Renneboog, L., 2014. Corporate governance rules and insider trading

profits. Review of Finance 18, 67–108. Daily, C., Dalton, D., Cannella, A., 2003. Corporate governance decades of dialogue and data.

Academy of Management Review 28, 371–382. Dechow, P.M., Sloan, R.G., Sweeney, A.P., 1995. Detecting earnings management. The

Accounting Review 70, 193–225. Dechow, P.M., Sloan, R.G., Sweeney, A.P., 1996. Causes and consequences of earnings

manipulation: An analysis of firms subject to enforcement actions by the SEC. Contemporary Accounting Research 13, 1–36.

Fama, E., French, K., 1993. Common risk factors in the returns on stocks and bonds. Journal of

Financial Economics 33, 3–56. Fama, E., French, K., 1997. Industry costs of equity. Journal of Financial Economics 43, 153–

193. Fishman, M., Hagerty, K., 1992. Insider trading and the efficiency of stock prices. The RAND

Journal of Economics 23, 106–122.

32

Ferreira, M., Laux, P., 2007. Corporate governance, idiosyncratic risk, and information flow.

Journal of Finance 62, 951–990. Fidrmuc, J., Goergen, M., Renneboog, L., 2006. Insider trading, news releases and ownership

concentration. Journal of Finance 61, 2931–2973. Francis, J., Philbrick, D., Schipper, K., 1994. Shareholder litigation and corporate disclosures.

Journal of Accounting Research 32, 137–164. Gompers, P., Ishi, J., Metrick, A., 2003. Corporate governance and equity prices. Quarterly

Journal of Economics 118, 107–156. Hartzell, J., Startks, L., 2003. Institutional investors and executive compensation. Journal of

Finance 58, 2351–2374. Hazarika, S., Karpoff, J., Nahata, R., 2012. Internal corporate governance, CEO turnover, and

earnings management. Journal of Financial Economics 104, 44–69. Healy, P., Palepu, K., 2001. Information asymmetry, corporate disclosure, and the capital

markets: A review of the empirical disclosure literature. Journal of Accounting and Economics 31, 405–440.

Helland, E., Sykuta, M., 2005. Who’s monitoring the monitor? Do outside directors protect

shareholders' interests? Financial Review 40, 155–172. Henderson, M., Jagolinzer, A., Muller, K., 2012. Hiding in plain sight: Can disclosure enhance

insiders’ trade returns? Working Paper, University of Chicago. Hermalin, B., Weisbach, M., 1998. Endogenously chosen boards of directors and their

monitoring of the CEO. American Economic Review 88, 96–118. Hoechle, D., Schmid, M., Walter, L., Yermack, D., 2012. How much of the diversification

discount can be explained by poor corporate governance. Journal of Financial Economics 103, 41–60.

Hoitash, U., Hoitash, R., Bedard, J., 2009. Corporate governance and internal control over

financial reporting: A comparison of regulatory regimes. The Accounting Review 84, 839–867.

Jagolinzer, A., 2009. SEC Rule 10b5-1 and insiders strategic trade. Management Science 55,

224–239. Jagolinzer, A., Larcker, D., Taylor, D., 2011. Corporate governance and the information content

of insider trades. Journal of Accounting Research 49, 1249–1274.

33

Jeng, L.A., Metrick, A., Zeckhauser, R., 2003. Estimating the returns to insider trading: A performance-evaluation perspective. Review of Economics and Statistics 85, 453–471.

Jensen, M., Murphy, K., 1990. Performance pay and top management incentives. Journal of

Political Economy 98, 225–264. Johnson, M.F., Kasznik, R., Nelson, K., 2001. The impact of securities litigation reform on the

disclosure of forward-looking information by high technology firms. Journal of Accounting Research 39, 297–327.

Johnson, M.F., Nelson, K.K., Pritchard, A.C., 2007. Do the merits matter more? The impact of

the Private Securities Litigation Reform Act. Journal of Law, Economics & Organization 23, 627–652.

Kaplan, S., Minton, B., 2012. How has CEO turnover changed? International Review of Finance

12, 57–87. Lakonishok, J., Lee, I., 2001. Are insider trades informative? Review of Financial Studies 14,

79–111. Laksmana, I., 2008. Corporate board governance and voluntary disclosure of executive

compensation practices. Contemporary Accounting Research 25, 1147–1186. Larcker, D., Richardson, S., Tuna, I., 2007. Corporate governance, accounting outcomes, and

organizational performance. The Accounting Review 82, 863–1008. Loughran, T., Ritter, J., 1995. The new issues puzzle. Journal of Finance 50, 23–51. Lee, I., Lemmon, M., Li, Y., Sequeira, J., 2014. Do voluntary corporate restrictions on insider

trading eliminate informed insider trading? Journal of Corporate Finance 29, 158–178. Li, F., Srinivasan, S., 2011. Corporate governance when founders are directors. Journal of

Financial Economics 102, 454–469. Manne, H., 1966. Insider trading and the stock market. New York: Free Press. Masulis, R., Wang, C., Xie, F., 2012. Globalizing the boardroom - The effects of foreign

directors on corporate governance and firm performance. Journal of Accounting and Economics 53, 527–554.

Morellec, E., Nikolov, B., Schurhoff, N., 2012. Corporate governance and capital structure

dynamics. Journal of Finance 67, 803–848. Muller, K.A., Neamtiu, M., Riedl, E.J., 2012. Do managers benefit from delayed goodwill

impairments? Working Paper, Pennsylvania State University.

34

Niehaus, G., Roth, G., 1999. Insider trading, equity issues, and CEO turnover in firms subject to securities class action. Financial Management 28, 52–72.

Parrino, R., 1997. CEO turnover and outside succession: A cross-sectional analysis. Journal of

Financial Economics 46, 165–197. Petersen, M., 2009. Estimating standard errors in finance panel data sets: Comparing approaches.

Review of Financial Studies 22, 435–480. Piotroski, J., Roulstone, D., 2005. Do insider trades reflect both contrarian beliefs and superior

knowledge about future cash flow realizations? Journal of Accounting and Economics 39, 55–81.

Ravina, E., Sapienza, P., 2010. What do independent directors know? Evidence from their

trading. Review of Financial Studies 23, 962–1003. Rogers, J.L., 2008. Disclosure quality and management trading incentives. Journal of

Accounting Research 46, 1265–1296. Rogers, J.L., Stocken, P., 2005. Credibility of management forecasts. The Accounting Review 80,

1233–1260. Rosenstein, S., Wyatt, J., 1990. Outside directors, board independence, and shareholder wealth.

Journal of Financial Economics 26, 175–191. Roulstone, D., 2003. The relation between insider-trading restrictions and executive

compensation. Journal of Accounting Research 41, 525–551. Seyhun, N., 1986. Insiders’ profits, costs of trading, and market efficiency. Journal of Financial

Economics 16, 189–212. Shleifer, A., Vishny, R., 1986. Large shareholders and corporate control. Journal of Political

Economy 94, 461–488. Walsh, J.P., Seward, J.K., 1990. On the efficiency of internal and external corporate control

mechanisms. Academy of Management Review 15, 421–458. Weisbach, M., 1988. Outside directors and CEO turnover. Journal of Financial Economics 20,

431–460. Wintoki, M., Linck, J., Netter, J., 2012. Endogeneity and the dynamics of internal corporate

governance. Journal of Financial Economics 105, 581–606.

35

Table 1 Summary statistics This table reports descriptive statistics for a sample of 463,527 insider transactions made by officers and directors between 1998 and 2011. Variables are defined in Appendix 2. Panel A shows summary statistics for the variables used in the regressions for the full sample period. Panel B compares internal governance measures before and after the 2002 enactment of the Sarbanes-Oxley Act (SOX). *** indicates that the mean (median) differences between two subperiods are significantly different from zero at the 1% significance level.

Panel A: Main regression variables

Mean Q1 Median Q3 Standard deviation

BHAR6m (%) -0.314 -12.563 0.420 12.851 23.624 Alpha6m (basis points) -1.152 -11.388 -1.271 8.802 18.929 ICG 0.000 -1.672 0.482 2.116 3.061 ∆ICG 0.385 -0.749 0.261 1.381 1.987 ICG1 0.000 -2.443 0.323 2.612 3.880 ∆ICG1 0.166 -1.281 0.109 1.586 2.569 ISS 36.780 33 38 41 5.749 ∆ISS 1.790 0 1 3 2.960 Size ($ millions) 18,508 1,220 3,262 11,170 48,706 MB 4.249 1.946 2.994 4.737 4.599 Prior6m (%) -14.537 -26.159 -9.990 2.800 30.088 TradeSize (%) 0.271 0.031 0.101 0.290 0.502 RecentTrades (%) 0.575 0.000 0.002 0.361 1.771 RND (indicator) 0.255 0.000 0.000 1.000 0.436 Loss (indicator) 0.076 0.000 0.000 0.000 0.265 Dispersion 0.100 0.017 0.034 0.072 0.351 ITP (indicator) 0.481 0.000 0.000 1.000 0.500 GC (indicator) 0.089 0.000 0.000 0.000 0.285 Panel B: Internal governance measures before and after SOX

Before SOX (1998-2002) After SOX (2003-2011) Test of difference

Mean Median Mean Median Mean (Median)

ICG -3.055 -2.721 0.583 0.889 -3.638*** (-3.610) *** ICG1 0.140 0.191 0.425 0.279 -0.285*** (-0.088) *** ISS -1.885 -1.745 0.360 0.645 -2.245*** (-2.390) ***

36

Table 2 Pearson and Spearman correlation coefficients among selected variables This table provides correlation matrix for a sample of 463,527 insider transactions made by officers and directors between 1998 and 2011. All variables are defined in Appendix 2. Pearson and Spearman correlation coefficients are presented in upper diagonal and lower diagonal, respectively. All coefficients are significant at the 5% level except for those in italic. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) BHAR6m (1) 0.769 -0.026 -0.011 -0.022 -0.017 0.038 0.037 0.001 -0.026 -0.008 0.068 0.036 -0.021 -0.018 Alpha6m (2) 0.795 -0.022 -0.029 -0.020 -0.047 -0.020 0.048 0.006 -0.018 -0.029 0.061 0.012 -0.004 -0.021 ICG (3) -0.028 -0.038 0.694 0.403 -0.233 -0.047 0.030 0.017 0.011 -0.026 -0.002 0.045 0.014 0.018 ICG1 (4) -0.004 -0.039 0.665 0.079 -0.347 0.019 -0.033 0.134 0.114 0.020 0.037 0.066 0.072 0.001 log (ISS) (5) -0.019 -0.008 0.363 0.073 0.118 -0.039 0.091 -0.092 -0.069 -0.032 -0.074 -0.066 0.008 0.023 log (Size) (6) -0.027 -0.039 -0.220 -0.352 0.155 0.324 0.066 -0.234 -0.186 0.119 -0.113 -0.116 -0.021 0.030 MB (7) -0.007 -0.065 -0.016 0.037 -0.050 0.398 -0.045 -0.011 -0.008 0.130 -0.031 -0.068 0.063 -0.010 Prior6m (8) 0.015 0.020 -0.002 -0.034 0.085 0.070 -0.020 -0.108 -0.012 -0.037 -0.012 0.001 -0.066 -0.028 TradeSize (9) 0.019 -0.007 0.093 0.226 -0.044 -0.363 0.001 -0.135 0.300 -0.017 0.006 0.021 0.038 -0.015 RecentTrades (10) -0.013 -0.008 0.121 0.145 0.044 -0.120 0.040 -0.041 0.169 0.001 -0.012 0.008 -0.021 -0.042 RND (11) -0.007 -0.033 -0.024 0.028 -0.029 0.102 0.189 -0.019 -0.014 0.015 0.050 -0.022 0.150 0.025 Loss (12) 0.072 0.055 0.022 0.041 -0.069 -0.117 -0.162 0.006 -0.002 -0.035 0.050 0.305 0.059 0.007 Dispersion (13) 0.027 -0.007 0.088 0.131 -0.076 -0.163 -0.148 -0.129 0.036 0.016 0.019 0.282 0.010 -0.017 ITP (14) -0.003 -0.003 0.021 0.058 0.001 -0.024 0.031 -0.074 0.058 0.005 0.150 0.059 0.070 0.325 GC (15) -0.025 -0.029 0.017 0.015 0.032 0.045 0.004 -0.038 -0.029 -0.039 0.025 0.007 -0.005 0.325

37

Table 3 Abnormal returns earned by insiders according to the quality of firms’ internal governance and type of insider trading This table reports average abnormal returns earned by insiders according to the quality of firms’ internal governance and type of insider trading for a sample of 463,527 insider transactions made by officers and directors between 1998 and 2011. We divide firms into three governance groups according to their ICG, which is the composite governance measure of board independence, CEO compensation, and institutional ownership concentration at a common scale with a zero mean and unit variance. A higher ICG indicates better internal governance. In each year, firms with top 30% and bottom 30% of ICG are classified as High and Low ICG firms, respectively. For sale portfolios, we multiply the abnormal returns earned by insiders by -1 to indicate that the numbers reported in the table are abnormal profits earned by insiders. High – Low is a zero-investment portfolio that is long in High ICG firms and short in Low ICG firms. In Panel A, abnormal returns are measured by six-month market-adjusted abnormal returns (BHAR6m). In Panel B, daily abnormal returns are measured by an intercept from the Carhart (1997) four-factor model estimated over the 180 calendar days subsequent to the transaction date (Alpha6m). In Panels A and B, the averages are reported on top and the medians are reported in parentheses. In Panel C, average monthly abnormal returns are measured by intercepts from the Carhart (1997) four-factor model using the monthly time-series portfolio approach (Alpha6mret). To form portfolios, in each month between July 1997 and December 2011, we assign a sample firm into a net purchase (net sales) group based on the net number of shares purchased by insiders of the firm during the month, and then, in each month starting from January 1998, we form net purchase (sale) portfolios using all firms belonging to the net purchase (sales) group at least once over the past six-month period. In a similar way, we form net purchase (sale) portfolios that are composed of only those firms in each governance group. For each portfolio, monthly value-weighted portfolio returns during the month of portfolio formation are calculated and used in the regressions. In Panel C, alphas (p-values) from each portfolio return regression are reported on top (below in parentheses). P-values are based on heteroskedasticity-adjusted standard errors. All values in parentheses are significant except for those in italics.

1998-2011 1998-2002 2003-2011 Purchases Sales Aggregate Purchases Sales Aggregate Purchases Sales Aggregate

Panel A: BHAR6m (%; six-month market-adjusted abnormal return) Sample size 31,840 431,687 463,527 9,235 64,867 74,102 22,605 366,820 389,425

Full sample 3.702 -0.641 -0.319 3.907 -2.118 -0.450 3.589 -0.455 -0.294 (2.893) (0.372) (0.425) (2.482) (-0.389) (0.725) (3.121) (0.452) (0.482)

High ICG 3.911 -3.248 -2.755 3.761 -6.087 -3.040 3.994 -2.905 -2.702 (3.171) (-0.821) (-0.694) (2.796) (-3.433) (-1.783) (3.380) (-0.641) (-0.608)

Medium ICG 3.922 0.423 0.637 4.409 -1.342 0.209 3.652 0.423 0.722 (2.748) (1.014) (1.125) (2.099) (0.434) (1.453) (3.109) (1.014) (1.059)

Low ICG 3.200 0.705 0.923 3.382 0.878 1.263 3.099 0.684 0.860 (2.807) (0.915) (1.028) (2.679) (1.631) (2.041) (2.878) (0.857) (0.864)

High – Low 0.711 -3.953 -3.678 0.379 -6.965 -4.302 0.895 -3.589 -3.562 (0.364) (-1.736) (-1.722) (0.117) (-5.064) (-3.824) (0.502) (-1.498) (0.382)

38

Panel B: Alpha6m (basis points; daily abnormal return based on daily individual transaction level regression) Sample size 31,840 431,687 463,527 9,235 64,867 74,102 22,605 366,820 389,425 Full sample 6.511 -1.557 -1.152 6.637 -5.176 -2.746 6.642 -1.102 -0.848

(6.081) (-1.635) (-1.260) (6.236) (-4.655) (-2.455) (5.996) (-1.295) (-1.115) High ICG 6.659 -3.851 -3.246 6.373 -7.087 -4.197 6.818 -3.460 -3.070

(6.001) (-3.636) (-3.101) (5.974) (-5.360) (-3.277) (6.016) (-3.391) (-3.089) Medium ICG 6.494 -0.870 -0.635 6.897 -5.491 -2.943 6.270 0.423 0.677

(6.013) (-1.268) (-0.945) (6.307) (-5.349) (-2.741) (5.850) (1.014) (1.679) Low ICG 6.385 -0.047 0.304 6.543 -2.801 -1.031 6.297 0.297 0.554

(6.250) (-0.187) (0.003) (6.392) (-3.461) (-1.701) (6.171) (0.119) (0.302) High - Low 0.274 -3.804 -3.550 -0.170 -4.286 -3.166 0.521 -3.757 -3.624

(-0.249) (-3.449) (-3.104) (-0.418) (-1.899) (-1.576) (-0.155) (-3.510) (-3.391)

Panel C: Alphamret (%; monthly abnormal return based on monthly portfolio regression) Sample size 168 168 168 60 60 60 108 108 108 Full sample 2.025 -2.407 -2.203 3.262 -3.872 -3.513 1.338 -1.690 -1.584

(0.001) (0.000) (0.000) (0.000) (0.000) (0.000) (0.008) (0.000) (0.000) High ICG 1.774 -3.266 -2.821 2.356 -4.416 -3.564 1.450 -2.609 -2.393

(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Medium ICG 2.562 -2.491 -2.302 4.732 -4.215 -3.840 1.301 -1.650 -1.567

(0.001) (0.000) (0.000) (0.000) (0.000) (0.000) (0.007) (0.000) (0.000) Low ICG 1.609 -1.875 -1.710 2.209 -2.690 -2.395 1.276 -1.491 -1.399

(0.008) (0.000) (0.000) (0.005) (0.000) (0.000) (0.006) (0.000) (0.000) High - Low 0.164 -1.391 -1.111 0.147 -1.727 -1.169 0.174 -1.118 -0.994

(0.201) (0.000) (0.002) (0.321) (0.016) (0.048) (0.218) (0.001) (0.005)

39

Table 4 Internal governance and profitability of insider trading This table presents the results of regressions of six-month market-adjusted abnormal returns (BHAR6m) earned by insiders from their trading on the effectiveness of internal corporate governance and other control variables. The sample consists of 463,527 insider transactions made by officers and directors between 1998 and 2011. Variables are defined in Appendix 2. P-values in parentheses are based on standard errors clustered at the firm level. ***, **, and * stand for statistical significance based on two-sided tests at the 1%, 5%, and 10% levels, respectively. Gov stands for internal governance measure.

Gov = Independence Gov = Compensation Gov = Institution Gov = ICG Gov = ICG1 Gov = log(ISS) Purchases Sales Purchases Sales Purchases Sales Purchases Sales Purchases Sales Purchases Sales

Gov -0.129 (0.730)

-0.321** (0.015)

2.847 (0.126)

-2.765*** (0.000)

0.657 (0.154)

-0.801** (0.027)

0.700 (0.189)

-0.585*** (0.000)

0.361 (0.191)

-0.319** (0.011)

0.784 (0.886)

-2.740** (0.025)

log (Size) -3.642*** (0.000)

-0.081 (0.885)

-4.075*** (0.000)

0.324 (0.507)

-3.501*** (0.000)

-0.124 (0.833)

-3.673*** (0.000)

-0.282 (0.620)

-3.143*** (0.000)

-0.523 (0.395)

-3.936*** (0.000)

-0.132 (0.850)

MB -0.410** (0.014)

0.373** (0.028)

-0.500*** (0.003)

0.367** (0.014)

-0.405** (0.015)

0.373** (0.027)

-0.410** (0.014)

0.381** (0.025)

-0.442*** (0.006)

0.393** (0.021)

-0.121 (0.299)

0.617*** (0.000)

Prior6m -0.037 (0.138)

0.020 (0.205)

-0.052** (0.035)

0.014 (0.350)

-0.035 (0.153)

0.020 (0.199)

-0.040* (0.099)

0.019 (0.216)

-0.040 (0.104)

0.020 (0.204)

0.110*** (0.001)

0.020 (0.297)

TradeSize -1.579 (0.182)

0.881 (0.194)

-1.707 (0.128)

0.734 (0.266)

-1.456 (0.215)

0.894 (0.188)

-1.450 (0.209)

0.835 (0.211)

-1.620 (0.171)

0.946 (0.159)

1.555 (0.594)

1.033 (0.347)

RecentTrades -0.780* (0.071)

-0.366** (0.025)

-0.825* (0.056)

-0.431** (0.015)

-0.776* (0.073)

-0.354** (0.029)

-0.776* (0.072)

-0.395** (0.017)

-0.776* (0.071)

-0.337** (0.038)

-0.877*** (0.002)

-0.041 (0.879)

RND 1.618 (0.205)

-1.021** (0.031)

1.747 (0.168)

-1.026** (0.023)

1.608 (0.203)

-1.037** (0.029)

2.175* (0.095)

-0.784* (0.094)

2.226* (0.090)

-0.715 (0.132)

0.610 (0.572)

-0.062 (0.907)

Loss -11.451*** (0.000)

8.104*** (0.000)

-10.167*** (0.000)

7.230*** (0.000)

-11.473*** (0.000)

8.097*** (0.000)

-11.710*** (0.000)

8.295*** (0.000)

-11.629*** (0.000)

8.385*** (0.000)

-5.370** (0.047)

9.936*** (0.000)

Dispersion 3.729*** (0.009)

0.952 (0.251)

3.729*** (0.009)

1.202 (0.143)

3.653** (0.012)

0.933 (0.258)

3.646** (0.011)

0.962 (0.247)

3.653** (0.012)

0.943 (0.250)

-1.397 (0.538)

0.262 (0.919)

ITP -1.368 (0.141)

-1.274** (0.021)

-1.275 (0.144)

-1.473** (0.012)

-1.326 (0.117)

-1.251** (0.022)

-1.094 (0.501)

-1.351** (0.018)

-1.179 (0.173)

-1.335** (0.018)

-0.673 (0.664)

-0.817** (0.049)

GC -2.527 (0.199)

-2.837** (0.012)

-2.727 (0.182)

-2.590** (0.019)

-2.289 (0.124)

-2.778** (0.013)

-1.978 (0.118)

-2.760** (0.014)

-1.973 (0.107)

-2.817** (0.011)

-1.402 (0.110)

-1.649** (0.027)

Intercept 18.204*** (0.000)

3.299 (0.249)

18.661*** (0.000)

2.715 (0.303)

17.134*** (0.000)

3.118 (0.282)

16.444*** (0.000)

4.928 (0.101)

15.613*** (0.000)

5.283* (0.096)

11.540 (0.547)

9.443 (0.592)

Fixed effects Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Industry /Year

Adjusted R2 18.76% 3.40% 19.85% 4.72% 18.85% 3.44% 19.15% 3.83% 18.96% 3.67% 16.15% 7.71% Sample size 31,840 431,687 31,840 431,687 31,840 431,687 31,840 431,687 31,840 31,687 11,795 242,811

40

Table 5 Sensitivity analysis This table presents the results of sensitivity analyses for the regressions of abnormal returns earned by insiders following their trading. The dependent variable is the six-month market-adjusted abnormal returns (BHAR6m) earned by insiders from their trading unless otherwise indicated. The sample consists of 463,527 insider transactions made by officers and directors between 1998 and 2011. Panel A reports the results using the alternative model specifications. Panel B presents the results including the interaction terms between ICG and ITP (GC). Panel C provides the results including the transaction by large shareholders as an additional sample to examine the role of governance on the profitability of transactions made by each type of insider. Panel D reports the results from 3SLS regressions in which industry median ICG, industry median ITP, and industry median GC are used as the instrumental variables for ICG, ITP, and GC, respectively. The reported results are the coefficient estimates on instrumental variables in the first three rows and the coefficient estimates on instrumented ICG in the fourth row. All variables are defined in Appendix 2. P-values in parentheses are based on standard errors clustered at the firm level. ***, **, and * stand for statistical significance based on two-sided tests at the 1%, 5%, and 10% levels, respectively. Panel A: Alternative model specifications Specifications Alpha6m as the dependent variable

Purchases 0.538

(0.221)

Sales -0.408***

(0.003)

Specifications Firm fixed effects

Purchases 0.562

(0.138)

Sales -0.808***

(0.001) Log(Following) as a transparency measure

0.732 (0.168)

-0.569*** (0.000)

Insider fixed effects 0.680 (0.127)

-0.839*** (0.002)

Earnings quality as a transparency measure

0.853 (0.146)

-0.630*** (0.000)

ΔICG as an internal governance measure

0.558 (0.142)

-0.446** (0.037)

Pre-SOX period as the sample 0.605 (0.117)

-0.614*** (0.003)

ΔICG1 as an internal governance measure

0.250 (0.214)

-0.170** (0.027)

Post-SOX period as the sample 0.794 (0.122)

-0.571*** (0.001)

Log(ΔISS) as an internal governance measure

0.246 (0.823)

-0.425** (0.046)

Trades aggregated at the firm-daily level

0.563 (0.246)

-0.568*** (0.000)

Panel B: Including the interaction terms between ICG and ITP (GC) Type ICG ITP GC ICG × ITP ICG × GC Purchases 0.548

(0.204) -1.028 (0.293)

-1.746 (0.288)

-0.774 (0.105)

-0.347 (0.525)

Sales -0.453*** (0.005)

-1.074 (0.130)

-2.440 (0.110)

-0.477** (0.031)

-0.253** (0.046)

Panel C: Including insider transactions made by officers, directors, and large shareholders Type ICG ×

Officer ICG ×

Director ICG × Large shareholder

Controls Sample size

Purchases 0.759 (0.206)

0.524 (0.253)

-1.041 (0.331)

Included 38,052

Sales -0.638*** (0.000)

-0.709*** (0.001)

-1.360 (0.369)

Included 604,349

Panel D: Key results of 3SLS regressions Specification Purchases Sales

ICG as a dependent variable and industry median ICG as an instrument 0.745*** (0.000)

0.748*** (0.000)

ITP as a dependent variable and industry median ITP as an instrument 0.091*** (0.000)

0.338*** (0.000)

41

GC as a dependent variable and industry median GC as an instrument 0.097*** (0.000)

0.550*** (0.000)

BHAR6m as a dependent variable and instrumented ICG, ITP, and GC plus other control variables as independent variables 0.903

(0.184) -1.325***

(0.000)

42

Table 6 Effects of internal governance on the profitability of insider trading for firms with high legal risk and for opportunistic insider trades

This table presents the results of regressions of six-month market-adjusted abnormal returns (BHAR6m) earned by insiders from their trading on the effectiveness of internal corporate governance (ICG) and other control variables. The sample consists of 463,527 insider transactions made by officers and directors between 1998 and 2011. Two variables are used to measure firms’ litigation risk. High Legal Risk is an indicator that takes the value of one if the ex-ante litigation risk estimated using the Rogers and Stocken (2005) approach is higher than the 90 percentile and zero otherwise. Opportunistic Trade is an indicator that takes the value of one for transactions by insiders who do not trade stocks in the same calendar month in the past three years and zero otherwise. Other variables are defined in Appendix 2. P-values in parentheses are based on standard errors clustered at the firm level. ***, **, and * stand for statistical significance based on two-sided tests at the 1%, 5%, and 10% levels, respectively.

Category indicator used:

High Legal Risk Category indicator used:

Opportunistic Trade Purchases Sales Purchases Sales

ICG: (1) 0.637 (0.141)

-0.387*** (0.010)

-0.387 (0.540)

-0.265** (0.012)

Category (indicator) -1.027 (0.540)

4.193*** (0.000)

1.167*** (0.036)

1.058*** (0.018)

ICG × Category (indicator): (2) 0.195 (0.680)

-0.631** (0.045)

-1.109 (0.186)

-0.585** (0.027)

log (Size) -3.284*** (0.000)

-1.856*** (0.004)

-3.867*** (0.000)

-0.271 (0.637)

MB -0.406** (0.013)

0.361** (0.029)

-0.465*** (0.009)

0.385** (0.027)

Prior6m -0.044* (0.063)

0.021 (0.171)

-0.046* (0.057)

0.019 (0.214)

TradeSize -1.404 (0.218)

0.885 (0.184)

-1.300 (0.262)

0.823 (0.224)

RecentTrades -0.769* (0.069)

-0.372** (0.026)

-0.757* (0.082)

-0.365** (0.029)

RND 2.191* (0.094)

-0.731 (0.109)

2.614** (0.026)

-1.512*** (0.003)

Loss -11.621*** (0.000)

7.787*** (0.000)

-13.016*** (0.000)

7.612*** (0.000)

Dispersion 3.693*** (0.010)

0.725 (0.395)

3.827*** (0.009)

1.393 (0.101)

ITP -1.026 (0.226)

-1.479** (0.012)

-1.136 (0.183)

-1.323** (0.018)

GC -1.958 (0.121)

-2.796** (0.010)

-2.150 (0.179)

-2.827** (0.012)

Intercept 14.998*** (0.000)

11.520*** (0.000)

17.957*** (0.000)

4.745** (0.013)

Fixed effects Industry /Year Industry /Year Industry /Year Industry /Year Adjusted R2 19.17% 4.32% 19.63% 3.95% Sample size 31,840 431,687 31,840 431,687 Test: (1) + (2) = 0 0.268 0.002 0.124 0.000

43

Table 7 Likelihood of adopting ex-ante preventive measures of insider trading This table presents the results of probit regressions in which the dependent variable is either an indicator that takes the value of one if the firm adopts a voluntary insider trading policy and zero otherwise (ITP), or an indicator that takes the value of one if the firm requires a general counsel’s pre-approval prior to insider trading and zero otherwise (GC). The sample consists of 10,339 firm-year observations between 1998 and 2011. Volatility is the annualized standard deviation of daily market adjusted stock returns. InsOwn is the ownership by firm officers. Ins_Trade is the ratio of total insider trading volume over a year to shares outstanding. Following is the number of analysts following a firm in a year. Litigation probability is the ex-ante litigation likelihood estimated using the Rogers and Stocken (2005) approach. Other variables are defined in Appendix 2. P-values in parentheses are based on standard errors clustered at the firm level.***, **, and * stand for statistical significance based on two-sided tests at the 1%, 5%, and 10% levels, respectively.

Dependent variable = Indicator for

adopting ITP (1)

Dependent variable = Indicator for adopting GC

(2)

ICG 0.006** (0.035)

0.013** (0.015)

log (Size) 0.052** (0.038)

0.110*** (0.001)

log (Volatility) 0.065** (0.011)

0.075*** (0.008)

InsOwn -0.677*** (0.002)

-0.873*** (0.000)

Ins_Trade -0.001 (0.326)

-0.001 (0.477)

log (Following) 0.141*** (0.000)

0.108*** (0.006)

Institutional ownership 0.268*** (0.005)

0.196** (0.015)

Litigation probability 0.212** (0.032)

0.246** (0.024)

Intercept 0.021 (0.893)

0.080 (0.332)

Fixed effects Industry /Year Industry /Year Adjusted R2 30.45% 7.87% Sample size 10,339 10,339

44

Table 8 Likelihood of taking ex-post disciplinary actions (forced CEO turnover)

The table provides the results of probit regressions of forced CEO turnover on the average six-month market-adjusted abnormal returns (Avg_BHAR6m) earned by CEOs from their trading. In columns (1) and (2), the sample consists of 9,954 firm-year observations with at least one CEO transaction during the previous fiscal year between 1992 and 2010. In columns (3) and (4), the sample is restricted to those between 1998 and 2010 during which data on ICG are available. The dependent variable is equal to one if there is forced CEO turnover in a given year and zero otherwise. Forced CEO turnover is assumed to occur if a CEO is forced to leave her position in the fiscal year following CEO transactions. Appendix 2 provides the details about how we classify CEO turnover as voluntary or forced. Avg_BHAR6m is the value-weighted average market-adjusted abnormal buy-and-hold return over 180 calendar days subsequent to the insider trading date for trades during the previous fiscal year. Ret is the difference between annual stock return of the firm and the median annual return of firms in the same Fama-French (1997) 48 industry. ROA is the ratio of earnings excluding extraordinary items to total assets. RetInd is the median annual returns of firms in the same industry. SG is the annual growth in firm’s sales. RetVol is the volatility of daily stock returns of the firm during the fiscal year. Lev is total liabilities as a ratio of total assets. Other variables are defined in Appendix 2. All control variables are lagged for one year to capture the characteristics of the firms before the year of forced CEO turnover. P-values in parentheses are based on standard errors clustered at the firm level. ***, **, and * stand for statistical significance based on two-sided tests at the 1%, 5%, and 10% levels, respectively.

Purchases (1)

Sales (2)

Purchases (3)

Sales (4)

Avg_BHAR6m -0.076 (0.707)

0.328** (0.042)

-0.153 (0.238)

0.170* (0.064)

Ret -0.597*** (0.000)

-0.164** (0.011)

-0.561* (0.067)

-0.162 (0.204)

ROA -0.630 (0.211)

-0.374 (0.290)

-0.697 (0.487)

-0.514 (0.397)

RetInd 0.008 (0.974)

-0.153 (0.320)

-0.225 (0.670)

-0.038 (0.897)

SG -0.179 (0.152)

0.022 (0.455)

0.158 (0.614)

-0.399* (0.087)

RetVol 3.092 (0.554)

8.689** (0.011)

9.063 (0.336)

6.699 (0.277)

MB 0.025 (0.137)

-0.008 (0.411)

0.035 (0.202)

0.009 (0.646)

log (Size) 0.122*** (0.002)

0.079*** (0.001)

0.138* (0.058)

0.043 (0.249)

Lev -0.676** (0.020)

0.130 (0.491)

-0.814 (0.116)

-0.213 (0.492)

ICG -0.028 (0.230)

-0.035*** (0.005)

Avg_BHAR6m × ICG 0.103 (0.288)

0.202** (0.012)

Intercept -10.644 (0.974)

-3.402*** (0.000)

-10.589 (0.973)

-2.954*** (0.000)

Fixed effects Industry /Year Industry /Year Industry /Year Industry /Year Pseudo R2 5.14% 1.41% 8.65% 2.62% Sample size 2,308 7,646 892 3,857

45

Appendix 1 Description of internal corporate governance (CG) component variables This appendix provides a detailed description of the construction of the internal corporate governance variables used in the tables.

Category Item CG item Good CG Definition Source Reference

Independence

1 Board independence + Percentage of outside directors on the board. RiskMetrics

Laksmana (2008), Hoitash et al. (2009), Li and Srinivasan (2011), Chen et al. (2012), Chhaochhaia et al. (2012), Hazarikaet al. (2012), Masulis et al. (2012), Morellec et al. (2012) and Wintoki et al. (2012).

2 Compensation committee independence

+ Percentage of outside directors on the compensation committee. RiskMetrics Laksmana (2008) and Chhaochhaia et al. (2012).

Compensation

3 CEO pay-performance sensitivity

+

Ratio of the value of all outstanding stock options at year-end plus the value of all shares owned by the CEO at year-end to total compensation, including salary, bonus, the value of all outstanding option grants, and shares owned at year-end.

ExecuComp Laksmana (2008), Li and Srinivasan (2011) and Hoechle et al. (2012).

4 Directors receive shares or options + An indicator that equals one if any nonexecutive

directors receive shares or options or both. ExecuComp Hoechle et al. (2012),

Institution

5 Institutional ownership +

Percentage of shares held by institutional investors.

Thomson Financial

Laksmana (2008), Li and Srinivasan (2011), Chen et al. (2012), Hazarika et al. (2012) and Morellec et al. (2012).

6

Independent & long-term institutional ownership

+ Percentage of shares held by the top five independent, long-term, and dedicated/quasi-indexer institutional investors.

Thomson Financial Chen et al. (2007).

Others

7 Board size - Number of directors on the board. RiskMetrics Laksmana (2008), Hoitash et al. (2009), Li and Srinivasan (2011), Chen et al. (2012), Hazarika et al. (2012), Hoechle et al. (2012), Masulis et al. (2012) and Wintoki et al. (2012).

8 Old director percentage - Percentage of directors who are older than 72. RiskMetrics Armstrong, Ittner, and Larcker (2012) and Hoechle et al.

(2012).

9 Busy director percentage -

Percentage of outside directors who hold three or more board directorships.

RiskMetrics Laksmana (2008), Hoitash et al. (2009), Hoechle et al. (2012) and Masulis et al. (2012).

10 Block independent director +

An indicator that equals one if a firm has an independent director that is a large shareholder and zero otherwise.

RiskMetrics Li and Srinivasan (2011), Masulis et al. (2012) and Armstrong et al. (2012).

46

11

Fraction of directors whose tenure predates CEO

+ Percentage of outside directors appointed before the current CEO took office. RiskMetrics Laksmana (2008), Hazarika et al. (2012) and Armstrong et

al. (2012).

12 CEO⁄ Chairman Duality -

An indicator that equals one if the CEO and the chairman of the board are the same person and zero otherwise.

ExecuComp Chen et al. (2012), Hazarika et al. (2012), Masulis et al. (2012) and Wintoki et al. (2012).

13 Insider ownership + Percentage of equity ownership by all board members and all top 5 officers.

RiskMetrics and ExecuComp

Li and Srinivasan (2011), Masulis et al. (2012) and Hazarika et al. (2012).

47

Appendix 2 Description of variables

This appendix provides a detailed description of the construction of the variables used in the main tables in the alphabetical order. Variable Definition

Alpha6m Average daily abnormal return estimated using Carhart’s (1997) four-factor model over 180 calendar days subsequent to the insider trading date, multiplied by -1 for insider sales transactions.

Alphamret

The intercept from time-series monthly portfolio regressions estimated using the Carhart’s (1997) four-factor model. To form the portfolio, in each calendar month, we first assign a sample firm into a net purchase (net sales) group based on the net number of shares purchased by insiders of the firm during the month. Then, in each month, we form net purchase (sale) portfolios using all firms classified as net purchase (sales) firms at least once over the past six-month period and calculate value-weighted portfolio returns. For aggregate portfolios, we include all firms in net purchase and sale portfolios, but multiply the returns of the firms in net sale portfolios by -1.

BHAR6m Market-adjusted abnormal buy-and-hold return over 180 calendar days subsequent to the insider trading date, multiplied by -1 for insider sales transactions.

Compensation The sum of the measures of the CEO pay-performance sensitivity and whether the directors receive shares or options at a common scale with a zero mean and unit variance.

Dispersion Monthly standard deviation of current-fiscal-year EPS forecasts divided by the mean of the forecasts for the most recent fiscal quarter.

Earnings quality

The absolute value of discretionary accruals estimated using the modified Jones model (Dechow et al., 1995).

Following Number of analysts following a firm in a year.

GC Indicator that takes the value of one for firms with a general counsel pre-approval requirement and zero otherwise.

ICG The sum of independence, compensation, and institution at a common scale with a zero mean and unit variance, with a higher value indicating better governance.

ICG1 The sum of independence, compensation, institution, and other seven governance attributes described in Appendix 1 at a common scale with a zero mean and unit variance, with a higher value indicating better governance.

Independence The sum of the measures of board independence and compensation committee independence at a common scale with a zero mean and unit variance.

Ins_Trade Ratio of total insider trading volume over a year to shares outstanding. InsOwn Ownership held by firm officers.

Institution The measure of institutional ownership concentration at a common scale with a zero mean and unit variance.

Institutional ownership Percentage of shares held by institutional investors

ISS

Corporate governance score constructed by Institutional Shareholder Services, equal to the number of minimally acceptable governance attributes met by a firm out of 64 governance attributes along eight major dimensions, with a higher ISS score indicating better governance. For a detailed description of the governance attributes used in the calculation of ISS, see Aggarwal and Williamson (2006) and Aggarwal et al. (2009).

ITP Indicator that takes the value of one for firms with an insider trading policy and zero otherwise. Lev Total liabilities as a ratio of total assets.

Litigation probability

Ex-ante litigation likelihood estimated using the Rogers and Stocken (2005) approach. Specifically, it is estimated using a probit model in which the dependent variable is an indicator that takes the value of one if a Rule 10b-5 lawsuit is filed against a firm in a given quarter and zero otherwise. The independent variables include firm size, beta, daily turnover, cumulative quarterly return, standard deviation of daily returns, minimum of daily returns, and skewness of daily returns and indicators for high-risk industries.

48

Loss Indicator that takes the value of one if net income before extraordinary items in the most recent fiscal year is strictly negative and zero otherwise.

MB Ratio of the market value of equity to the book value of equity in the most recent fiscal quarter.

Prior6m Cumulative market-adjusted excess return over 180 calendar days prior to the insider trading date, multiplied by -1 for insider sales transactions.

RecentTrades Sum of absolute values of the daily net numbers of shares purchased by all insiders of a firm during ten days prior to the transaction date, scaled by the total number of shares outstanding of the firm.

Ret Difference between annual stock return of the firm and the median annual return of firms in the same Fama-French (1997) 48 industry.

RetInd Median annual returns of firms in the same industry. RetVol Volatility of daily stock returns of the firm during the fiscal year.

RND Indicator that takes the value of one if a firm reports non-zero R&D expenditures in the most recent fiscal year and zero otherwise.

ROA Ratio of earnings excluding extraordinary items to total assets. SG Annual growth in firm’s sales. Size The inflation-adjusted market value of equity at the end of the most recent fiscal quarter.

TradeSize Absolute value of the net number of shares purchased by all insiders of a firm on the transaction date divided by the total number of shares outstanding of the firm.

Volatility Annualized standard deviation of daily market adjusted stock returns.

49