• No results found

SEC fraud punishment and stock performance : study on the influence of sanctions imposed by the Security and Exchange Commission as a consequence of fraud on the stock performance of these companies

N/A
N/A
Protected

Academic year: 2021

Share "SEC fraud punishment and stock performance : study on the influence of sanctions imposed by the Security and Exchange Commission as a consequence of fraud on the stock performance of these companies"

Copied!
41
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

SEC fraud punishment and stock performance

Study on the influence of sanctions imposed by the Security &

Exchange Commission as a consequence of fraud on the stock

performance of these companies

Economics & Finance

Felix Snoeck Henkemans Studentnumber: 10385320

Supervisor: J. Lemmen

Bachelor Thesis – University of Amsterdam 29-06-2015

(2)

This document is written by Student Felix Snoeck Henkemans who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract: This study examines the effect of fraud punishment by the Security and Exchange Commission on stock pricing. The immediate stock-price effects and the pre- and post-announcement drifts are examined. We use 3 different samples: the total sample, a warning sample and a fine sample. To capture the effects of the mortgage crisis recent data is used. Moreover, we examine the effects for the market-model described by MacKinlay. We find a statistically significant positive immediate price reaction for the warning sample. For the other samples results are less significant. We conclude that the type of fraud punishment matters and that the outcome of the fraud punishment is not known to the market.

JEL-classifications: G2, G3, G18, G38

(3)

Table of contents

1. Introduction 1

2. Prior Research 2

2.1 Fraud, the concept 2

2.2 Evidence from prior research 2

2.3 Efficient Market Hypothesis 4

2.4 Recent research 5

2.5 Impact of the financial crisis 6

2.6 Explaining the negative reactions on stock performance 7

2.7 Hypotheses 7

3. Experimental design 10

3.1 Data 10

3.2 Method 11

3.2.1 Measuring & analysing abnormal returns 13

3.2.2 Estimation of the market model 14

3.2.3 Cumulative abnormal returns & statistical properties of

abnormal returns 16

3.2.4 Aggregation of returns through time and across securities 18

3.2.5 Testing the null hypothesis and power of the tests 19

4. Results and analysis of results 20

5. Conclusion and discussion 24

5.1 Conclusion 24

5.2 Discussion 24

6. References 26

(4)

1. Introduction

After major fraud scandals involving companies like Enron, Worldcom, Royal Ahold and Parmalat there is an on-going debate over how to create good control and regulation. This is obviously needed to prevent more fraud scandals. Why do companies commit fraud? Johnson et al. (2009, p. 115-116) mention the economic theory of crime framework, motivated by Becker’s (1968) paper. This theory states that the expected utility of the payoff exceeds the expected disutility of getting caught and punished. Therefore, agents have an incentive to commit fraud.

First, the motivation for this thesis is to give an insight in the sanctions that the Security & Exchange Commission (SEC hereafter) imposes when fraud is noticed and their consequences on stock performance. By looking at the stock returns the damage of a fraud penalty for shareholders can be analysed. The sanctions can be financial (fines) or non-financial

(warnings). If there really is a negative effect on stock returns, the sanctions could be helpful because they take away incentives to commit fraud. If the penalty for committing fraud is substantial, managers will be reluctant to commit fraud.

Second, this study tries to address the question if there are any changes visible since the beginning of this century. Has punishment become heavier during recent years? After the crisis starting in 2007 where countless scandals occurred, there might be more heavy

punishment during recent years. Does the SEC actually use heavier penalties? Are effects of a fraud punishment on stock performance larger than before the scandals in the beginning of this century due to loss of confidence by shareholders? The general research question reads as follows:

What is the influence of sanctions imposed by the Security & Exchange Commission as a consequence of fraud on the stock performance of these companies?

We proceed as follows. Chapter 2 provides background on fraud actions, their sections and prior research in this area. Chapter 3 describes the research method and data. Chapter 4 contains the empirical findings. Finally, chapter 5 contains a conclusion and summary discussion of the empirical findings.

(5)

2. Prior research

This research examines the influence of fines on the stock return. Before a well-built hypothesis can be formulated, one must built a theoretical framework consisting of prior research. In the past a countless number of event studies have taken place. Some of these studies contain a very useful basis for this research. This study aims to address the question if fines as fraud punishment have an effect on the performance of stock.

2.1 Fraud, the concept

Fraud is a very broad concept. According to Apostolou et al. (2000, p. 181), corporate financial fraud refers to “intentional misrepresentation of amounts or disclosures in the financial statements.

Sadighi Firozabadi et al. (1998) use the broader concept earlier used by Simmons in his 1995 paper. Sadighi et al. state that fraud occurs when all the following elements exist:

’’ an individual or an organisation intentionally makes an untrue representation about an important fact or event;

the untrue representation is believed by the victim (the person or organisation to whom the representation has been made);

the victim relies upon and acts upon the untrue representation;

the victim suffers loss of money and/or property as a result of relying upon and acting upon the untrue representation (Sadighi Firozabadi et al. 1998, p.4-5).’’

In this study the punishment for fraud in the broader definition above is used. All types of fraud that the SEC punishes are included.

2.2 Evidence from prior research

A second question that is addressed in this study is if there is a movement towards more heavy punishment and the effect of this heavier punishment on stock performance. The punishment of companies for fraud is not a process started during recent years. Karpoff et al. (1999) examine defence fraud by looking at the wrist slap hypothesis. This hypothesis states that the average penalty for committing defense fraud is extremely low. Their findings however do not support this hypothesis. ’’[…] Public announcements of (alleged) fraud are associated with large and statistically significant declines in stock price […].’’ (Karpoff et al., p.823). Karpoff & Lott (1993) examine the effect of fraud punishment on reputation and stock price. For their data sample they find a mean change of the market value equal to- $6.418

(6)

million. This loss in market value indicates a significant drop in stock performance due to a lower stock price. In their study they use an event window of (0,1) and (0,2) and an estimation window of (-131,-31).

Fich & Shivdasani (2007, p.315) analyze cumulative abnormal returns (CARs hereafter) based on a time window of 100 days prior to the announcement of a penalty until the

announcement. They find: ’’the CARs are close to zero until approximately 20 trading days, or one calendar month, prior to the filing and then become sharply negative, averaging -16% during the (-20, -3) window relative to the lawsuit filing date. This phenomenon is called the pre-announcement drift. Over the (1,0) interval, the abnormal returns average -5.95% (Fich &Shivdasani, 2007 p.316).’’ This estimate is comparable to Karpoff and Lott (1993), who find two-day excess returns of -4.56% for a sample between 1978 and 1987, as well as Karpoff, Lee, and Martin (2005), who show a market reaction of -7.00% in their 1978-2002 sample. As can be seen from the prior research mentioned above, the filing of a financial fraud class action suit is associated with a significant negative revaluation of the firm’s equity.

This research suggests that there is a negative effect of fraud penalties on stock performance for both market value and the cumulated abnormal return.

There are more studies that confirm the above findings. Bhagat et al. (p.15) use the 150-day period from day -171 to day -22 as the estimation window. They find a decline in shareholder wealth of 1.73% with a CAR of -0.93%. Other studies find even more obvious effects. An average two-day excess return of -4.2% for a sample of 159 earnings restatements from 2000 to 2001 is documented by Agrawal and Chadha (2005). This study also examines the medium term effect and find the medium-term effects significantly larger (CAAR -10% &-17% for event window (-30,0) and (60,90) respectively). The CAAR stated in their study stands for the cumulative average abnormal return (CAAR hereafter).

Ferris & Pritchard (2001) examine stock price reactions to securities fraud class actions. They use papers by Griffin, Robert Kellogg, Francis et al., Niehaus & Roth and Roberta Romano to construct their theoretical framework. Robert Kellogg finds negative stock price reactions around the month of discovery but also in the months preceding the discovery. A possible explanation for this by Kellogg is that some information leaked in an earlier stage about financial misrepresentations. Griffin et al. find that there are also long-term effects, effects outside the normally used (-1,1) event window. ‘’Griffin finds a long-term negative drift for

(7)

stock prices following the announcement of the misrepresentation that forms the basis for the lawsuit (Ferris & Pritchard, p.6).’’

Comparing all studies mentioned above a similarity occurs in the form of multiple event windows. At least one event window is centred on the day of the announcement. Furthermore, multiple event windows are chosen to capture the pre- and post-announcement drift. The choice for this event windows comes from the Efficient Market Hypothesis.

2.3 Efficient Market Hypothesis

Fama (1969) states there are 3 forms of the efficient market hypothesis (EMH). These 3 forms will be discussed briefly. First, Fama discusses the weak form hypothesis, in which only historical prices influence current stock prices. Second, he discusses the semi-strong hypothesis, in which not only historical prices but also obviously publicly available information influences prices.

Finally the strong market hypothesis is discussed. In this last hypothesis it is assumed that all available information, publicly and private, is reflected immediately in the stock prices. According to this hypothesis, the stock price effect will occur immediately after the announcement at date 0.

The EMH is of importance when choosing an event window. This study assumes the semi-strong form of the EMH. Why is the semi-semi-strong form assumed instead of the semi-strong form? The reason is that the strong form would imply a reflection of all private information in stock prices. Undoubtedly, the managers of a fraud firm have got information about the fraud and this would cause an immediate stock price reaction when fraud starts and not during the event window when the SEC imposes the fraud penalty. When the SEC notices fraud and as a result releases an enforcement action, we can determine the event window. However, with the strong-hypothesis this is not possible, because it is impossible to gather private information and notice when managers start committing fraud.

Except for Bhagat et al. (1998) all prior research uses multiple event windows to examine the effects of fraud on stock performance. The EMH implies that a small event window centred on date 0 is the best window due to immediate price reflection of news. This raises the question why prior research uses multiple event windows.

(8)

One reason is to look at the mid-term effects of stock performance. However, the most important reason multiple researchers mention is that they expect that prices adjust very quickly, but not immediately. Griffin et al. (2000) conclude bad news is not fully incorporated by the market at the announcement. This implies a small imperfection of the EMH.

Studies from Marciukaityte (2006), Agrawal & Chadha (2005) and Griffin et al. (2000) examine the so-called post-announcement drift by setting their event windows larger than a few days following the announcement. Griffin et al. (2000) state that the post-announcement drift persists for at least three weeks. Chapter 3 data & method will discuss this issue further and takes into account the importance of the post-announcement drift when choosing proper event windows.

2.4 Recent research

The financial crisis of 2007 has been of enormous influence on the world and on the investor climate in particular. After all the scandals were revealed before and during the financial crisis the need for better regulation and governance has become enormous.

Agrawal & Chadha (2005) describe the four major changes that have taken place since scandals like Enron and Worldcom, ’’First, three of the Big 4 audit firms have either divested or publicly announced plans to divest their consulting businesses.’’(Agrawal & Chadha, p.372). Second, a former member of the Big 5 audit firms, Arthur Andersen, has gone out of business. Third, the Sarbanes-Oxley Bill of 2002 was adopted. In this law multiple corporate governance rules are written down that apply for all public companies with stock traded in the United States. Finally, in the end of 2003 the New York Stock Exchange, the NASDAQ and the American Stock Exchange adopted additional corporate governance rules. These rules apply to the major part of companies with stock listed on these 3 markets.

With this new method of government governance one would expect the amount of fraud cases to be substantially lower. As can be seen in table 1 the amount of enforcement actions has not lowered during recent years, this is a interesting finding taken into account that more heavy punishment should make firms reluctant to commit fraud.

(9)

2.5 Impact of the financial crisis

The subprime mortgage crisis of 2007 has had an enormous influence on the financial system. Two aspects of the financial crisis in particular are of interest for this study. First, the

confidence under investors. Schwarz (2008) states that the crisis ultimately caused a general confidence loss on the financial markets. The loss of confidence since the financial crisis could influence the results of this study because investors will now be reluctant to invest in a fraud firm. This could cause a significant drop in value of the firm and a following drop in the CARs.

Second, the influence of government governance on the financial crisis. Remarkably, while prior studies like Kirkpatrick (2009) advise changes in government structure, no important changes have taken place compared to the changes in the beginning of this century.

Concluding, the aspect of confidence since the financial crisis is of interest for this study.

Table 1 SEC Enforcement Actions

Enforcement Actions by Fiscal Year 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Broker-Dealer 94 75 89 67 109 70 112 134 121 166 Delinquent Filings n/a 91 52 113 92 106 121 127 132 107

FCPA n/a n/a n/a n/a n/a n/a * 20 15 5 7

Insider Trading 50 46 47 61 37 53 57 58 44 52

Investment Adviser/Investment Co. 97 87 79 87 76 113 146 147 140 130 Issuer Reporting and Disclosure 185 138 219 154 143 126 ** 89 79 68 99 Market Manipulation 46 27 36 53 39 34 35 46 50 63 Securities Offering 60 61 68 115 141 144 124 89 103 81

Other 98 49 65 21 27 35 31 39 13 50

Total Enforcement Actions 630 574 655 671 664 681 735 734 676 755

* Prior to FY 2011, FCPA, the Foreign Corrupt Practices Act, was not a distinct category and FCPA actions were classified as Issuer Reporting and Disclosure. The FCPA essentially takes action against bribery. ** Prior to FY 2011, this category included FCPA actions, which are now tracked as a distinct category. Source: https://www.sec.gov/news/newsroom/images/enfstats.pdf

Johnson et al. (2009) use a sample with relatively recent data (1991-2005). After imposing multiple restrictions on their sample data they find 87 cases of fraud. In this more recent sample they find similar conclusions as Karpoff et al, (1993), Agrawal & Chadha (2005) and Eliever & Shivdasani (2007). ’’We find a mean cumulative raw return from one day before

(10)

the announcement to one day after of −14.9%, which is significantly different from zero at the 0.01 level (Johnson et al., p126).’’ Moreover, they find that a majority of the companies, 80%, experiences negative stock price reaction when potential fraud information is revealed. In their study Johnson et al. (2009) argue that news about fraud events does not come out on a single day. Therefore they use a more broad event window of −5 to +30 to compute the CARs. The 1% level of significance is used, and the average CAR is −22.7%.

Table 2 presents all the results of prior research, the most recent research of Johnson uses data until 2005. This study aims to provide more recent results, especially on stock performance after the financial crisis.

2.6 Explaining the negative reactions on stock performance

Financial fraud typically implies a substantial drop in stock performance. To interpret the results provided by our data set in section 3 it is helpful to understand the effect of a fraud sanction on the performance of stock. Direct penalties only account for a small portion of the value loss and the major part results from damaged reputation of firms (Karpoff et al., 2005). This raises the question why a loss in reputation causes a decline in stock performance. A firm that has been charged for fraud experiences substantial reputational costs. These costs arise due to a higher cost of capital, loss of credibility with respect to customers and contracting of suppliers (Jarrell & Peltzman, 1985).

The expected penalty for committing fraud should be equal to the total social costs of the fraud. In their 1993 paper, Karpoff and Lott state that the mean ratio for dollar sanction to loss on government procurement fraud is .29. This means that for every dollar of social costs caused by fraud, the penalty for the company is only 29 cents. The U.S. Sentencing

commission raised corporate fraud penalties in 1991 by over twenty-fold in response to the perceived small punishment of fraud committing firms. Concluding, not only fraud penalties tend to decrease stock performance, actually the reputational loss seems to have a major impact on stock performance.

2.7 Hypotheses

This study studies two hypotheses to understand the effect of fraud penalties on stock performance. The reputation hypothesis holds that the announcement of a fraud penalty will result in a loss of reputation. This loss of reputation leads to a lower credibility and higher

(11)

costs. These factors will ultimately have a negative effect on stock performance.

The second hypothesis is the government governance hypothesis. This hypothesis holds that governance has changed since the beginning of this century. Heavier fraud punishment in recent years results in greater loss in value for the company and following this value loss a decrease in CAR. Taken into account the recent financial crisis and the loss of confidence under investors, a loss in reputation will result in a large decrease of CAR. Concluding, this hypothesis holds that the decrease in CAR for the firms in our recent sample will be greater than the decrease in CAR found by earlier studies.

Table 2 Summary of literature and their results

Author (s) Country Time

period

Estimation window**

Results () & Event window []

Karpoff & Lott (1993) US 1978-1987 (-131, -31) Market value (-6,418 mln): [0,1], [0,2] Brown & Warner (1983) US - * (-244,-6) -*, [-5,5] MacKinlay (1997) US -* (-270,-20) -*, [-20,20] Fich & Shivdasani (2007) US 1998-2002 (-100,0) CAR (-16%) :[-20,3] (-5,95%): [0,1] Bhagat et al. (1998) 1981-1983 (-171,-22) CAR (-0.97%):[-1,0] Agrawal & Chadha (2005) US 2000-2001 *** CAR (-5,6%): [-1,1] (-4,2%): [0,1] (-17%):[-30,0] (-10%): [60,90] Johnson et al. (2009) US 1991-2005 (-3,-1) years CAR (-14,9):[-1,1] (-22,7%): [-5,30] Griffin et al. (2000, working US 1995-1997 (-64,0) CAR (0,5% ; -1,8%): [-1],[0], [1], [1,2], [2,15], [2,50]

(12)

paper) Kellogg (1984) US 1967-1979 (-220,-2) (-31,7%): [-20,10] **** Marciukaityte et al. (2006) US 1978-2001 (-90,-30) CAR (-5,01%):[1,0] Long term: 2,95%): [1yr] 2,35%): [2yr] 5,25%): [3yr] 15,30%): [4yr] (-20,82%): [5yr]

*These studies examine how to perform an event study, they do not examine stock

performance and are included in the table to show the window selection of event studies in general.

** Estimation window in days unless reported otherwise.

*** Uses the value-weighted CRSP index returns in the event window for the estimation of the normal returns.

(13)

3. Experimental Design

3.1 Data

To collect a data sample of fraud firms, we search for the enforcement releases of the SEC. The releases are summarized under the Accounting and Auditing Enforcement Releases (AAER). We find 204 fraud cases for the years 2007-2012.

Next, we select the set of Standard and Poor’s 1500 in the Wharton database. All enforcement releases that do not concern firms in the S&P 1500 are excluded from the sample. For the estimation window, we need data in the (-270, -20) window. The literature discussed in chapter 2 does not reach a consensus over the estimation window. Therefore we follow the suggestion of MacKinlay. For the event window we need data of these firms until at least 20 days before and after the announcement of the SEC. All firms that do not have these data are excluded from the sample.

The total usable sample contains 111 fraud cases. This sample contains 50 cases where the firm only received a warning or received a warning in combination with a fine for the

manager. In the remaining 61 cases the firm received a fine or had to pay prejudgment interest and a payback of illegal-gotten gains. Prejudgment interest is an increase of the amount of fines because of the time that elapsed between the fraud and the announcement of the penalty. In fact this in the opposite of an opportunity cost compensation: the fraud firm has fraud gains from the moment of the fraud and has to pay these back when their penalty is announced.

We perform multiple regressions, over the total sample and also separately over the two samples of 50 and 61 cases respectively. The motivation behind the separation is that a fine for the manager may hurt only the manager and not the firm so we consider this as

punishment that does not directly hurt the company. The same motivation holds for warning-only cases in this category: the effect on stock price could be different than when an actual fine is imposed.

Some studies use the Wall Street Journal as the source to select fraud firms. Johnson et al. mention an important feature of the AAERs. ‘’AAERs represent cases where the SEC believes that there is sufficient evidence of accounting or auditing problems to bring a case against a firm or its executives (Johnson et al. p. 119).’’ In some cases the Wall Street Journal

(14)

accusation appears to be wrong and the firms did not commit fraud. Therefore, this study focuses on fraud firms as mentioned by the AAERs. For the selected firms in our sample we take the daily stock data including dividend. The dividends are included because this study examines the total stock return, not only capital gains.

3.2 Method

This study examines the influence of fines and warnings on the stock return. Section I contains the basis for this study. The economic theory that is necessary for this study is written down in the previous section. This research will use the event study method described by MacKinlay (1997). In his article MacKinlay uses a step-wise explanation on how to conduct an event study in Economics & Finance.

MacKinlay states that there is no unique structure but there is a general flow of analysis. In this study this general flow of analysis will be described and followed.

First we specify the time period that will be used. During this research daily stock return data are needed because differences in stock prices will generally occur immediately after the announcement of the fraud penalty. The pre- and post-announcement drift have to be studied carefully too. Daily data are the most precise data and will therefore be used in this study. Because this thesis does not only cover the effect of the penalty but also examines the trend of heavier punishment during recent years we have to take multiple years into account.

Therefore this study examines fraud sanctions during the years 2007-2012.

After we have specified the time period we have to set multiple criteria to specify our sample. One criterion is that the company is listed on a major stock exchange, and as mentioned before this exchange is the Standard & Poor’s 1500.

Of course, another criterion is that a company has been penalized for fraud. Furthermore, financial institutions and utilities are excluded because penalties for financial institutions may be very different than penalties for other companies. The major institutions that penalize firms for fraud like the AFM in the Netherlands were reluctant to give any information on penalties for financial firms. Their argument to conceal all information about fines for financial

institutions is to prevent panic.

After the determination of the criteria it is necessary to set up a model, which can be tested using a statistical program. MacKinlay (1997) states that to measure the event’s impact, a

(15)

measure of the abnormal return is needed. ‘’The abnormal return is the actual ex post return of the security over the event window minus the normal return of the firm over the event window. The normal return is defined as the expected return without conditioning on the event taking place.’’(MacKinlay, p. 15) Putting this in a formula gives for firm i and event date the abnormal return formula of:

Where is the abnormal return, is the actual return and is the normal return for time period . is the conditioning information for the normal return model. To calculate normal returns the daily stock return data will be taken over the (-250, -20) window prior to the fraud announcement.

For the expected normal return a market model to estimate the normal return is needed. There are two models that can be used to calculate the normal return: the constant mean return model and the Capital Asset Pricing Model.

The model used in this study is the Capital Asset Pricing Model (CAPM). The CAPM is stated in the following formula:

E(R) = Expected return = Risk-free return = Beta-coëfficient

= Expected market return

For the risk-free return we take the return on US treasury bonds. US treasury bonds are considered risk-free; they currently have a return level almost equal to zero.

To calculate the expected market return: we need the expected return of the market portfolio, the S&P 1500, and the risk-free rate.

(16)

1. Investors can buy and sell all securities at competitive market prices (without incurring taxes or transactions)

2. Investors hold only efficient portfolios of traded securities- portfolios that yield the maximum expected return for a given level of volatility.

3. Investors have homogenous expectations regarding the volatilities, correlations, and expected returns of securities. (Berk & DeMarzo, 2011 p.357-358)

MacKinley (1999) mentions that although the constant mean return model is perhaps the simplest model, Brown and Warner (1985) find it often yields results similar to those of more sophisticated models. The choice for the CAPM in this study comes from the fact that the CAPM is a model constructed on the basis of economic theory. Any academic research always prefers using a model, which has economic theory underlying it.

With the CAPM formula above we capture the second part of the abnormal return formula:

To calculate the first part of the formula: the actual return data have to be analysed. The actual return of a stock is captured in the following formula:

Total Stock Return =

P0= Initital Stock Price

P1= Ending Stock Price (Period 1) D= Dividends

The initial stock price is the stock price prior to the announcement of the penalty. The ending stock price is the price at the end of the event window.

With the CAPM and the formula for the actual return the abnormal returns are calculated.

3.2.1 Measuring & analysing abnormal returns

To facilitate the analysis of abnormal returns it is feasible to use different notations for the time windows that this study uses. MacKinlay (1997) defines the event date as =0 , =T1+1

(17)

to =T2 as the event window and finally =T0+1 to =T1 as the estimation window. Furthermore L1= T1-T0 and L2=T2-T1 represent the length of the estimation window and event window respectively. This research examines fines and warnings. These penalties are announced at a certain date. Why should an event window be used and not just the date of the announcement? ’’Even if the event being considered is an announcement on given date it is typical to set the event window length to be larger than one. This facilitates the use of abnormal returns around the event day in the analysis (MacKinlay, p. 19).’’

3.2.2 Estimation of the market model

The CAPM is used as the market model in this study. Daily stock data are analysed within this model. Brown & Warner (1985) mention that the combination of daily data and a cross-sectional approach can bring difficulties. One of the most important issues is the non-normality of the data. They state that there is some evidence that the cross-sectional daily return distribution converges to a normal distribution. However: ’’A chief concern here is whether and for what sample size this result applies to the excess returns, even though the assumptions of the Central Limit Theorem (at least in its standard version) are violated with these data (Brown & Warner, p.5).’’ They find that the mean excess return is close to normal for sample sizes of 50. Figure 2 plots their findings.

(18)

Figure 2 Distribution of mean excess returns for sample size of 50

Ordinary least squares (OLS) is the best estimation method for the market model under general assumptions. Moreover, MacKinlay (1997, p.17) imposes another assumption: the asset returns are jointly multivariate normal and independently and identically distributed through time. Under this assumption the OLS estimator is efficient.

The OLS estimators of the market parameters for an estimation window of daily observations are:

(1)

(2)

(3) where

and

(19)

The return for security in event period is and the return for the market in this period is .

In the next section we examine the statistical properties of abnormal returns. Furthermore, we explain the use of cumulative abnormal returns. These CARs are also used in prior research.

3.2.3 Cumulative abnormal returns & statistical properties of abnormal returns

Now that the market model parameters are estimated, the abnormal returns can be calculated. First, Mackinlay defines the abnormal returns as:’’[…] the disturbance term of the market model calculated on an out of sample basis.’’ (MacKinlay, p. 20)

The formula for the estimator of abnormal returns is:

(4)

In this formula the represents the abnormal return sample for firm i in the event window for the time window , which represents T1 + 1, . . . , T2. In the section ’’estimation of the market model’’ the use and efficiency of OLS is explained. This implies that the abnormal returns are jointly normally distributed with a conditional mean equal to zero, the adequate conditional variance is:

(5)

The second part of the variance is due to sampling errors in the parameters and . If the estimation window L1 becomes large, the parameter sampling error approaches zero. The disturbance variance is then the variance of the abnormal returns. This shows the importance of a large estimation window, this study uses a 230 days estimation window (-250, -20). The prior literature does not reach consensus on the size of the estimation window. We will therefore follow the 230 days event window used by MacKinlay.

The observations of the abnormal return are now independent through time. So under the hypothesis H0 that a penalty for fraud has no influence on the return behaviour, the distribution of the abnormal return in the event window is:

(20)

(6)

With formulas (5) and (6), the statistical properties of the abnormal returns are known. This study aims at drawing an overall conclusion on abnormal returns so we need a measure that aggregates the abnormal returns across all securities and also through time. This is where the cumulative abnormal returns are needed.

First, the formula for the CARs through time is as follows:

(7)

Formula (7) gives the CAR aggregation from . From figure 1 in section ‘’Measuring & analysing abnormal returns’’ T1+1 to T2 represent the event window. For the CAR the following holds: T1 T2.

Important here is the statistical property of the variance, namely as L1 increases:

(8)

However, this formula has to be adjusted for the parameter estimation errors derived from formula (5) when L1 has a small value. So formula (8) shows the importance of a large estimation window again.

The cumulative abnormal return distribution under the null hypothesis is:

(9)

Now that the aggregation of the returns through time is known, we need the formula for aggregation across securities because one cannot draw an overall inference based on the data of one security.

When aggregating across securities one problem arises: the clustering effect. Clustering arises when the included securities overlap in their event window. To overcome this problem the Central Limit Theorem (CLT) is used. Keller (2012, p.306) states: ‘‘the sampling distribution of the mean of a random sample drawn from any population is approximately normal for a sufficiently large sample size. The larger the sample size, the more closely the sampling

(21)

distribution of will resemble a normal distribution.’’ An important condition for the CLT is a large sample size. This study uses a sample size larger than 100 data points. With this data amount the CLT can be used.

3.2.4 Aggregation of returns through time and across securities

The aggregation of abnormal returns for time period follows from the in formula (4). Furthermore we take N as the number of events. Now the sample abnormal return aggregation formula is:

(10)

With the corresponding variance:

(11)

Again, this variance only holds for large values of L1 (the estimation window).

Using formula (10), the average abnormal returns formula, one can calculate the cumulative abnormal return through time and across all securities.

This formula uses the same derivation as formula (7). The difference is the use of the aggregated abnormal returns .

(12)

This holds for all intervals in the event window. The formula for the variation of the CARs follows from formulas (11) and (12). It is essentially a summation of the variations of .

(13)

Once again, the clustering effect is of importance. The assumption of no clustering effects is needed to set the covariance terms of the variance estimator equal to zero. This implies the following formula for tests concerning the null hypothesis (abnormal returns equal to zero):

(22)

(14) Formulas (1) to (14) contain all information needed to test our hypotheses mentioned in section I.

Now we arrive at the last important note of the abnormal return calculation. The variation of the error term is unknown and therefore an estimator is needed before we can calculate the variance of formula (11). The appropriate estimator according to Mackinlay (p. 24) is the commonly used sample variance measure of obtained from the market model regression.

3.2.5 Testing the null hypothesis and power of the tests

Using all information in the previous sections the null hypothesis can be tested using:

However, the outcome of these tests is not a sufficient answer for the research question. The power of the tests has to be analysed. It is desirable to look at the effects of fraud penalties on return behaviour on different significance levels.

To perform a statistical test on the null hypothesis the critical region has to be defined. MacKinlay (1997) defines this critical region in the following way:

or

where c(x) represents the normal cumulative distribution function. In this region the null hypothesis will be rejected.

(23)

4. Results & analysis of results

Total sample empirical findings: In table 3 the results of our regression on the total sample

containing 111 securities are presented. First of all, we observe a positive CAAR at the 5% level in the (0, +1). The (0, +1) positive CAAR is significant at the 5% level. The MacKinlay market-model return is 0.69% respectively. Second, we analyse the results in the other event windows. For the (-1,0) window there is no significant positive CAAR for the market-model returns.

The (-20, -1) window that represents the pre-announcement drift shows no significant positive CAAR. The same holds for the CAAR of the post-announcement drift: the -0.50% CAAR in the (0, +20) window is statistically insignificant. So, we do observe a positive stock price reaction for the (0, +1) window. This is the immediate stock price reaction. We do not observe, however, a pre- or post-announcement drift.

In appendix 3 the daily average abnormal returns (AARs hereafter) are calculated for the (-20, +20) window. We observe both positive and negative AARs and therefore we conclude there is no pre- or post-announcement drift. We do observe statistically significant positive and negative daily AARs. However, there are both positive and negative AARs in both drifts. Therefore we ascribe these to the coincidence factor. The 0.69% positive effect suggests that a fraud sanction has a positive effect on the stock price, a result that is contrary to prior research results. Next, we separate the sample in 2 categories as mentioned in the data section to give a more clear explanation on the stock price effects.

Table 3 CAARs for the total sample with MacKinlay market-model returns

Days N Mean Cumulative Abnormal Return Precision Weighted CAAR Positive: Negative Generalized Sign Z (-20, -1) 111 0.75% 0.31% 55:56 0.260 (-1, 0) 111 0.30% 0.53% 60:51 1.210 (0, +1) 111 0.69% 0.45% 63:48 1.780* (0, +20) 111 -0.50% 0.08% 52:59 -0.309

(24)

Figure 3 Daily abnormal returns for the total sample

Warning and manager fine sample (warning sample hereafter) empirical findings: In table 4

the results of our regression on the warning sample are presented. Again, we do not see a pre-announcement or post-pre-announcement drift: both (-20, -1) and the (0, 20) window show insignificant results.

For the (0, +1) sample we find a positive CAAR of 0.75%, significant at the 5% level. An explanation for this result is that the announcement of the SEC that only a fine for the manager or a warning is imposed may have a positive effect on the stock price. The sanction is considered better than expected. An analysis of the pre- and post-announcement windows suggests that firms do not receive information on the severity of the penalty, because no significant pre-announcement stock price effect shows up.

In appendix 4 the daily AARs are calculated for the (-20, +20) window. Again, we observe both positive and negative AARs and therefore we conclude there is no pre- or

post-announcement drift due to the coincidence factor described in the previous section.

-0,60% -0,40% -0,20% 0,00% 0,20% 0,40% 0,60% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Daily average abnormal returns total

sample

Daily abnormal returns total sample

(25)

Table 4 CAARs for the warning and manager fine sample with MacKinlay market-model returns Days N Mean Cumulative Abnormal Return Precision Weighted CAAR Positive: Negative Generalized Sign Z (-20, -1) 50 0.02% -0.55% 23:27 -0.296 (-1, 0) 50 -0.04% 0.12% 23:27 -0.296 (0, +1) 50 0.75% 0.60% 30:20 1.686* (0, +20) 50 -0.84% 0.26% 24:26 -0.013

* denotes statistical significance at the 0.05 level

Figure 4 Daily abnormal returns for the warning sample

Fine, prejudgment interest and illegal-gotten gains sample (fine sample hereafter) empirical findings: Table 5 presents the main empirical findings for the fine sample. We observe an

important difference with the warning sample for the (0, +1) window. In the warning sample there is a 0.75% significant positive CAAR for the MacKinlay market-model. In the fine sample there is no significant positive CAAR for the (0, +1) window. Furthermore, we also do not observe a significant positive CAAR for the (-1, 0) window. This evidence suggests that if the sanction is a warning or a fine for the manager only, it is considered to be better than expected and there is a positive stock price effect. If, however, the company does receive

-1,40% -1,20% -1,00% -0,80% -0,60% -0,40% -0,20% 0,00% 0,20% 0,40% 0,60% 0,80% 1,00% 1,20% 1,40% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Daily average abnormal returns

warning sample

Daily average abnormal returns warning …

(26)

a fine or has to repay illegal-gotten gains plus prejudgment interest the sanction does not have a positive stock price effect. This result emphasizes the importance of the separation of the 2 samples: we observe that the type of fraud punishment matters.

The results for the pre- and post-announcement drift for the fine sample are similar to the results of the total sample and warning sample: there is no significant effect. From appendix 5 we observe that there are statistically significant daily AARs, both positive and negative, but as explained before we ascribe this to the coincidence factor.

Table 5 CAARs for the fine, prejudgment interest and illegal-gotten gains sample with MacKinlay market-model returns

Days N Mean Cumulative Abnormal Return Precision Weighted CAAR Positive: Negative Generalized Sign Z (-20, -1) 62 2.08% 0.95% 33:29 0.758 (-1, 0) 62 0.45% 0.68% 36:26 1.521 (0, +1) 62 0.39% 0.22% 32:30 0.504 (0, +20) 62 0.10% 0.13% 28:34 -0.512

Figure 5 Daily abnormal returns for the fine sample

-1,00% -0,80% -0,60% -0,40% -0,20% 0,00% 0,20% 0,40% 0,60% 0,80% 1,00% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20

Daily average abnormal returns

fine sample

Daily average abnormal returns fine sample

(27)

5. Conclusion & Discussion 5.1 Conclusion

This study provides an insight on the impact of fraud sanctions imposed by the SEC on the stock price. The general research question reads as follows: What is the influence of sanctions imposed by the Security & Exchange Commission as a consequence of fraud on the stock performance of these companies?

Contrary to prior research, we find a significant positive CAAR in the (0, +1) window for the warning sample. Moreover, we do not find a significant negative CAAR for the (0, +1) window for the fine sample or total sample. Some prior research also found a pre- or post-announcement drift. The results of this study show insignificant drifts on both sides of the (0, +1) window.

This research also provides information on the predictability of SEC fraud sanctions. The absence of announcement drifts suggests that information regarding the fraud sanction is unknown to the market until the day of the announcement. The effect of the announcement is immediately observable in the warning sample.

We tested two hypotheses: the reputation hypothesis and the government governance hypothesis. The reputation hypothesis is not confirmed: a fraud sanction does not result in a loss in firm value. Of course, there could still be a loss in reputation. However, this effect may not translate in a negative CAAR and is therefore not observed in this study. The government governance hypothesis is also not confirmed: there are no significant negative CAARs in our results while we expected a more severe negative stock price reaction due to heavier

punishment by the SEC.

We conclude that the type of fraud sanction influences the stock price effect and that the change in government governance has not lead to a more negative stock price reaction.

5.2 Discussion

This study intends to shed light on the effect of SEC fraud sanctions on stock price during recent years. The results differ from prior research; a clear difference is observed in the form of a positive CAAR for the warning sample. Prior research found significant negative CARs in the range between -0,5% and -22,7%. Furthermore, studies like Agrawal & Chadha (2005) found significant pre- and post-announcement drifts. This study does not observe significant

(28)

pre-and post announcement drifts.

The reason for this difference may be that prior research does not always take the AAERs as the measure of a fraud announcement. Some studies use the Wall Street Journal as their source for fraud announcement.

Another possible cause for the difference in results is the separation of the sample. This study separates the total sample in a warning and fine sample while prior research takes all fraud cases in one general sample. The negative CAARs of severe punished firms may in this case fade out the positive CAARs of the warning-only firms. This could possibly make a positive effect unobservable.

Moreover, this study focuses on the pre- and post-announcement drifts on a 20-day window basis. A limitation of this approach is the possibility that the market receives information on the fraud sanction more than 20 days before the announcement. For instance, if the Wall Street Journal publishes an article with a possible fraud case against a firm. This could cause insignificant pre-announcement drifts because information is reflected in stock prices before the (-20, 0) window.

The suggestion for further research is a study that examines the announcement drifts using a larger window. Marciukaityte et al. (2006) for instance use windows up to 5 years after the announcement to capture the post-announcement drift. For this study these event windows were unusable due to the recent data in the sample. Moreover, the study should separate the different types of sanctions to examine the effect of different types of fraud sanctions. A suggestion for a more precise study on the CAARs is to separate the firms in the data sample per market.

Another cause of possible bias in the results is the difference in the amount of data available for each year examined. The SEC may have punished more severe in some years than other years and an uneven distribution of data over the years may have caused biased results.

(29)

References:

Agrawal, A., & Chadha, S. (2005). Corporate governance and accounting scandals. Journal of

Law and Economics, 48(2), 371-406.

Agrawal, A., Jaffe, J.F., & Karpoff, J.M. (1999). Management turnover and governance changes following the revelation of fraud. Journal of Law and Economics, 42(1), 309-342.

Bhagat, S., Bizjak, J.M., & Coles, J.L. (1998). The shareholder wealth implications of corporate lawsuits. Financial Management, 27(4), 5-27.

Brown, S.J., & Warner, J.B. (1985). Using daily stock returns: The case of event studies.

Journal of Financial Economics, 14(1), 3-31.

Fama, E. (1969). Efficient Capital markets: A Review of Theory and Empirical Work. The

Journal of Finance, 25(2), 383-417.

Ferris, S.P., & Pritchard, A. C. (2001). Stock price reactions to securities fraud class actions under the private securities litigation reform act. The University of Michigan Law and

Economics Research, Paper No. 01-009.

Fich, E.M., & Shivdasani, A. (2007). Financial Fraud, Director Reputation, and Shareholder Wealth. Journal of financial economics, 86(2), 306-336.

Francis,J., Philbrick, D., & Schipper, K. (1994). Shareholder Litigation and Corporate Disclosures. Journal of Accounting Research, 32(2), 137-164

Griffin, P.A., Grundfest, J., & Perino M.A. (2000). Stock Price Response to News of

Securities Fraud Litigation: Market Efficiency and the Slow Diffusion of Costly Information.

Working Paper no.208, Stanford Law School.

Johnson, S.A., Ryan Jr., H.E., & Tian, Y.S. (2009). Managerial Incentives and Corporate Fraud: The Sources of Incentives Matter. Review of Finance, 13(1), 115-145.

Karpoff, J.M., Lee, D.S., & Martin, G.S. (2005). The cost to firms of cooking the books.

(30)

Karpoff, J. M., Lee, D. S., & Martin, G. S. (2007). The legal penalties for financial misrepresentation. Working Paper no. 2-7, University of Washington, 1-53.

Karpoff, M., Lee, D. S., & Vendrzyk, V.P. (1999). Defense Procurement Fraud, Penalties, and Contractor Influence. Journal of Political Economy, 107, 809-842.

Karpoff, J.M., & Lott, J.R., (1993). The reputational penalty firms bear from committing criminal fraud. Journal of Law and Economics, 36, 757–802.

Keller,G. (2012). Managerial Statistics. Canada: Ontario: Joe Sabatino.

Kellogg, R.L. (1984). Accounting activities, security prices, and class action lawsuits. Journal

of Accounting and Economics, 6(3), 185-204.

Kirkpatrick, G. (2009). The corporate governance lessons from the financial crisis. OECD

Journal: Financial Market Trends, 1(3), 61–87.

MacKinlay, A.C. (1997). Event Studies in Economics and Finance. Journal of Economic

Literature, 35(1), 13-39.

Marciukaityte, D., Szewczyk S.H., Uzun, H., & Varma R. (2006). Governance and Performance Changes after Accusations of Corporate Fraud. Financial Analysts Journal, 62(3), 32-41.

Niehaus, G., & Roth, G. (1999). Insider Trading, Equity Issues and CEO Turnover in Firms Subject to Securities Class Action. Financial Management, 28(4), 52-72

Persons, O. (2006). The effects of fraud and lawsuit revelation on U.S. executive turnover and compensation. Journal of Business Ethics, 64(4), 405–419.

Rezaee, Z. (2005). Causes, consequences, and deterrence of financial statement fraud. Critical

Perspectives on Accounting, 16(3), 277–299.

Soltani, B. (2014). The Anatomy of Corporate Fraud: A Comparative Analysis of High Profile American and European Corporate Scandals. Journal of Business Ethics, 120(2), 251-274.

Swarcz, S. L. (2008). Protecting Financial Markets: Lessons from the Subprime Mortgage Meltdown. Minnesota Law Review,93 ,373-406.

Yiu, D. W., Xu, Y., & Wan, W.P. (2014). The Deterrence Effects of Vicarious Punishments on Corporate Financial Fraud. Informs Journal, 25(5), 1549-1571

(31)

7. Appendices Appendix 1 Selected securities Company Fine in dollar x1000 () = prejudgment interest [] = illegal-gotten gains

payback Date PERMNO

MBIA Inc. 50000+[1000] 20070129 79714

RenaissanceRe Holding Ltd. 15000 20070206 83728

The Dow Chemical Company 325 20070213 20626

Veritas Software Corp. Warning 20070221 80055

Bennett Environmental Inc. Warning+75* 20070221 89047

NorthWestern Corp. Warning 20070307 90458

Collins&Aikman Corp. Warning 20070326 30440

Atlas Air WorldWide Inc. Warning 20070329 91262

Tenet Healthcare Corp. 10000 20070402 52337

Baker Hughes Inc. 10000+(3133)+[19945] 20070426 75034

Capitol Distributing LLC Warning+50* 20070502 92474

Motorola Inc. Warning+(18000)+[7000] 20070508 22779

Penthouse International Inc. Warning 20070510 x

The BISYS Group Warning+([25000]) 20070523 77426

International Business Machines Corp. Warning 20070605 12490

Allied Capital Corp. Warning 20070620 79788

Cambrex Corp. Warning 20070620 11707

CVS Caremark Corp. Warning 20070629 17005

OM Group Inc. Warning 20070718 79724

Quovadx Warning 20070717 87648

ConAgra Foods Inc. 45000 20070725 56274

Cardinal Health Inc. 35000 20070726 21371

Delta&Pine Land Company(Turk

Deltapine) 300 20070726 79255

Aspen Technology Warning 20070726 80957

Integrated Silicon Solution Inc. Warning 20070801 81256

First BanCorp 8500 20070807 11285

Integrated Electrical Services Warning 20070830 85768

Saks Inc. Warning 20070905 11382

Berger, Apple&Associates Warning 20070913 x

Beutel Accountancy Corp. Warning 20070913 x

Bray& Associates Warning 20070913 x

Bujan&Associates Warning 20070913 x

Forbush&Associates Warning 20070913 x

F.X. Duffy&Co Warning 20070913 x

Ferro Corp. Warning 20070913 21135

Electronic Data Systems Company Warning+(132)+[359] 20070925 83596

(32)

Federal Home Loan Mortgage Corp. 50000 20070927 75789

Tidewater Inc. Warning 20070927 50606

Nortel Networks Corp. 35000 20071015 58640

MQ Associates Warning 20071018 x

Lucent Technologies Inc. 1500 20071221 83332

AXM Pharma Inc. Warning 20080214 x

Westinghouse Airbrake Technologies Warning+(29)+[259] 20080214 81677

Biovail 10000 20080324 80307

Savvides & Partners/PKF Cyprus 106+(49)+[106] 20080415 x

Broadcom Corp. 12000 20080422 85963

McCann-Erickson WorldWide inc. 12000+[1000] 20080501 x

Interpublic Group of companies Inc. Warning 20080501 53065

UTStar Inc. Warning 20080501 87825

GlobeTel Communications Inc. Warning 20080501 90686

Willbros Group. (1400)+[8900] 20080514 83834

Brooks Automation Inc. Warning+(29)+[259] 20080519 81241

Faro Technologies Inc. Warning+(439)+[1411] 20080605 85372

CitiGroup Inc. Warning 20080616 70519

NEC Corporation Warning 20080617 x

Sycamore Inc.

Warning+420*+(58)*+

[12040]* 20080709 87343

El Paso Corp. Warning 20080711 77481

HCC Insurance Holdings Warning 20080722 78033

Ernst & Young LLp Warning+(537)+[2382] 20080805 x

Prudential Financial Inc. Warning 20080806 89258

Con-Way Inc. 300 20080827 41929

United Rentals Inc. 14000 20080908 85663

American Italian Pasta 7500+275*+(32)*+[752]* 20080915 85411

Blue Coat Systems Warning 20081112 87377

Zurich Financial Services 25000+[1000] 20081211 x

Siemens AG 450000+[350000] 20081215 88935

Stewart Entreprises Inc. Warning 20081229 77045

Rk Dhawan & Co. 240* 20080923 x

ITT company 250+(397)+[1041] 20090211 12570

KBR inc. 402000+[177000] 20090211 91579

Halliburton company 402000+[177000] 20090211 23819

Meridian Holdings 250* 20070927 x

Research in Motion Limited 76850 20090205 86745

Pediatrix Medical Group/Mednax

Services Warning 20090305 82272

PowerCold Cooperation Warning+75 20090311 x

Quest Software 144*+(5808)* 20090312 87182

Allion Healthcare Inc Warning 20090318 90717

Mercury Interactive 425 20090320 79718

(33)

Mercury Interactive 300 20080917 79718

Delphi Corporation 30 20090327 76099

Take-Two Interactive Software 3000 20090401 84761

Take-Two Interactive Software 1000+(1143)+[4118] 20070214 84761

Stratum Holdings Warning 20090414 x

Ingram Micro Inc. Warning 20090512 84168

Monster Worldwide Inc. 2500 20090518 84342

WellCare Health plans 10000 20090518 90272

Apogee Technology Warning+30 20090519 12118

TMT Capital Corporation Warning 20090521 x

CSK Auto Corporation Warning 20090526 26060

Pollard Kelley Auditing Services Inc. Warning 20090527 x

United Industrial Corporation Warning 20090529 41371

Dyadic International Inc. Warning 20090604 90682

Blackout Media Corporation ( First

Canadian) Warning 20090612 x

Comverse Technology Inc. Warning 20090618 10942

Ulticom Inc. Warning 20090618 88202

LSB Industries Inc. Warning 20090717 49488

West Marine Inc. Warning 20090723 79884

Avery dennison Corporation warning 20090728 44601

MedQuist Inc. Warning+75* 20090312 77575

General Electric Company 50000 20090804 12060

Escala Group Inc. Warning 20090323 79167

General Motors Corp. Warning 20090122 12079

Terex Corporation 8000 20090812 58318

United Rentals Inc. 14000 20080908 85663

Entrade Inc. Warning 20090819 31472

Verifone Holdings Inc. Warning+25* 20090911 90657

The Hain Celestial Group Inc. Warning 20090903 80167

Tenet HealthCare Corporation

Warning+500*+(251)*+

[1780]* 20090911 52337

Tenet HealthCare Corporation Warning 20070402 52337

Dana Holding Corporation Warning 20090911 92570

China Holdings Inc. Warning 20091030 92475

Symbol Technologies 3300 20091102 73940

Merge Healtcare Inc.

Warning+140*+(167)*

+[570]* 20091104 85739

Safenet Inc. 1000+365*+(235)*+[1792]* 20091112 78106

Bancinsurance Warning 20091116 76308

Home Solutions Warning+180* 20091130 89767

Black Box Corp. Warning 20091204 78172

Ernst & Young LLP Warning 20091217 x

UTStarcom Inc. 3000 20091231 87825

NATCO group Inc. Warning+65 20100111 87539

(34)

Assurant Inc. 3500 20100127 90038

Tsukuda-America Inc. Warning 20100128 x

Verint Systems Inc. Warning 20100303 89397

Morgan Asset Management Inc. Warning 20100407 x

Morgan Keegan & Company Inc. Warning 20100407 x

Collins& Aikman Corp. Warning 20070326 80713

Collins& Aikman Corp.

Warning+400*+(2376)*+

[4424]* 20100419 80713

Spongetech Delivery Systems Inc. Warning 20100505 x

Diebold Inc. 25000 20100602 40440

Lucent Technologies Inc. Warning+65* 20100607 83332

Snamprogetti Netherlands 240000 20100707 x

Technip 240000+[98000] 20100628 89196

Trident Microsystems Inc. Warning+400*+[(1177)]* 20100716 78208

Dell Inc. 100000+4000* 20100722 11081

Sunrise Senior Living Inc. Warning+50*+(31)*+[83]* 20100723 83558

General Electric Company 1000+(4080)+[18397] 20100727 12060

Navistar International Corp. Warning 20100805 12503

Universal Corporation Inc. 4400+[4581] 20100806 16555

Alliance One International Corp. 9450+[10000] 20100806 29867

Affiliated Computer Services Inc. Warning 20100909 80913

Sunopta Inc. Warning+(6305)*+[46]* 20100924 78418

ABB Ltd. 35510+([22804]) 20100929 88953

Office Depot 1000 20101021 75573

LocatePlus Holding Corporation Warning 20101014 x

GlobalSantaFeCorp. 2100+(1063)+[2694] 20101104 x

Transocean Inc. (Ltd) (1283)+([5981] 20101104 79237

Pride International 32625 20101104 12074

Royal Dutch Shell 30000+[18150] 20101104 90793

Panalpina Inc. 70560 20101104 x

Noble Corporation 2590+(1282)+[4294] 20101104 90537

Tidewater Inc. 217+(881)+[7223] 20101104 50606

Vitesse SemiConductor Corp. 3000+(63)*+[136]* 20101210 77173

Hudson Highland Group 200 20110110 89698

NIC Inc. 500 20110112 87066

NutraCea Warning 20110113 x

Maxwell Technologies Inc. 8000+(695)+[5654] 20110131 51960

Deerfield Capital Corp. Warning+(300)+[977] 20110202 90752

ArthroCare Corp. Warning 20110209 83113

DHB Industries (Point Blank Solutions

Inc,) Warning 20110228 86355

KPMG Australia Warning +(760)+[1982] 20110228 x

International Business Machine

Corporation 2000+(2700)+[5300] 20110318 12490

Ball Corp. 300 20110324 57568

(35)

Price Waterhouse &Co. 6000 20110405 x

Satyam Computer Services Limited Warning 20110405 88995

Kentucky Energy Warning 20110408 x

Kempisty&Company Warning 20110408 x

SouthPeak Interactive Corp. Warning+50* 20110421 x

Rockwell Automation Inc. 400+(590)+[1771] 20110503 84381

Michael Baker Corp. Warning 20110511 x

Thor Industries Inc. 1000 20110513 76081

GSI Group Inc. Warning 20110513 82261

Livingstone&Haynes 130 20110606 x

Morgan Asset Management Inc. 75000+(4500)+[20500] 20110622 x

Morgan Keegan & Company Inc. 75000+(4500)+[20500] 20110622 x

LaBarge Inc. 200 20110630 47255

Armor Holdings Inc. 3680+(458)+[1552] 20110713 83189

Diageo plc Warning+(2067)+[11306] 20110727 76592

LPB Capital d/b/a Family Office Group

LLC Warning 20110927 x

Watts Water Technologies Inc. 200+(820)+[2756] 20111013 10606

Koss Corporation Warning 20111024 48072

Kempisty&Company Warning 20111214 x

Aon Corp. 1764+(3128)+[11416] 20111220 61735

Life Partner Holdings Inc. Warning 20120104 89888

JBI Inc. Warning 20120104 x

R.E. Bassie Inc. 75 20120110 x

Imperiali Inc. Warning 20120111 x

Symmetry Medical Inc. Warning+25* 20120130 90515

Smith&Nephew PLC Warning+(1399)+[4028] 20120206 87444

Biomet 17280+(1142)+[4433] 20120326 18092

Sinotech Energy Limited. Warning 20120423 12396

FalconStor 2900 20120628 89099

Wyeth LLC Warning +(1659)+[17217] 20120808 15667

Pfizer Inc. Warning +(10307)+(16033) 20120808 21936

Peak Wealth LLC Warning 20120810 x

Gold Standard Mining Corp. Warning 20120629 x

Gruber&Company Warning 20120629 x

China Sky One Medical Inc. Warning 20120904 92646

Tyco International Ltd. Warning+(2566)+[10565] 20120924 45356

Sunrise Solar Corp. Warning 20121001 x

* Fines, illegal-gotten gains payback and prejudgment interest imposed on managers of the firm

(36)

Appendix 2

(37)
(38)

Appendix 3

(39)

Appendix 4

(40)

Appendix 5

Mean abnormal returns for the fine, prejudgment interest and illegal-gotten gains sample

(41)

Appendix 6

CAARs for the different samples

CAARs for the total sample

CAARs for the warning and manager fine sample

Referenties

GERELATEERDE DOCUMENTEN

This research set out to create a comprehensive understanding of the literature related to the challenges associated with the diffusion of electric vehicles and has

The general mechanical design of the Twente humanoid head is presented in [5] and it had to be a trade-off between having few DOFs enabling fast motions and several DOFs

In this thesis is researched whether there are patterns in the involvement of perpetrators and the use of fraud techniques regarding the time span of

In this study, the term advertising fraud is understood to refer to ‘misleading business practices between organisations involving sales techniques aimed at building up trust

Moreover, the market betas of the portfolios with high customer satisfaction results (both based on relative and absolute ACSI scores) are considerably lower compared

occupational fraud and classified them according to the classification framework of Shields (1997). Overall, the findings of the literature showed that executives’ main incentive

As the weather variables are no longer significantly related to AScX returns while using all the observations, it is not expected to observe a significant relationship

Tobin’s q is measured as the market value of common equity plus the book value of total assets minus common equity and deferred tax all divided by the book value of total assets..