Is Disciplinary Action against a Firm Threat Enough for. Corrective Actions in Peers?

Transcription

1 Is Disciplinary Action against a Firm Threat Enough for Corrective Actions in Peers? Buvaneshwaran Venugopal Preliminary Version (January 2016) Abstract External disciplinary actions against companies are taken to punish inappropriate behavior of companies. But more importantly they are supposed to be a deterrent to others from committing the same mistake. This paper investigates if external disciplinary mechanism is an effective deterrent. Specifically, does litigation of a firm forces its peers to take corrective actions and address their agency issues? Using 2615 class action lawsuits filed against companies during the period 1996 to 2013, this paper shows that external disciplinary mechanism leads to corrective measures even from peers who were not litigated. When a class action lawsuit is filed against a firm, other companies reduce their over-investment by 3.5%, increase leverage and decrease top executives compensation by 11.2%. Boards become less myopic and extend their performance evaluation horizon by increasing their CEO s stock vesting schedule i.e. reducing CEO s newly vested shares by 12%. The effects felt by the peers range from 10 to 24% of effects felt by the target of an external disciplinary action. C. T. Bauer College of Business, University of Houston; bgvenugopal@uh.edu

2 Introduction People will engage in criminal and deviant activities if they do not fear apprehension and punishment..., general deterrence theory focuses on reducing the probability of deviance in the general population. Rational Choice and Deterrence Theory. Separation of ownership and control is considered to be the source of numerous agency problems and moral hazard. These issues are combated using either internal disciplinary actions or external disciplinary mechanisms (Shleifer and Vishny, 1997). Internal actions are taken typically by a concerned board or institutional investors by threating a vote. External actions are taken by either the regulator, by opening investigations, or the shareholders, by filing class action suits. The objective of external disciplinary actions is to punish companies that have not been acting in the best interests of it s shareholders. But more importantly, such external disciplinary actions serve as a threat to other detractors in the industry and deter companies in the market from wrongdoings. The aim of this paper is to address the question: Is external disciplinary action on a firm an effective for other companies in the industry to take corrective actions addressing agency issues? Existing literature has extensively studied the impact of external disciplinary action on the company it is targeted at. But little is known about the spillover effects of such actions on other firms in the industry. This study attempts to fill this gap in understanding the impact of external corporate disciplinary mechanism. External disciplinary actions such as Security Exchange Commission (SEC) investigations and shareholder initiated Securities Class Action (SCA) lawsuits have been shown to negatively impact the growth and reputation of companies (Karpoff et al., 2008), leading to executives losing their jobs (Karpoff et al., 2008) and board members leaving (Fich and Shivdasani, 2007; Brochet and Srinivasan, 2014). Such disciplinary actions attract huge median attention and further add to prolonged negative publicity for the companies involved. Given that companies that face investigations or lawsuits experience low stock performance (Gande and Lewis, 2009) and huge reputation losses, other companies in the same industry would want to avoid any disciplinary actions directed 1

3 on them. Thus, an external disciplinary action against one firm in the industry could act as a shock to the governance systems and force others in the industry to take corrective steps. The external disciplinary shock to an industry can take the form of SEC investigation, class action lawsuits or any publicized litigation against a company or it s executive. In this paper I use class action lawsuits filed against publicly listed companies during the period 1996 to 2013 to identify the occurrence of disciplinary shocks 1 The data was collected from Standford Law School s Securities Class Action Clearinghouse database. Following a class action law suit, we would expect that the targeted company would reduce it s over-investment (McTier and Wald, 2011), compensation to CEO (Banerjee et al., 2015), make changes to it s board structure (Ferris et al., 2007), etc. These are corrective actions to address agency problems that were present in the company. Kim and Skinner (2012) show that litigations are clustered around industries in a point of time. Thus, a class action suit filed against a company would make it s peers worry if they would be the next target. This is in-line with Becker (1968) theory of deterrence effect, where potential offenders would be deterred from committing a crime fearing punishment. Assuming that the industry participants are closely watching the proceedings of external disciplinary shock to their industry, they would also undertake certain corrective actions that would be appropriate for their firm. This would be reflected in their operational decisions or incentive mechanisms during years following a shock to the industry. On the operational decisions side, I look at over-investment, payouts, cash retained and amount of debt added to the financing structure of companies in the year following a disciplinary shock to the industry. Whereas on the incentives side I study the changes to compensation, bonus and stock vesting schedule of CEOs and top executives. The results show that there is a 3.5% decrease in over-investment of peer firms in year following the shock. This is about 10% of the effect felt by the target company of the class action suit. The spillover effect for reduction in dividends, increase in cash retained and leverage ranges about 10 to 20%. For the years prior to a disciplinary shock, there are no such spillover effects, suggesting that class action suit filed against a peer in 1 The next version of this paper also includes SEC & DOJ initiated investigations against companies or executives to identify external disciplinary actions. 2

4 th industry jolts other companies to take actions to address issues in their company. The total compensation of CEOs and top executives falls about 3% in the year following a class action suit in the industry. But the surprising result comes from the vesting schedule. Following a shock to the industry, the number of newly vested shares to the CEO drops by 11.2%. The corresponding number for other top executives is about 10%. Though a reduction was expected, the magnitude of reduction is high, suggesting that the shock might have been a wake-up call for the board to reevaluate their prior decisions. Edmans et al. (2015) and Gopalan et al. (2014) suggest that vesting schedule of CEO s shares represent the investment horizon focus of a company. For example, if a company is myopic in its focus then the vesting schedule will be short i.e. the number of newly vested shares each year would be high. Coupled this with the reduction in newly vested shares result of this paper, we can say that companies tend to increase their investment and performance horizon following a disciplinary shock to their industry. Overall this study shows that an external disciplinary shock to a firm makes peers in the industry to take corrective action. Thus a credible threat of litigation may be enough to make market participants fall in line. This paper differs from existing literature by looking at the corrective actions peer firms take following a disciplinary shock. The current literature, with very few exceptions, focuses mostly on the effects of litigation on the company it is targeted on. While it is paramount to understand the effect of punishment on the target, the aim of publicly levying punitive damages on detractors is to deter other from committing the same mistakes. Thus understanding the spillover effects of disciplinary actions is important. This paper is closely related to the literature that studies corporate governance disciplinary mechanisms 2. Specifically relevant are studies on external disciplinary mechanisms. Karpoff et al. (2008) looked at impact investigations by SEC and DOJ has on executives accused of wrongdoings. The authors conclusively showed that 93% of the individuals lose their jobs. Banerjee et al. (2015) show that over-confident CEOs are more likely to attract class action lawsuits in the near future. Peng and Röell (2008) and Armstrong et al. (2013) executives whose compensation are more sensitive to stock price changes tend to misreport more and eventually face class action lawsuits. 2 Please see Shleifer and Vishny (1997) for a survey of tools available to stakeholders. 3

5 Humphery-Jenner (2012) showed that following a class action lawsuit, internal and external disciplinary mechanisms kick-in and the firm is more likely to be taken over as a market disciplining approach. This paper differs in that it looks at the impact of external discipline on peer firms. This paper is also related to the growing peer effects literature. Bourveau and Spira (2014) is the closest to this paper. The authors investigate the impact of class action lawsuits filed against companies after a M&A deal. They find that following a lawsuit, peers in the industry respond by closing high quality deals and reduce suboptimal investments. Servaes and Tamayo (2014) study the impact a hostile takeover has on other firms in the industry. The argue that peers cut spending and adopt more takeover defenses. These papers are very close in spirit to my paper. The peer group effect literature has also concentrated on CEO compensation (Bizjak et al., 2011), and corporate investments. The reminder of this paper is organized as follows: Section 1 discusses the main hypotheses and provides empirical predictions; data is discussed in section 2; section 3 discusses the empirical tests and results; finally section 4 provides concluding remarks. 1 Hypotheses and Empirical Predictions When disciplinary actions are taken against a firm, peers in the industry are bound to take notice. Getting litigated would result in huge reputation losses (Karpoff et al., 2008), prolonged stock under-performance (Gande and Lewis, 2009) and huge financial penalties (Strahan, 1998; Arena and Julio, 2015). Fearing such repercussions, peers in the industry would take corrective measure to avoid the spotlight of disciplinary action. Thus the fear of being punished or identified as an offender should keep other companies in the industry from committing an offense (Becker, 1968). Therefore we would expect that following a disciplinary shock to the industry, all companies would try to reduce their agency issues. These corrections can take the form of operational decisions or changes to incentive mechanisms. McTier and Wald (2011) showed that companies that have been over investing are more likely to 4

6 attract a lawsuit. Therefore after a shock, peer companies would reduce their over-investments. This is our first hypothesis. Also in an attempt to create a better bonding mechanism as suggested by Jensen (1986), companies may increase their level of debt. This may lead to an increase in cash and decrease in payouts to shareholders. Secondly, companies may also change their compensation strategy. For example, firms that have been overpaying their CEOs and executives may become more cautious when the industry is faced with a disciplinary shock. This would lead to a decrease in compensation packages of CEOs and other top executives. Additionally, companies that have been focused on short-term gains may want to change their focus. Armstrong et al. (2013) suggest that executives whose compensation is more sensitive to stock price movements are more likely to misrepresent earnings. This would attract litigation when fraud is uncovered. Companies that are myopic in their investment and performance horizons are more likely to be those that set compensation packages that are highly sensitive to stock price movements. Thus a disciplinary or governance shock to the industry may make boards reevaluate their strategy. This should be reflected in the vesting schedule of CEO s stocks and options. After an external shock, we can expect the number of newly vested stocks to drop. Overall, a disciplinary shock to the industry via litigation of a firm should make other companies in the industry take corrective actions. 2 Data The data for this study comes from four different sources: Stanford Law School s Securities Class Action Clearinghouse, Compustat, Execucomp and RiskMetrics-ISS databases. To identify which companies have faced a lawsuit, I use Stanford Law School s Securities Class Action Clearinghouse (SCAC) database. As a result of Private Securities Litigation Reform Act (1995) the SCAC contains information on around 4090 class action suits filed starting from the year The database contains information on date a class action suit was filed, which is typically the date of first complaint; period relevant to the suit; current status (ongoing, dismissed, settled); 5

7 information on plaintiffs and defendants; case summary, rulings, etc. The objective of this study is to identify the effect external disciplinary action on a firm has on its peer firms or the industry. Therefore, ideally I would like to include not just class action lawsuits but also investigations opened by SEC on companies and any other litigation a firm might have faced as a disciplinary shock to the system (Next version of this paper includes such shocks in the analysis). Panel A of table 1 shows the sector-wise distribution of class action suits. Manufacturing sector, which includes consumer products, chemical, etc. faced the highest number of lawsuits followed by the Technology sector. Panel B shows that 2001 was the year with highest number of lawsuits filed, many of which came from the Technology sector. Recent years have seen increased filings in Finance and Real Estate sectors as well. All filings from SCAC were downloaded using a perl web-crawling script and primary fields such as Company name, Ticker, Date filed, Class period and Status are converted into table format. Using ticker symbols, companies were matched with financial information in Compustat North America file, which contains financial information on more than 29,309 companies. To be included in the sample, the firm should have financial information for at least three years prior to the filing date and at least one year after the suit was filed. This criteria ensures that the company is known in the industry prior to the litigation and peers take notice of the lawsuit proceedings. We were able to match 2615 lawsuits with publicly traded companies. Using North American Industry Classification (NAICS) system we identify the two digit industry codes which faced a litigation each year. In each of the 20 industry groups, if at least one firm faced a litigation in a given year then the industry and it s companies are said to be subject to an external disciplinary action during that year. Having identified the years and industries that were subject to a class action suit, next we merge this data with Execucomp database, which has executive compensation information for roughly 3,400 companies. This reduces our sample size substantially. For each matched company, we have data on the CEO s total compensation, bonus, shares owned by the CEO, number of vested and un-vested shares/options held by a CEO, new vested shares/options, CEO s age, tenure, etc. Similar information is available for other top executives as well. Additionally, I create two 6

8 variables: (i) CEO-Chairman, which identifies CEO s dual role as CEO and Chairman of the board and (ii) CEO-Chair-President, which identifies if the CEO-Chairman is also the President of the company. These variables represent power a CEO holds over his board. The CEO-Chairman has higher control over the board s voting decisions, whereas a CEO-Chair-President leaves the board with no recourse (Morse et al., 2011). Next, the data is merged with ISS (formerly Riskmetrics) data to get information on boards. Since ISS contains data only till 2013, I restrict the study to the period 1996 to From ISS, we can determine the board size, % independent directors on the board, busyness of the board members 3, CEO committee-ships, etc. Merging with ISS further reduces our sample size. Hence during analysis, we test our hypotheses at different stages of the merging process i.e. first with the biggest sample SCAC with Compustat, and next with the fully merged dataset SCAC, Compustat, Execucomp and ISS. I do not use companies from utilities and financial sector for this analysis as they are regulated differently. 2.1 Key variables The objective of this study is to measure the impact of any external disciplinary action taken against a single firm on the entire industry. Litigation in industry is a binary variable that indicates the occurrence of such a shock. It takes a value 1 if a class action suit was filed against a firm in the industry in a given year, 0 otherwise. Therefore, the variable Litigation in industry is equal to 1 for all firms in the industry for a year if at least one firm faced a lawsuit. No. of lawsuits captures the magnitude of disciplinary action by counting the number of class action suits against firms in a given industry. These two litigation variables are the primary independent variables of interest in this study. Together they measure the incidence and magnitude of external disciplinary shocks the an industry in a given year. When an firm is under public scrutiny, the management typically takes corrective actions to make sure that it presents a good image in the future. If the level of scrutiny is sufficiently high then 3 A director with more than three directorships is considered busy Fich and Shivdasani (2006). 7

9 we can expect other firms (peers of the litigated firm) in the industry to take similar actions as a preemptive measure. Thus ensuring that they have a clear record in the event they are litigated or at minimum they avoid litigation. The corrective actions can take two forms: changes to firm s operations or changes to incentives of top management. Firm operations or actions can include decreasing over-investment, changes to payouts dividends and repurchases and cash retained. Changes to top management incentives would include changes in compensation or shares vesting schedule of CEOs and other top executives. In this study I consider four actions at the firm level. Companies with free cash flows can either invest them into new projects, pay them out to shareholders or retain them (McTier and Wald, 2011). Each action may have different consequences. First, Over-investment of a firm is defined as the ratio of capital expenditures to total assets over the industry-wide ratio. This variable captures excess investments made by CEOs who might be interested in building empires, starting divisions that would increase their human capital (Jensen, 1986). We could expect that such agency costs increasing actions would decrease after litigation of a peer firm has brought the industry under scrutiny. Following external disciplinary action to a peer firm in the industry if companies are expected to decrease their over-investments then they have two ways of handling excess free-cash flows: pay them out as dividends or retain the cash in reserves. Dividends and Cash Retained are two other corrective actions a company can take. Finally, Repurchase is another method of using free-cash flows. On the incentives front, to avoid the spotlight of litigation companies could reduce compensation packages of their top executives. In this paper, I track compensation by looking at the CEO s Total Compensation, Bonus and number of newly vested shares. A captured board may have been paying it s CEO a very high salary (Morse et al., 2011; Gopalan et al., 2014). But a shock via class action lawsuit to peer firm may make the board reevaluate it s compensation decisions. This could lead to a reduction in the total compensation and bonus issues to the CEO. On the contrary, litigation threat may force a myopic board into looking at long-term benefits rather focusing on short-term gains. This would be reflected in changes to the stock and options vesting schedule of the CEO. Following Edmans et al. (2015), I construct CEO Vested Shares as the number of newly 8

10 vested stocks and options to capture changes in the myopic behavior of the board. 2.2 Descriptive statistics Table 2 summarizes the panel data used for this study. Mean of the binary variable Litigation in industry indicates that roughly 63% of firms in the industry faced an external disciplinary shock during the sample period. The average No. of lawsuits for an industry in a given year was 23. Median shows that both litigation measures in panel A are positively skewed. This means that there are certain industries that inherently face more lawsuits compared to others. Therefore, we would have to conduct a industry level analysis to gain more insight into the effects of external disciplinary actions. Panel B summarizes the firm action variables discussed in section 2.1. The average Over-investment is and standard deviation is 0.92 which means that there is a huge variation in overinvestment across firms. The ratio of dividends to total assets is roughly 2.4% indicating that companies payout about 2% of their assets as dividends to their shareholders. On average, firms repurchase million shares per year in the sample. But it has to be noted that only a very small fraction of firms repurchase their shares in our sample. Companies in our sample on average hold 32% of their sales as cash reserves. Panel C summarizes financial characteristics of companies in our sample. Leverage, Intangibles and Market value are in millions. It has to be noted that panels B and C summarizes characteristics of all firms (over 29,000) in the Compustat universe. Characteristics of CEOs and top executives are summarized in panel D of table 2. It has to be noted that Execucomp covers only 3,407 companies in our Compustat sample and hence the number of firm-year observations drop drastically when including compensation variables in our analysis. The average CEO is about 55 years and is in tenure for about seven years before retirement. CEOs in our sample hold less than 1% of their company s shares and about 50% of the firm-year observations had a CEO with dual role of Chairman of the board. CEOs on average earn 5 million in total compensation per year with a bonus of 0.5 million. The number of newly vested shares in an average year is million. This number varies highly meaning that there 9

11 are some boards that are extremely myopic as expected. Board variables are summarized in panel E. The average size of a board is 10 members with 8.6 years experience in the board. 69% of the board members are independent directors. Following Fich and Shivdasani (2006) I call a director as busy if he sits on more than 3 boards. In our sample 12.3% of the directors are busy independent directors. Independence and busyness of a director plays an important in effective governance of the firm. An independent director may be easily swayed by the CEO but at the same time has less stake (other than reputation) in the well-being of the company. A busy independent director may simply not have the time to appropriately govern due to lack of time. Correlation between key variables are reported in table 3. 3 Empirical analysis In this section I discuss the multivariate analysis on impact litigation of a peer firm has on other firms in the industry. I start by analyzing the factors that contribute to increase in probability of a firm facing external disciplinary actions, class action lawsuit in this case. Next we would look at the effect incidence of class action lawsuit against a firm in the industry has on other companies. The impact of discipline would lead to changes in two fronts: firm operational decisions and executive incentive mechanisms. 3.1 Probability of external disciplinary action taken against a firm A firm can attract scrutiny and external disciplinary actions if it shows signs of mismanagement, poor governance, performance or high level of over-investment in projects compared to industry average 4. Table 4 summarizes Logit regressions run to identify factors that contribute toward increase in probability of a firm getting litigated. The sample for this analysis includes all firms (litigated and non-litigated) in Compustat universe. The dependent variable is all regressions is a binary variable ClassAction t+1 that takes a value 1 if 4 The next version of the paper looks at diversification efforts also. 10

12 a class action lawsuit was filed against a firm in year t+1, 0 otherwise. In column 1, variables that represent firm s operations/corrective actions are regressed on ClassAction t+1. We would expect that firms that are over-investing compared to the industry would be more likely to face lawsuits compared to others. This is confirmed in column 1. Companies that pay their shareholders high dividends, repurchase more shares or retain more cash are less likely to face a class action lawsuit. Except for leverage all corrective action variables are statistically significant and robust to including additional variables. The standard errors are heteroskedasticity robust and clustered at the industry level. In addition, to control for unobserved variables, dummies for firm, industry and years are included in the logit regressions. In column 2, I include firm characteristics in the logit regression. It shows that companies with high stock return performance are less likely to be litigated. Combined with the coefficient of dividends, we can say that companies with good stock performance and history of returning that money to shareholders are less likely to face external disciplinary action. The coefficient of intangibles in column 2 indicates that companies with high level of intangibles are less likely to be sued. This is in contrary to the results of McTier and Wald (2011) where tangible goods were seen as a deterrent to class action suits. The different result, though not robust, could be due to increase in strength and performance on technology sector companies in the past few years. Columns 3 and 4 include CEO and board variables into probability analysis. The results show that companies with CEOs of longer tenures are less likely to be litigated. This could be due to the proven talent of CEOs in the market. The coefficient of CEO Vested Shares, which is the number of newly vested shares or options of the CEO is positive. The variable captures the myopic behavior of the board and is qualitatively equivalent to the CEO compensation duration measure discussed in Gopalan et al. (2014). When the duration or number of newly vested shares is high, it means that the board values short-term performance more compared to long-term performance. This myopic behavior of the board and subsequent CEO compensation mechanism may induce the CEO into taking more risks for short-term benefits and worsen agency costs (Peng and Röell, 2008; Armstrong et al., 2013). The positive sign of this coefficient in columns 3 and 4 indicates that firms with myopic vision attract more class action lawsuit next period. 11

13 To check how size of a firm differentially affects the probability of litigation, the full sample is divided into quintiles based on total assets held by a firm at the end of each year. Column 5 repeats the analysis using large firms (top quintile) while column 6 with small firms (lowest quintile). The results discussed so far hold for both large and small firms. The probability of getting litigated is higher if a small firm over invests compared to large firms. However, it has to be noted that in our sample large firms are more susceptible to litigation compared to small ones due to sheer size and attention. 3.2 Effect of peer litigation on firm operational corrections Now that it is clear from table 4 that corrective action variables i.e. variables that reflect investment activities and payouts play an important role in the probability of getting litigated, we have to look at the changes to these variables after a disciplinary shock is felt by the industry. To do this, companies that faced a class action lawsuits are removed from the sample. This ensures that the direct effect felt by these companies does not upward bias the effect of class action suits on other companies in the industry. The primary independent variables of interest in these analysis is: first, Lawsuit t, a binary variable that takes a value 1 if at least one company in the industry has faced a class action lawsuit in year t; second, No of lawsuits t, i.e. number of lawsuits faced by companies in an industry during the year t. A Effect on over-investment Since over-investment was the primary driver of litigation probability in table 4, I start by looking at the effect of peer litigation on over-investment during years after a class action suit was filed. Table 5 summarizes the results of this test. The dependent variable is Overinvestment t+1, which is the ratio of capital expenditures to total assets in excess of industry mean ratio. All independent variables are lagged (from year t). All regressions include year and industry fixed effects. The standard errors are heteroskedasticity robust clustered around industry and year level. Columns 1 and 2 include only firm level variables in addition to litigation variables. It can be seen 12

14 that when a peer firm faces a litigation in year t, on average other firms in the industry decrease their over-investment by 8.4%. And when the number of litigations increase by one standard deviation, peer firms in the industry decrease their over-investment by 3.5%. The corresponding decrease in over-investment for the litigated firm (company that faced class action suit) is 31.3%. Comparing these numbers reveals that at least 10% of the effect felt by the target of disciplinary action is transmitted to other firms in the industry. This strongly suggests that external disciplinary action taken against any firm in the industry has an impact of other companies and lead to corrective actions even among firms that did not face the disciplinary action. Therefore litigation against a peer firm can be threat enough for other firms to take corrective actions to their operations and address agency issues. Columns 3 and 4 includes board and CEO characteristics to the analysis. The litigation variable remain significant in these regressions. Results show that firms with good stock performance, high sales growth, intangible assets, more independent boards over-invest in the following year. To further investigate the differential impact of litigation on different firm sizes, as discussed in section 3.1 the sample is divided into quintiles. Columns 5 and 6 repeat the analysis for large firms whereas columns 7 and 8 for small firms. The results show that incidence of class action lawsuits in the industry affects only large firms and not the small ones. Only large firms reduce the level of over-investment in the year following a class action suit in the industry. This make sense given that large firms typically have large cash reserves. Another difference between large and small firms is that independent directors in large firms is associated with higher level of overinvestment. Whereas they have the opposite effect in small firms. This could be due to the fact that independent directors in large firms have a huge business to oversee which at times could be difficult while smaller firms are more concentrated and could be easier to govern. B Effect on dividends, cash retained, repurchases and leverage Companies with free cash flows could distribute them via dividends, repurchases or retain the cash for future investments opportunities. It has been shown in the literature that following a class 13

15 action suit, companies change their strategy in handling free cash flows. Then this must be reflected in the behavior of peer firms in the industry that was recently hit by a disciplinary action. At minimum, peer firms would attempt to evade future litigation by taking corrective actions to free cash flow disbursement strategies. Jensen (1986) suggests a stricter bond mechanism by increasing the amount of debt. Following this we could expect a decrease in payouts to shareholders and increase in leverage. Table 6 test these hypotheses. In these regressions, the primary variable of interest is No.oflawsuits t rather than Lawsuits t because as the intensity of scrutiny increases, the magnitude of corrective actions would also increase. Moreover, we would expect that firms are more likely to notice and take action when there are many disciplinary actions taken against companies in an industry. The dependent variable in column 1 is Dividend t+1. As expected, the coefficient of No. of lawsuits t is negative and significant implying that firms in the industry where at least one firm faced a class action suit reduce the amount of dividends paid to shareholders. One standard deviation increase in the number of lawsuits in the industry leads to 3% decrease in the amount of dividends paid. The dependent variable in column 2 is Repurchase t+1, the logarithm of number of share repurchased by the company in year t+1. The coefficient is positive, though statistically insignificant, showing that companies try to reduce their shareholder base after the industry was hit by a disciplinary shock. Column 3 shows that companies retain more cash in years following a shock whereas increase the amount of debt (column 4). Together these results show that companies reduce the amount they payout to shareholders and increase debt in an effort to move towards stricter bonding mechanism. All the regressions include the lagged dependent variable also as a control variable. Column 5 to 8 repeat the analysis by restricting the sample to large firms. Here No. of lawsuits t becomes significant for repurchases as well. This oculd be due to the fact that it is large firms that repurchase their stock in many instances. 14

16 3.3 Effect of peer litigation on top executives incentives In this section we turn our attention to changes in executives incentive mechanism following a disciplinary action in the industry. An external disciplinary action not only affect the CEO of the firms (Banerjee et al., 2015) but also on the board of directors (Ferris et al., 2007) and other executives. To test what is the impact of class action suit on other companies in the industry, I look at total compensation, bonus and, stock and options vesting schedules of CEOs and other top executives. A Effect on CEO incentives When a company in the industry is hit by a disciplinary action, others in the industry are bound to take notice of it. Thus they would take corrective actions to fix problems in their firms or at least avoid attracting a lawsuit. One contentious area that has attracted a lot of academic and media attention is CEO compensation. It is natural for industry participant to look at peers as well when one of the companies is facing litigation. Therefore we could expect that following a disciplinary shock in an industry, compensation of CEOs would decrease. Table 7 tests the above hypothesis. The dependent variable in column 1 is Total compensation of the CEO in year t+1. All independent variables are lagged i.e. from year t. The coefficient of No of lawsuits t shows that in the years following disciplinary action, non-litigated firms in the industry reduce their CEO s total compensation by 2.9%. The corresponding number for litigated firms is at least 5 times higher. As expected coefficients of returns, size, intangibles and CEO duality are positive. Column 2 repeats the same regression with logarithm of Bonus as the dependent variable. The coefficient of No of lawsuits t is not significant here. In column 3, the dependent variable is logarithm of number of newly vested shares next period. This is a measure of corporate short-termism or managerial myopia. We would expect that following an external disciplinary action, boards and CEOs reduce their myopic behavior and focus on long-term objectives. Consistent with our expectations, the coefficient of No of lawsuits t is nega- 15

17 tive and highly significant. Economically, there is a 11.2% reducing newly vested shares following a one standard deviation increase in the number of class action lawsuits. This clearly shows that external disciplinary action taken against a peer firm threatens other firms in the industry to take corrective steps and address agency issues in their own companies. Columns 4 to 6 repeats the analysis for large firms. The results are qualitatively similar. In unreported tests, for small firms there is a reduction only in total compensation and not in bonus or newly vested shares. B Effect on top management incentives Having studied the impact of class action lawsuits on CEO incentives, the next step is to look at the incentives of other top executives in the firm. For this analysis, we look at the median incentives rather than average. The dependent variables are median total compensation, bonus and newly vested shares. Table 8 reports the results of these tests. It can be seen that following an external disciplinary action, total compensation and newly vested shares of the top executives decrease. This is in-line with the expectation that the board and top management change their myopic behavior and attempt to take corrective actions to address agency issues. 4 Conclusion External disciplinary actions such as class action law suits, SEC investigations and other litigations are important corporate governance mechanisms to punish offenders. But the true purpose of such punishments is to act as a threat against potential defectors as well. Academic literature has investigated and established the negative effects a disciplinary action has on the target company. But little is know about the threat it poses to others in the industry and how the threat makes peer firms to take corrective actions as a preemptive measure to avoid litigation. This paper attempts to fill this gap in the literature by looking at the impact class action lawsuits have on peers companies in an industry where at least one firm has faced a class action lawsuit. The results show that there is a 3.5% decrease in over-investment activity. This reduction is 10% of the impact 16

18 felt by the firm that faced the class action suit. Following a disciplinary shock in the industry, other companies reduce cash payouts to shareholders and increase leverage. Also, the compensation of CEOs and other top executives decrease by at least 3% in the years following a disciplinary action in the industry. The number of newly vested shares for CEOs goes down by 11.2% which shows that boards become less myopic and reduce their focus on short-term performance. Overall the results indicate that disciplinary action taken against a firm can be threat enough for peers in the industry to address their agency issues. 17

19 References Arena, M. and B. Julio (2015). The effects of securities class action litigation on corporate liquidity and investment policy. Journal of Financial and Quantitative Analysis 50, forthcoming. Armstrong, C. S., D. F. Larcker, G. Ormazabal, and D. J. Taylor (2013). The relation between equity incentives and misreporting: The role of risk-taking incentives. Journal of Financial Economics 109 (2), Banerjee, S., M. Humphery-jenner, and M. Tham (2015). Executive overconfidence and securities class actions. Becker, G. S. (1968). Crime and Punishment : An Economic Approach. Journal of Political Economy 76 (2), Bizjak, J., M. Lemmon, and T. Nguyen (2011). Are all CEOs above average? An empirical analysis of compensation peer groups and pay design. Journal of Financial Economics 100 (3), Bourveau, T. and S. M. Spira (2014). Do M&A Lawsuits Discipline Managers Investment Behavior? Brochet, F. and S. Srinivasan (2014). Accountability of independent directors: Evidence from firms subject to securities litigation. Journal of Financial Economics 111 (2), Edmans, A., V. W. Fang, and K. Lewellen (2015). Equity Vesting and Managerial Myopia. Ferris, S. P., T. Jandik, R. M. Lawless, and A. Makhija (2007). Derivative Lawsuits as a Corporate Governance Mechanism : Empirical on Board Changes Surrounding Filings. Journal of Financial and Quantitative Analysis 42 (1), Fich, E. M. and A. Shivdasani (2006). Are Busy Boards Effective Monitors? Journal of Finance 61 (2), Fich, E. M. and A. Shivdasani (2007). Financial fraud, director reputation, and shareholder wealth. Journal of Financial Economics 86 (2), Gande, A. and C. M. Lewis (2009). Shareholder-Initiated Class Action Lawsuits: Shareholder Wealth Effects and Industry Spillovers. Journal of Financial and Quantitative Analysis 44 (04), 823. Gopalan, R., T. Milbourn, F. Song, and A. V. Thakor (2014). Duration of Executive Compensation. Journal of Finance 69 (6), Humphery-Jenner, M. L. (2012). Internal and external discipline following securities class actions. Journal of Financial Intermediation 21 (1), Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. Karpoff, J. M., D. S. Lee, and G. S. Martin (2008). The Cost to Firms of Cooking the Books. Journal of Financial and Quantitative Analysis 43 (03),

20 Karpoff, J. M., D. Scott Lee, and G. S. Martin (2008). The consequences to managers for financial misrepresentation. Journal of Financial Economics 88 (2), Kim, I. and D. J. Skinner (2012). Measuring securities litigation risk. Journal of Accounting and Economics 53 (1-2), McTier, B. C. and J. K. Wald (2011). The causes and consequences of securities class action litigation. Journal of Corporate Finance 17 (3), Morse, A., V. Nanda, and A. Seru (2011). Are Incentive Contracts Rigged by Powerful CEOs? Journal of Finance LXVI (5), Peng, L. and A. Röell (2008). Executive pay and shareholder litigation. Review of Finance 12 (1), Servaes, H. and A. Tamayo (2014). How Do Industry Peers Respond to Control Threats? Management Science 60 (March), Shleifer, A. and R. W. Vishny (1997). American finance association. Journal of Finance 52 (2), Strahan, P. E. (1998). Problems. Securities Class Actio, Corporate Governance and Managerial Agency 19

21 Table 1 Distribution of class action lawsuits in the sample. Panel A: Sector-wise distribution Sector All Manufacturing 940 Information & Technology 511 Finance & Insurance 463 Services 258 Retail & Wholesale Trade 137 Mining & Construction 91 Utilities 61 Health care 55 Real Estate 51 Transportation 18 Others 30 Total 2615 No. of Suits Panel B: Year-wise distribution Year No. of suits 20

22 Table 2 Panel Summary. This table summarizes the panel data used for this study. Panels A, B and C use Stanford Securities Class Actions Clearinghouse and Compustat merged data. Panel D and E report summary of variable by merging data with Execucomp and ISS-Riskmetrics. The sample period covered is 1996 to Variable Mean Stdev. p25 Median p75 N Panel A: Litigation Litigation in industry No. of lawsuits Panel B: Firm Actions Over-investment Dividend Repurchases Cash Retained Panel C: Firm Characteristics Leverage Intangibles Stock Return Market Value Growth Tobin s Q Institutional Holding % Panel D: CEO & Top Executive Characteristics CEO Age CEO Tenure CEO Shares Owned % CEO-Chair CEO-Chair-President CEO Total Compensation CEO Bonus CEO Vested Shares Median Top Execs. Compensation Median Top Execs. Bonus Median Top Execs. Vested Shares Panel E: Board Characteristics Board Size Board Experience Board Directorships % Independent Directors % Independent Busy Directors

23 Table 3 Correlation between key dependent and independent variables. (1) No. law suits Over-investment Dividend Repurchases Cash Retained Leverage CEO-Chair CEO Vested Shares No. law suits 1 Over-investment Dividend Repurchases Cash Retained Leverage CEO-Chair CEO Vested Shares p <.1, p <.05, p <.01 22

24 Table 4 Probability of a company being sued. The following table reports logit regressions where the dependent variable is Lawsuit i,t+1, a binary variable which take a value 1 if the company faced a class action lawsuit in year t+1. All independent variables are from year t. The sample period for these regressions is 1996 to Standard errors are reported in parentheses. They are clustered at industry and year level. *, ** and *** denote significance at 10%, 5% and 1% level respectively. (1) (2) (3) (4) (5) (6) Corrective Actions Overinvestment t (0.052) (0.028) (0.037) (0.055) (0.037) (0.035) Dividends t (1.560) (3.509) (2.543) (4.130) (2.755) (5.287) Repurchases t (0.031) (0.022) (0.024) (0.029) (0.019) (0.139) Cash Retained t (0.017) (0.021) (0.196) (0.264) (0.167) (0.034) Leverage t (0.200) (0.165) (0.016) (0.059) (0.109) (0.718) Firm Characteristics Returns t (0.069) (0.119) (0.143) (0.079) (0.088) Intangibles t (0.018) (0.031) (0.029) (0.019) (0.033) Ln(Market V alue) t (0.022) (0.046) (0.069) (0.039) (0.099) Growth t (0.047) (0.196) (0.314) (0.111) (0.058) CEO Characteristics CEOAge t (0.008) (0.009) CEO T enure t (0.014) (0.014) CEO V ested Shares t (0.021) (0.032) % CEO Shares (0.713) (1.230) CEO Chair (0.231) (0.295) Board Characteristics Board Size t (0.019) % Independen Dirs t (0.295) Observations Pseudo R Industry F.E Yes Yes Yes Yes Yes Yes Year F.E Yes Yes Yes Yes Yes Yes Sample Full Full Full Full Large firms Small firms 23

25 Table 5 Effect of industry disciplinary shock on over-investment. The following table reports result for tests on reduction of Over-investment following a disciplinary shock in the industry. The dependent variable is logarithm of over-investment of a company in excess of industry mean. The sample period is from 1996 to Fixed effects for industry and years are included in all models. Standard errors are reported in parentheses. They are heteroskedasticity robust and clustered at industry and year level. *, ** and *** denote significance at 10%, 5% and 1% respectively. (1) (2) (3) (4) (5) (6) (7) (8) Lawsuit t (0.013) (0.017) (0.021) (0.163) No. of lawsuits t (0.010) (0.013) (0.010) (0.060) Returns t (0.005) (0.018) (0.017) (0.018) (0.013) (0.014) (0.057) (0.066) Ln(Market V alue) t (0.003) (0.003) (0.006) (0.006) (0.006) (0.006) (0.055) (0.056) T obin s Q t (0.000) (0.000) (0.003) (0.003) (0.002) (0.002) (0.029) (0.029) Growth t (0.007) (0.011) (0.026) (0.023) (0.030) (0.026) (0.082) (0.079) Leverage t (0.000) (0.000) (0.007) (0.006) (0.015) (0.015) (0.138) (0.096) Intangibles t (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.016) (0.014) Board Size t (0.001) (0.001) (0.002) (0.002) (0.014) (0.015) % Independen Dirs t (0.020) (0.022) (0.024) (0.027) (0.122) (0.130) CEOAge t (0.001) (0.001) (0.001) (0.001) (0.002) (0.002) CEO T enure t (0.001) (0.001) (0.001) (0.001) (0.010) (0.011) CEO V ested Shares t (0.004) (0.004) (0.005) (0.006) (0.019) (0.022) % CEO Shares (0.064) (0.093) (0.175) (0.240) (0.356) (0.358) CEO Chair (0.011) (0.013) (0.015) (0.019) (0.050) (0.045) Observations Adjusted R Industry F.E Yes Yes Yes Yes Yes Yes Yes Yes Year F.E Yes Yes Yes Yes Yes Yes Yes Yes Sample Full Full Full Full Large firms Large firms Small firms Small firms 24