Do the AFM enforcement preferences influence a firm’s likelihood of committing a violation?

Amsterdam Business School

Do the AFM enforcement preferences influence a

firm’s likelihood of committing a violation?

Name: Nick Lamers

Student number: 10169830

Thesis supervisor: dhr. dr. W.H.P. Janssen & dhr. prof. dr. V.S. Maas

Date: 20-06-2016

Word count: 11839

MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam

Statement of Originality

This document is written by Nick Lamers, 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 thesis researches whether the Dutch AFM has preferences in its enforcement actions. The AFM is the equivalent of the SEC and has the right to punish companies if they conduct in misbehavior. This thesis contributes to the existing literature in two ways; this kind of research hasn’t been done yet towards the AFM and it will help to understand the behavior of this institution.

The preferences of the AFM are in this study expressed in the amount of penalty they give to the punished companies. The variables used in the model are firm size, distance towards the AFM-office and the auditor of the punished company. Firm size is measured in total assets, distance is measured in kilometers and for auditor a proxy is used when they have a Big 4 audit firm. Most of the data are hand collected through the AFM register and the remaining data comes from the Amadeus database.

After performing a statistical regression, it can be concluded that larger firms receive higher penalties. For each 1% increase in assets, an increase of 0.27% in penalty is received.

Furthermore, firms closer to the AFM receive higher penalties. Each kilometer a firm is located further away from the AFM, will decrease the penalty with 1.03%. So, summarizing the AFM has enforcement preferences which are expressed in the amount of penalties they give. Larger firms receive higher penalties, while firms further away from the AFM receive lower penalties.

Table of contents

1 Introduction ... 6

1.1 Background information ... 6

1.2 Research question ... 7

1.3 Contribution and outline ... 8

2 Literature review ... 9

2.1 The AFM ... 9

2.1.1 Organization of the AFM ... 9

2.1.2 Enforcement process ... 10

2.1.3 Measures of the AFM ... 12

2.1.4 Financial penalties ... 13

2.2 Enforcement preferences and costs ... 14

2.2.1 SEC’s enforcement preferences ... 14

2.2.2 Costs of punishments ... 16

2.3 Role of the auditor ... 17

2.4 Hypotheses ... 18 3 Data selection ... 20 4 Methodology ... 23 4.1 Research model... 23 4.2 Variables ... 23 4.2.1 Size ... 23 4.2.2 Distance ... 23 4.2.3 Big 4 auditor ... 24 5 Results... 25 5.1 Descriptive statistics ... 25

6 Conclusion ... 30

6.1 Conclusion... 30

6.2 Limitations... 31

6.3 Further research... 32

1 Introduction

1.1 Background information

The Dutch financial authority has punished in 2015 much less than in the previous years (Financieele Dagblad, 2015). The AFM (Autoriteit Financiële Markten), which is the Dutch equivalent of the SEC, has almost a decline of 50% in punishments compared to 2014. The reason for this decline is still unknown and also the AFM itself would not react to this fact. They first want to discuss it intern. These punishments are for individuals and for companies. For example, two directors of Imtech get a fine of €1 million and €1.35, while the Dutch bank ABN Amro was punished by €750,000. Reasons were respectively misleading and keeping back of sensitive information and unfair and non-professional behavior.

But the reason for the decline is still a guess. Was there really less lawbreaking and undesired behavior or was the AFM less strict compared to the previous year. It also might be that the AFM has some preferences about how they are doing their investigations, which companies are getting more control and how many controls they perform each year. They also can have a preference by punishing some firms harder in financial way than other firms. In the United States, companies are less likely to conduct in financial misbehavior the closer they are located to the SEC (Kedia & Rajpogal, 2011). Coval & Moskowitz (2001) claim that information doesn’t travel any further than 100 kilometers, so firms outside the range of 100 kilometers from the SEC are less able to monitor the activities of the SEC. The SEC is comparable to the AFM; they both have the task to regulate companies and the financial market in their country. Further, Kedia & Rajpogal (2011) conclude that when the SEC had enforcement actions in the

neighborhood, this also decreases the likelihood of committing financial misbehavior. The main reason is that firms closer to the SEC are better informed about their actions. Additionally, Correia (2014) found that firms with political connections are less likely to be investigated by the SEC. Companies with larger donations to politicians or political parties face also lower penalties when they get caught. So, based on Kedia & Rajpogal (2011) and Correia (2014) the SEC does have some enforcement preferences, based on different factors. DeFond et al. (2011) also found location as important factor in the SEC preferences. They concluded that audit firms are more likely to issue a going concern report to firms who are located farther away from a SEC-office.

Files (2012) concludes that when firms try to cooperate with the SEC after a restatement, the likelihood of being sanctioned increases. The reason for this is that the SEC is in this way better able to build a strong case against the firm. On the other hand, the financial punishments are lower compared to companies who are less willing to cooperate. Feroz et al. (2008) found

that smaller audit firms receive more severe punishments from the SEC than bigger firms. An explanation could be that bigger firms have greater resources and therefore perform better audit services. DeHaan et al. (2014) state that the revolving-door phenomenon causes harder

punishments. SEC-lawyers who know they will be joining another company in a few years, want to show their abilities by making strong cases which lead to higher punishments.

The potential costs for firms are also important, because the purpose of the penalties is meant to scare the companies off committing a financial crime. According to Karpoff et al. (2008), when firms are getting caught, they face two different kind of costs; legal costs and market costs. Legal costs are the penalties or fines companies received from the SEC, while market costs are punishments by the market, like reputational damage. They conclude that market costs have a much higher impact than legal costs. For every dollar lost through

restatements, a loss of $3,08 (rounded) is made. From this $3,08, only $0,36 is the result of legal sanctioning. The remainder of $2,71 is the result of lost reputation in the market. Glaeser et al. (1996) state additionally that the potential costs of a crime are for each company or individual different. The same crime could be differently judged by different companies and have therefore different estimations about the potential economic costs of the crime.

1.2 Research question

This thesis will be doing further research on the topic of Kedia & Rajpogal and Correia, but in another country. Based on their articles, the SEC has enforcement preferences in the USA. These preferences are especially based on physical location and political contributions. More existing literature researched the role of the SEC and its activities. Where is SEC is responsible in the USA, the AFM is in The Netherlands the main institution for regulatory oversight. I do think that this study reveals other results than for the SEC, which are stated in the next paragraph. So, when the SEC has enforcement preferences, why wouldn’t this happen for the AFM? Therefore, the research question of this thesis is:

Do the AFM enforcement preferences influence a firm’s likelihood of committing a violation?

The preferences of the AFM will be measured by three variables; firm size, distance to the AFM office and the auditor of the firm. In the research model, it will be investigated whether these three factors have influence on the amount of the penalty given by the AFM to the company in violation. The preferences of the AFM will be therefore expressed by the amount of the given

penalty. Three different hypotheses will explain the expectations of this study, and these hypotheses will be tested with a statistical regression.

1.3 Contribution and outline

This thesis contributes in two ways. First, this kind of research hasn’t been done in The

Netherlands yet, so in that way it will contribute to the existing literature and fill in the gap that consists. As mentioned before, in the United States the SEC has been a lot chosen as topic for the studies. In The Netherlands, the AFM has a similar role as the SEC but is not investigated yet. Less research has been done towards the activities and potential preferences of the AFM, while it is playing a major role in the legal environment of the Dutch companies. On contrary of the SEC researches, I do expect different outcomes for this study towards the AFM. For

example in The Netherlands, I think distance will play a smaller role than in the United States, because of the fact that the country is much smaller and the infrastructure is better. Furthermore, political contributions might play a smaller role, because the political climate is totally different in the Netherlands.

Second, it has also a practical relevance. The AFM is the main institution for financial issues in The Netherland and has the main task of oversight of the Dutch market. It has the right to punish companies for financial misbehavior and to give financial and non-financial penalties. Therefore it is important that such kind of an institution works in the right way and don’t have any lacks in their processes. This thesis will contribute to get a better understanding of the activities of the AFM. It will reveal potential preferences in their enforcement actions, on which companies can decide if they have to adjust their activities.

In the second paragraph the background theory will be presented, which is necessary to answer the research question. It will also cover an explanation for the three hypotheses. In the third paragraph the data selection will be revealed. The fourth paragraph will present the research methodology of the study and the fifth paragraph discusses the results of the statistical

regressions. At last, the sixth paragraph will give the conclusion and the answering of the research question. It also covers the limitations of this study and some recommendations for further research towards the AFM.

2 Literature review

2.1 The AFM

The AFM (Authoriteit Financiële Markten / Authority Financial Markets) is the Dutch equivalent of the SEC. The main purpose of the AFM is to monitor the Dutch market and companies who are active in this market. The AFM started in 2002 and is indirect a

governmental institution. It is the successor of the Stichting Toezicht Effectenverkeer and located in Amsterdam. The minister of Finance is responsible for the appointment of the directors and also has to approve the budgets. The purpose of the AFM is also legally recorded in the Wft (Wet financieel toezicht / law financial oversight). In art. 1:25 Wft (2006) is described that the AFM is the responsible institution to oversee the Dutch financial market and to make sure that

companies in the market are following the rules. When necessary, the AFM is permitted to warn or punish companies who are violating the laws.

2.1.1 Organization of the AFM

As mentioned earlier, the AFM is under supervision of the minister of Finance (AFM, 2016). This minister is responsible for the budget of the AFM and has also the right to appoint members of the direction. Besides that, the minister has also the right to hire and fire members of the supervisory board. This board has the job to approve the planning, the budget and the financial statements. The main task of the supervisory board is to keep an eye on the actions of the direction. Under this direction, four different divisions are structured: the General Counsel, Compliance & Integrity, a Fine-functionary and the Internal Audit Service.

Other than the relation with the minister of Finance, the AFM has more governmental relationships. On their website they describe also the following relations. In order to check the Dutch pension funds, the AFM works closely with the ministry of Social Affairs and Security. Further, the AFM has an advisory role to the ministers when developing new laws on this topic. Additionally, the AFM holds close relations with De Nederlansche Bank (DNB, The Dutch Bank), from which the Dutch Kingdom is the owner. This bank has the monopoly of pressing money. The AFM and the DNB are working together in different fields, for example on giving or drawing back licenses of financial institutions. The AFM emphasizes the collaboration with the DNB, because they both consider it important to indicate the risks on the Dutch markets.

2.1.2 Enforcement process

Before coming to a final punishment, a lot of activities happen in order to decide which kind of measure is appropriate. The measures of the AFM will be described in the next paragraph. The first condition for a penalty is that there has to be a violation. In The Netherlands different kinds of laws are applicable for the AFM. But most of the violations are disclosed in the Wet op het

Financieel Toezicht (Wft / Law on Financial Oversight). In this law not only the main tasks of the

AFM are disclosed, also the “criminal” financial activities can be found. Other important laws on which the AFM could determine the punishment are Wet handhaving Consumentenbelang (Law protecting Consumer’s interest) and Wet toezicht Accountantsorganisaties (Law oversight on audit firms).

Karpoff et al. (2008) give in their research a schematic overview about the enforcement process of the SEC, which consists of three periods; the violation period, the enforcement period and the regulatory period.

Source: Karpoff, J. M., Lee, D. S., & Martin, G. S. (2008). The cost to firms of cooking the books. Journal of Financial and Quantitative Analysis, 43(03), 581-611.

Before the SEC comes in action, first there is a violation period in which the violation happens. In The Netherlands this can be a violation of the laws from the Wft, but also a financial crime according to the Criminal Code. After that, a trigger event must happen, when the SEC gets an indication about a possible violation. This can come from a whistleblower, a newspaper or from the company itself by submitting some restatements (Karpoff et al., 2008). The SEC performs also routine actions in which firms randomly can be drawn and get a visit of the SEC. During this visits the SEC can also detect violations. Might the trigger event show a potential violation, the SEC prepares an informal inquiry in which they ask the firm for information about the activities around the violation. When the SEC is not satisfied, it can decide to perform a formal investigation. Firms who are getting a formal investigation can make a public statement that they are under formal oversight (Karpoff et al., 2008). If it appears that the firm really has violated a

law, legal actions are taken. This is the start of the regulatory period. According to Files (2012), firms who actively cooperate with the SEC have a higher likelihood of getting a penalty. Firms are more reluctant by giving insight in their activities, which results that the SEC is better able to build a stronger case against this firm. On the other hand, Files also proves that these firms receiving lower penalties compared with firms who are more closed towards the SEC. In this way the SEC wants to reward the firm for their cooperation.

During the regulatory period the final punishment will be decided. All stakeholders to the violation can be questioned and the SEC has the collect more evidence around the violation to come to a final judgement. In the United States some cases can go through the Department of Justice if the violation serious enough (Karpoff et al., 2008). After the SEC has done its final judgement the firm has the opportunity to make an objection to the punishment.

For the AFM the whole process is quite similar. After the violation period a trigger event has to happen in order to activate the AFM. A special factor is that in The Netherlands recently a new law for whistleblowers is accepted, which must give the whistleblower more protection to the negative consequences. Due to this fact, the amount of whistleblowers increased with 32% to an amount of 74 individual cases compared with the previous year (NRC Handelsblad, 2016). When the trigger event indicates a potential violation, the AFM can decide to perform an investigation. About the measures and the financial penalties a detailed explanation will be discussed in the next paragraph.

After determining the punishment, the firm can decide to make objection. A firm has three possibilities to reduce or delete the punishment. First, they can make objection at the AFM itself. If they decide that the punishment is justified, the firm can go to a judge who will also take a look at the punishment. After that, the firm has also the possibility to go to the College van

Beroep voor het Bedrijfsleven (CBb, College of objection for business). When they decide that the

punishment is right, the penalty will be final. Sometimes not only the AFM punishes the company, if the crime is serious enough the firm or individual can also be prosecuted by the Dutch Court and get a criminal record.

When the final punishment is a financial penalty, the AFM is legally obliged to make the penalty public (AFM, 2016). This goes by sending out a news article with the details about the company, the amount of the penalty and which law is violated. When the violation is of the heaviest kind (Category 3, see below), the AFM must made public immediately. This means that after communicating the penalty to the firm, the AFM has five working days to send out their news article. With smaller violations, the publication of the penalty will be made when it

becomes final and no objections can be made. All the punishments can be found in the register of the AFM, which is accessible on their website.

A company has the opportunity to make objection to the publication of the penalty. Looking into the registers, sometimes the name of the company is blurred in the news article. The AFM only decides this when their main task of oversight of the market can be

compromised. Sometimes the market costs for a firm are higher than the legal costs (Karpoff et al., 2008). This means that they will get punished more by the market than by the AFM.

Objectively, the AFM should not bother about this, but if for example a major bank with a lot of Dutch customers can go bankrupt by stating their name in the news article, it can make a big reputational loss. If the AFM decides to prevent this bank, because it might go bankrupt due to this reputational loss, the AFM can make the decision to bring the penalty publicly without the name of the bank on it. In this way, it will prevent all the citizens from losing their money and the bank is still getting punished by the financial penalty.

2.1.3 Measures of the AFM

In order to fulfill the main tasks of AFM, the institution has different kinds of measures it could take to execute their tasks. The AFM has the right to execute the next measures:

- Conversation or warning about violation of norms. In this case the AFM provides the company an oral or written warning about a violation of norms. The particular company has to make a plan about how this problem will be resolved.

- Public warning through press-release. The AFM could give out a press-release about the possible misbehavior of a company.

- Company is put on a warning-list. The AFM keeps a kind of black list, where the company will be put on. The companies on the list are more checked than other companies.

- Instruction. A company gets an instruction from the AFM to make within a stated timeframe clear improvements on the given warning.

- Provisional penalty. The company gets a provisional penalty and when the misbehavior isn’t resolved within a stated timeframe, this penalty will become executed.

- A fine. This fine will be given at once. So the company doesn’t get time for

improvements, but just has to pay the given fine. This will be the main topic in this research.

- Dismissal of directors. When it is clearly that a director of a company didn’t act according the law, it is also possible that only the director will get punished instead of a whole company. In this case, the AFM has the right to dismiss this director.

- Withdraw of license. The company could lose his license for acting in the financial market.

- Promise. In this case, a company has to make a public promise to do not involve in misbehavior in the future and make improvement in this behavior on the short term. As you can see, some measures are very look-a-like and it is also not excluded that different measures are combined to punish companies. As last step it is also possible that the AFM makes a declaration at the Public Prosecution Service. In this way, a company or a person could also get a criminal prosecution. By deciding the kind of measure, the AFM keeps in mind the different circumstances. For example the advantage of the company gets by breaking the law, the

timeframe of the misbehavior and to what kind the crime could be assigned to a whole company are all possible standards in deciding the measure.

2.1.4 Financial penalties

Since this study mainly focuses on the financial penalties from the AFM, some extra rules about the process deciding the amount of the penalty are explained. First, the AFM divides the

violations into three different categories. For each category a base amount for the penalty is determined, just as the maximum and minimum deviation of the amount. After that the AFM looks at different factors that might play a role during the violation and therefore can influence the received advantage of the violation. Based on these factors the base amount will be lower or increase the final penalty. The amount of the penalty depends on a few factors; the seriousness and duration of the violation, the extent to which the company is responsible for the violation, the recidivism of the violation and the advantage reached through the violation (AFM, 2016). At last also capacity of the firm plays a role.

The AFM has three categories of penalties; penalties smaller than €10.000 is Category 1. These are the smaller violations which occur relatively more. The €10.000 is the base amount and can be decreased on the hand of the factors mentioned above. Category 2 are the penalties between €500.000 and €1.000.000. Here is also the €500.000 the base amount and it is dependent on the different factors if this will be lowered or be higher. Category 3 has a base amount of €2.000.000 and handles the heaviest violations. Based on the culpability, the base amounts can be lowered. For recidivists, another rule is determined. If a firm gets punished twice for the same reason within five years, the amount of the penalty will be doubled. Furthermore, the base

amount can also increase if the firm’s violation has a big impact on the market or on individual clients or customers. Finally, the capacity of the firm does also play a role according to the AFM (2016). When a firm might get bankrupt due to the fine or has very less capital accessible, the AFM could decide to lower the penalty. When the firm is already in the bankruptcy process, the AFM still punishes the company and looked at the present capital before the bankruptcy was claimed.

2.2 Enforcement preferences and costs

2.2.1 SEC’s enforcement preferences

Where the AFM is still uninvestigated, different studies investigated the SEC and those studies found some preferences in the activities of the SEC. Kedia & Rajgopal (2011) researched how SEC’s enforcement actions influenced corporate misconduct. In their article they used two different hypotheses: the “differentially informed criminal” hypothesis and the “constrained cop” hypothesis. The differentially informed criminal focuses on the physical location of a firm and the relation to misconduct. Proxies for this hypothesis are the location of a firm and distance to the SEC-office and past activity in the neighborhood. The authors predict that firms are less likely to commit a violation when they are better informed about the SEC’s activities. The other hypothesis states that firms are keeping the possible constraints that the SEC has, in their mind while deciding a possible violation. Examples of these constraints are time or budged. After performing a statistical analysis, the authors could make different conclusions. First, companies who are physical closer the SEC are less likely to conduct in financial misreporting. Also, companies with more SEC-activities in the neighborhood are also less likely to misbehave. On the other hand, Kedia & Rajpogal found evidence for their constrained cop hypothesis. They state that SEC is more likely to investigate firms in the neighborhood. So it could be that the location close to the SEC office is ambiguous. There is less misreporting, but firms are also more investigated. At last, firms with more media attention also have a higher concentration of

enforcement actions. They argue that these firms have more public interest and should therefore better be watched. The SEC follows this idea and performs more enforcement actions against these firms, which are often the larger companies.

Correia (2014) did a kind of similar research, but wanted to find out to what extent there is a relation between SEC enforcement actions and political connections of the companies. In the article she listed firms who made donations to politics in the USA and made a statistical analysis if these firms are favorited by the SEC. With as main hypothesis, she stated that firms

with long-term connections are less likely to be prosecuted by the SEC and get lower penalties. Using PAC-contributions as proxy, the main conclusion is supporting the hypothesis. Firms showing their loyalty to politicians by donating funds seem to get a favorable treatment by the SEC. However, the result of lobbying appears to be less effective, but not for former SEC-employees. When lobbyists with prior connections to the SEC use these connections effectively, this could also decrease the costs of penalties. Heese (2015) also researched if there is a relation between the politics and the SEC, and found that some political factors may contribute to the preferences of the SEC. Indirectly, the politics will influence the voters by sparing large

companies during times of elections and areas with high unemployment rates are also less likely to be subject of enforcement actions. So, Heese concluded that the SEC enforcement

preferences could reflect the interest of the voters and the SEC retrieves its preferences from political factors.

Files (2012) also claims a sort of preferences at the SEC. She found that when firms actively try to cooperate with the SEC after a restatement, there is a higher likelihood of being punished. The SEC is able to get more sensitive information from the firm than when they have to perform an external investigation. Therefore with more information, the SEC is able to build a stronger case against this firm and increase the likelihood of finding misbehavior. On the other hand, the SEC wants to be reluctant towards companies who ask for cooperation, because they want to stimulate cooperation. So Files also found that when firms are caught, they face lower financial penalties as “reward” for their cooperation. So, when firms are cooperating with the SEC, they increase the likelihood of getting punished, but face lower financial penalties.

Furthermore, Feroz et al. (2008) found evidence of preferences in firm size. In their article about the relation between SEC enforcement and accounting and auditing firms, they found that smaller firms are receiving more severe punishments. In this case it is specific about smaller audit firms. Feroz et al. state that the reason for this is that bigger firms have greater resources and therefore have more quality in the firm, with as result that they perform less failures. So, therefore the SEC has to perform fewer actions against these bigger companies.

Firm size could also play a role in the SEC’s enforcement preferences, according to Heese (2015). He concludes that larger firms are less likely to be victim of the SEC’s enforcement preferences. This counts especially for years in which a presidential election is planned, but also but the other years. In the research towards the political connections and the SEC, he also finds that larger companies receive less comment letters. The main reason for this fact might be that the SEC uses fewer resources for the reviews of bigger firms and prefers the

smaller firms. Further, Heese argues that location also matters for firms. If companies are located in politically important states, it is less likely that they will be subject of an enforcement action of the SEC. In this way, it looks like politicians will influence indirectly there voters by favoring these large companies with a lot of employees.

Additionally, DeHaan et al. (2014) investigated if the revolving door phenomenon has influence on the enforcement actions of the SEC. The revolving door phenomenon means that employees have a short time at a firm and then move on to another firm. So, like a revolving door, they come in the company and within short time they go out. According to DeHaan et al. this phenomenon has influence on the activities of SEC-lawyers. They state that there are two different kind of lawyers; the ones who want to make quick career and the ones who want to build strong relationships with firms. The first group is part of the revolving door and DeHaan et al. conclude that this group punishes harder than other SEC-lawyers. Because this group wants to make quick career, they have to show their abilities to other firms and therefore they work more aggressively. In this way, this way of working leads to more punishments and to higher financial penalties.

2.2.2 Costs of punishments

When firms are considering of committing financial misbehavior, they know there could be a change of getting caught. And when they get caught, there could be some financial punishments. Different researches show how these firms are handling these potential penalties. Karpoff et al. (2008) wrote an article about the costs of getting caught by cooking the books. They took a sample of 585 enforcements actions of the SEC towards companies who had financial

misrepresentations. The main purpose of the article is to find out what the costs for a company are. The authors make a difference in the kind of costs. First, there are legal costs, for example the fines or financial penalties given by the SEC. Second, there are market costs. These costs are not given by any institution but are the result of punishments by the market. Examples are reputational loss or a major decrease in stock price. After their statistical analysis they conclude that punishments from the market are much larger than legal punishments. Stated in dollars, for every dollar that is misstated, a loss of $3,08 is made. This (rounded) number is the combination of legal costs and market costs, where the legal costs only made up to $0,36. The other part, $2,71, is the result of reputational loss. So, when firms are punished by the SEC for financial misbehavior, they shouldn’t look only to the fine given by the SEC. A bigger worry should be the punishments from the market, which are often a way bigger, but also harder to predict. The prediction of the potential costs is uncertain before committing a crime, according to Sah (1991).

Where the benefits are more reliable to measure, the potential costs are considered very uncertain.

Glaeser et al. (1995) claim that expected costs of a crime are for everyone different. Each company or individual face different circumstances and couldn’t possibly assign the same costs to the same crime. Therefore, when estimating the potential economic costs of a company after committing financial misbehavior, it is dependent on the circumstances what costs companies do assign to the potential crime.

2.3 Role of the auditor

In the preferences of the SEC the role of the auditor might play a role. Francis & Yu (2009) claim that under Big4 audit offices, bigger firms deliver better audit quality than smaller firms. Bigger firms have a higher likelihood of issuing going concern reports and these rapports have a higher predicting power. In their research they took the Big 4 firms issuing reports about SEC-registrants. When audit firms provide high quality audits, there should be a smaller likelihood for firms to conduct in misbehavior and committing a violation punished by the SEC. Francis & Yu concluded that larger Big 4 offices perform better audits with higher quality, where the size of the firm is based on the fees they receive for their audits. Larger Big 4 firms have more resources to perform their audits and therefore deliver better quality. In this way, firms who have larger Big 4 offices as auditor get higher quality audits and misstatements and other violations would be discovered in an earlier stage. Therefore, the SEC doesn’t have to come in action for these companies, because their activities are already judged in a proper way. The SEC should therefore more focus on companies with smaller auditors, who perform according to Francis & Yu audits of lower quality.

Dechow et al. (1996) found that firms manipulating earnings are less likely to have an audit committee. In their study towards the motives for earnings manipulation at firms who are subject to SEC enforcement actions, they found that only 58% of the firms committing a violation have an audit committee. Furthermore, they found that 59% of these firms in violation have a Big 6 firm as external auditor. This means that a small majority has committed a violation even though there are audited by a big audit firm. These results are to some extent contradicted to the results of Francis & Yu. Where Francis & Yu stated that larger Big 4 firms provide better audit quality, Dechow et al. found that even though the majority of firms getting punished are audited by a Big 6 firm. A reason for this might be that firms with bigger audit firms are relatively more getting punished by the SEC.

Location of the audit firm is also playing a role according to DeFond et al. (2011). Non-Big 4 firms, are more likely to issue a going concern opinion when they are closer located to a SEC office. When they are further away from the SEC, they feel less pressure from the

institution and are more willing to cooperate with their clients and give a favorable report. They also have less the feeling that they could get caught when the closest SEC office is relative far. On the other side, for Big 4 firms this conclusion can’t be made. DeFond et al. found no

evidence that location of the Big 4 firms is of influence in the amount of going concern opinion. Big 4 firms are more likely to be consequent, and location of their firm doesn’t matter in their audits. At last, they also found that Big 4 firms are better informed about the activities of the SEC. They have better connections and could receive more information about the SEC’s preferences.

2.4 Hypotheses

Based on the previous literature I come to the following hypotheses. First, Feroz et al. (2008) claim that smaller audit firms receive more severe penalties from the SEC. Bigger firms perform better, because they have more resources in the firm. Furthermore, bigger firms are getting more attention from media and stakeholder. They have more pressure from the outside to perform well and the public interest is larger than with smaller firms. Heese (2015) argues that larger firms are less likely to be subject to enforcement actions and also receive fewer comment letters. Especially this second result could indicate that the SEC allocates fewer resources towards larger companies and therefore has a preference for smaller companies. Kedia & Rajpogal (2011) argue that firms with more media attention have a higher concentration of SEC enforcement actions. Larger companies appear more often in the newspapers, which are functioning like a kind of watchdog. Therefore I expect that:

H1: bigger firms are more likely to receive lower penalties.

Second, Kedia & Rajgopal (2011) claim that firms closer located to the SEC are better informed about their activities and therefore are less likely to conduct in financial misbehavior. They stated two main reasons for this outcome. First, the SEC is resource constrained. Due to limited resources, the SEC prefers to investigate firms closer to their own office to reduce travel costs. Second, they conclude that firms closer to the SEC are better informed about the activities of the SEC. They have better insights about the preferences of the SEC and could therefore adjust their own activities.

Coval & Moskowitz (2001) support this conclusion to some extent. They concluded that information has a maximum travel distance. Information has a reach of 100 kilometers, so firms within this range should get information about the activities of the SEC. Firms outside of this 100 kilometers have a disadvantage and are less informed about the actions and preferences of the SEC. They have less the opportunity to adjust their own activities towards the preferences of the SEC and therefore a higher likelihood to commit a violation. Based on this literature, I expect that

H2: firms closer to the AFM are receiving lower penalties.

Third, from Big 4 firms you should expect that they deliver high quality audits. These Big 4 firms are large companies with a lot of employees and resources. Francis & Yu (2009) support this claim and state that between Big 4 companies, larger firms provide better audit quality. In this way misstatements and other violation are detected by the audit firm. When no suspected behavior is found, the SEC has also less incentive to investigate this firm. Therefore, you could expect that SEC will investigate relatively more firms with who have smaller audit firms as external auditor. These smaller firms perform audit of lower quality and are less able to find misstatements and violations. DeFond et al. argue that Big 4 firms are better informed about the activities of the SEC. Due to better connections, they are better able to receive the information about the potential preferences of the SEC.

However, Dechow et al. (1996) found that almost 60% of the firms who are investigated by the SEC have a Big 6 firm as auditor. The Big 6 is comparable to the Big 4; these are the biggest audit firms in their time. Firms who are manipulating their numbers should be detected by their audit firm, before the numbers are put in the annual report. After that the SEC comes in and could decide if it is necessary to investigate this company.

Based on the previous literature, Big 4 audit firms provide higher audit quality. They are better able to detect violations before the SEC does. Therefore I expect that the SEC has a preference for smaller companies, who have smaller audit firms with less audits of high quality. So, the last hypothesis is

3 Data selection

To make a statistical analysis, first the right data had to be selected. A major part of the data is hand collected and a part comes from databases. First, this thesis only focuses on the financial penalties the AFM has given. The AFM has a lot of different enforcement actions, as described in the theoretical part. The AFM is legally obliged to make their given penalties public, while the other punishments are not publicly available. Therefore financial penalties is the best way to uses this measure. Where many databases have SEC enforcement actions registered, no database has a collection of the activities of the AFM. Consequently, these financial penalties have to be hand collected from the AFM register. This register could be received via the website of the AFM and contains, next to all other kind of information, the new articles which the AFM has to make public with the details about the penalties. In this new articles, most of the time the violator is mentioned, just as the amount of the penalty, the legal violation and information about the enforcement process. A disadvantage is that this register goes back to 2007, while the AFM was founded in 2002. So only new articles from 2007 until 2015 can be found in the register and these years are therefore selected in the data sample. From annual reports of the AFM (2002-2015) the actual quantity of penalties is received, but in these reports no mention is made about the punished companies.

Year 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Total

Amount

penalties 13 19 14 19 10 6 20 52 53 36 22 20 31 17 332

However, as mentioned before the AFM-register goes only back to 2007. Due to this fact, penalties given before 2007 are not included in the data selection. So, from a total of penalties of 332, only 257 might be found in the news articles of the AFM. Additionally, the register of the AFM did not have an easy search function on their website and the results were shown in a very proper way. Consequently, 182 financial punishments were collected out the AFM-register.

To come to a final selection, this number was further reduced due to several reasons. First, the AFM punishes also individual persons and these penalties are not included in the data selection. With these punishments, the AFM has concluded that individual persons are more in violation rather than the whole company. Therefore, AFM would not hold a whole company responsible for the actions of an individual and only punishes this man/woman. For example directors of Royal Imtech or ABN Amro received individual penalties for misbehavior or

maladministration. Since databases do not have data about individuals, these punishments couldn’t be used in the selection.

Second, not all news articles did contain a company’s name. Due to privacy or legal actions the name of the company is not always shown in the news article. Companies can make an objection to the AFM when they really think their firm has a big disadvantage when the penalty is made public. However, they have to come up with really persuasive arguments and judges deciding very often to forbid making the company’s name public. In this cases most of the times the privacy of the firm was respected or a big reputational loss was the reason to keep the penalty anonymously. In these news articles only the violation and the amount of the penalty was given, but not the company name so these punishments also fall outside of the selection.

At last, the database didn’t include all the punished companies. Via WRDS and Bureau Van Dijk, the Amadeus database is used, because this database has the most West European companies. Unfortunately, it was not able to get data from all of the punished companies and these are also kept outside of the selection. Some firms aren’t registered in the database or the necessary information was absent. This leaves a total amount of 77 penalties.

After coming to a final company selection, most of the observations come from the Amadeus-website. The variable firm distance is hand collected. Via the Amadeus database the official firm’s addresses were collected. On the website of the AFM the address of the AFM was found (Vijzelgracht 50, Amsterdam). After that via Google Maps the physical distance from the AFM to the company was calculated. It has to be mentioned that the shortest route every time was taken, not the fastest. The other variables were found in Amadeus and exported.

Total amount of penalties 2002-2015 332

Amount penalties given in 2002-2006 (75) .

Penalties 2007-2015 257

Penalties not found in register of the AFM (75) . Total amount penalties received from register 182

Penalties lost due to several reasons (105) . Total amount penalties in data selection 77

Karpoff et al. (2014) did research towards database research in SEC enforcements actions and also found that is very difficult to get acceptable data. They also hand collected a sample of SEC punishments and conclude that there are four factors who play a factor in this kind of researches. Karpoff et al. name the late initial revelation date of the announcements, scope limitations, events that are less subject to financial misstatements and incomplete and partial data omissions. These factors could influence the way in which database research towards the SEC has been done. This study also struggles with some of these factors and especially missing data causes a large decrease in observations. As result of these decreases, a smaller sample is left than wished before and with this number of observations it would be harder to find a result that is really generalizable. Potential conclusions in this study are therefore useful to a certain extent. They could be used for a better understanding of the AFM, but do not cover all of the punishment and are not generalizable for the all of the activities of the AFM.

4 Methodology

4.1 Research model

In order to test the three hypotheses, this study will make use of the next model. AMOUNT PENALTY = ß0 + ß1SIZE + ß2DISTANCE + ß3AUDITOR

To find out if the AFM has preferences, the amount of penalty is taken as dependent variable. In the ideal scenario the likelihood of a firm to get punished was researched, but this is too hard to operationalize. Therefore, this study will focus on the amount of the given penalty and look if the AFM has a preference by giving a different amount of penalties to certain firms. The independent variables will be explained next.

4.2 Variables

4.2.1 Size

Firm size could be an important factor in the enforcement preferences of the AFM. Feroz et al. (2008) are concluding for example smaller audit firms receive more severe penalties from the SEC. In this research total assets is taken as proxy for firm size. Heese (2015) uses different kinds of proxies as firm size, like total assets, total employees and accounting quality. This study uses total assets as measure for firm size. For the major part total assets was included in the database, however not all and these firm are kept outside the regression. But total assets are in this study the best way to express firm size. Total employees was another option like Heese did, but a lot of smaller firms in the observation had one or even no employees. This is the case when for example a Dutch BV only has a director which is not seen as employee. Accounting quality was too difficult to operationalize in this study. So, total assets of the company was used as proxy for firm size. However, due to the big differences in assets the natural logarithm is used in this case. While there are a lot of smaller firms with relatively smaller total assets and there are a few bigger companies, it causes a non-linear distribution of the total assets. This can cause a biased view and that has to be avoided. Thus, in order to bring the total assets closer to each other the logarithm of ln total assets (lnta) was taken.

4.2.2 Distance

Second, it will be researched if distance has an influence on the penalty. According to previous research (Kedia & Rajpogal (2011), Coval & Moskowitz (2001)), distance was an important factor in enforcement preferences. Heese (2015) argues additionally that the area of the firm is

also important, especially during election times. Kedia & Rajpogal (2011) stated that due to limited resources, the SEC had preferences to investigate firms closer to their own office to make less travel costs for example. Further, Coval & Moskowitz (2001) claim that information has a limited travel distance of 100 kilometers. In this study a distance of 80 kilometers is taken. In this range from Amsterdam the other three major cities in The Netherlands are located (Rotterdam, Utrecht and The Hague). This area is also called The Randstad and is considered to be to most important area in the field of economics, politics and culture. Fur this study it is important that in The Randstad the infrastructure is relative better than outside this area. Outside The Randstad there is relatively less highway and more small villages, which decreases the quality of the infrastructure. Therefore, outside The Randstad travel costs can be higher because firms are harder to reach. For this variable is a dummy created in Stata. As mentioned, there is a distinction made in firms closer than 80 kilometers distance to the AFM office and firms outside of this range.

In the study all locations of the punished companies were retrieved from Amadeus. After this the distance from the AFM office in Amsterdam to company was found via Google Maps. All firms who are 80 kilometers or closer located were separated from firms who are located further away with the help of a dummy variable. Also, two penalties for Ernst & Young were kept outside of the research. These penalties were addressed to their English office located in London, which is more than 500 kilometers away from the AFM-office. In contrast, the second furthest distance was 181 kilometers. So, this could be a potential threat as outlier to the reliability of the research model and therefore deleted from the list.

4.2.3 Big 4 auditor

Firms with a Big 4 firm as auditor receives better audit quality. Therefore misstatements and violations are more likely to be detected by the auditor. Francis & Yu (2009) argue that larger Big 4 firms produce better audits of higher quality. Larger firms have more and better resources to produce there audits and have therefore a higher quality. Also, Big 4 firms are better informed about the preferences of the SEC, due to a better network and connections (DeFond et al., 2011). On the other hand, a research from Dechow et al. (1996) concluded that a majority of firms investigated by the SEC has a Big 6 firm as auditor. About 60% received a high quality audit and is still under investigation by the SEC. In this study a dummy is created for the Big 4 auditors. Via Amadeus the auditor of the firms has been searched and added to the data file. For the firms who didn’t have an auditor stated from the database is assumed that these firms had no Big or Mid 4 firm as auditor, but a smaller or no auditor.

5 Results

5.1 Descriptive statistics

In this paragraph the results of the statistical tests will be shown. First, a total sample of 56 exclusive penalties was used in the test. In table 5.1 the distribution of the included penalties through the years can be found. As mentioned before, due to the fact that not all data or companies were available in the database some penalties are unfortunately lost in the process. You can see the difference between the real amounts of penalties given in a year versus the penalties included in the research in table 5.1.

Table 5.1: Distribution of penalties

Year Freq. Real amounts 2007 1 6 2008 1 20 2009 11 52 2010 9 53 2011 11 36 2012 6 22 2013 8 20 2014 6 31 2015 3 17 Total 56 257

The mean penalty given by the SEC to the companies was €190.331,90. The lowest fine was only €500, while Your Finance B.V. received the highest penalty of €2.000.000 for trading without a license. Due to limited data, total assets (ta) had a smaller number of observations, namely 50. The average total assets of the company was €250.000.000. Furthermore, the average distance from the AFM to the punished companies was about 68 kilometers. This is about the range from the office of the AFM in Amsterdam towards The Randstad, the area in which the four largest Dutch cities are located (Amsterdam, Rotterdam, Utrecht, The Hague).

Table 5.2: Variables summarized

Variable Obs Mean Std. Dev. Min Max

penalty 56 190331.9 447769.7 500 2000000

ta 50 2.50e+08 1.02e+09 7511 6.46e+09

km 56 68.28571 54.92001 1 181

Penalty is the amount of the penalty in euro’s. Ta is Total Assets and is also in euro’s. Km stands for kilometers.

When looking at the numbers of the auditor, the most firms do not have a Big- or Mid 4 company. A dummy variable is made to include the auditor in the research model .The Big 4 firms are PwC, EY, KPMG and Deloitte and receive the number 1 in the sample. Firms with the Mid 4 as auditor (Baker Tilly Berk, Mazars, BDO and Grant Thornton) get number 2 and the others who not have a Big or Mid 4 received the number 3 as dummy.

Table 5.3: Frequency of auditors Auditor Freq.

1 7

2 4

3 45

Total 56

From table 5.2 it could be concluded that a majority has a small or no audit firm as external auditor. Only 12,5% has a Big 4 company and even less company used a Mid 4 company as auditor. Depending on firm size, total employees and other factor it is not always obliged in The Netherlands to have an external auditor.

5.2 Regressions

To make up a good research model, different regressions were produced to find the most reliable model. When performing the original model, with the unadjusted variables, it does not lead to a proper model. In table 5.4 you can find the first regression. Although this first model has a reasonable R-squared of 0.2125, the p-values are very high and insignificant for firm size and distance. Only the Big 4 auditor is for 1% significant, but this might come to the low amount of observations with one of the major four firms as auditor.

Table 5.4

Penalty Coef. Std. Err. t

Total assets -.000055 .000074 -0.74

Kilometer -437.6787 1200.574 -0.36

Big4-auditor 680860.4*** 217445.7 3.13

_cons 152183 106809.9 1.42

AMOUNT PENALTY = ß0 + ß1SIZE + ß2DIST + ß3AUD. *** significant at 1%, ** significant at 5%,

*significant at 10%

As mentioned in the methodology part, the natural logarithm for total assets was taken. This because the values of the total assets between the companies are much diversified. To make a more linear relationship the logarithm is taken. Additionally, also the amount of the penalties will be expressed in the logarithm. For the same reason as the assets, in order to bring the values somewhat closer to each other the ln(penalty) is taken.

Table 5.5

Ln(penalty) Coef. Std. Err. t

Ln(total assets) .1695315 .1071507 1.58

Kilometer -.0103935** .0049887 -2.08

Big4-auditor 1.437412 .9955867 1.44

_cons 8.208311 1.530287 5.36

AMOUNT PENALTY = ß0 + ß1SIZE + ß2DIST + ß3AUD. *** significant at 1%, ** significant at 5%,

*significant at 10%

With a R-squared of 0.3615 this model already looks some better, but still only kilometers is significant for 5%. Although the p-values of the other variables are highly decreased and are close to 10%, they are still not significant. Therefore, chosen is to drop one variable and in this case it is the Big 4 variable. For this variable counts that only a small part of the penalties is given to a firm who is client of a Big 4 firm. Therefore it can cause a biased view on the model. So, table 5.6 will show a regression without the auditor variable.

Table 5.6

Ln(penalty) Coef. Std. Err. t

Ln(total assets) .2754926*** .0789661 3.49

Kilometer -.0102567** .005045 -2.03

_cons 6.934695 1.264789 5.48

The last regression shows a proper model to explain the amount of the penalties. Both variables are significant, where total assets is significant for 1% and kilometers at 5%. The R-squared is on contrary a little decreased compared to the previous regression, but is with 0.3325 still

acceptable.

Using the last regression model, where ln of the amount of the penalty is explained by the ln of total assets and the distance in kilometers, a reliable model is designed. Where both the variables are significant for the model, also the R-squared is reasonable. Because the ln is used, you have to transform the coefficients when using the model. According to the last regression a change of 1% in total assets will lead to a change of 0.27% increase in the penalty. Further, a change of one kilometer will decrease the amount of penalties by 1.03%.

Total assets is positively correlated to the amount of penalty. In other words, for each positive change in assets of 1%, the given penalty will increase with 0.27%. This means that firms with higher assets will receive higher penalties from the AFM. This is the opposite of the first hypothesis, where it was expected that larger firms would receive lower penalties. Kedia & Rajpogal (2011) argued more attention from media towards bigger companies caused less high concentration of SEC enforcement action. Also Feroz et al. (2008) found that smaller audit firms receive more severe fines and Heese (2015) argues that larger firms are less likely to be part of enforcement actions of the SEC. This study reveals the opposite, where an increase of assets leads to an increase of the penalty. A reason for this result might be that larger firms are better able to catch up the higher penalty. The AFM itself also declares that when deciding the amount of the penalty, the capacity of the firm is considered (2016). They argue that they take into account the capacity of the firm. If a big multinational has a violation and gets a fine of €1.000.000, a smaller firm might get for the same violation a lower penalty.

For the distance variable, the results from the regression are also contradictory with the hypothesis. The hypothesis predicted that firms closer to the AFM receive lower penalties. Based on literature of Kedia & Rajgopal (2011) and Coval & Moskowitz (2001) it was concluded that firms closer to the SEC are better informed about their activities, also because information has a travel distance of 100 kilometers maximum. Results from the regression are showing different results for the AFM. In this case, each kilometer further away from the AFM office leads to a 1.03% decrease in the amount of penalty. So, how further the firm is located from the AFM, the lower the penalty is they receive. There could be two reasons for this result. First, with the upcoming technologies it is much easier to receive information. Where the research of Coval & Moskowitz was about 15 years ago, a lot of changes happened with cell phones, e-mail and

internet. A lot of information can be easily reached in contrast to some years ago. In this way, firms in Maastricht or Groningen can also get information about the AFM and is it not necessary to be located close to AFM office. Secondly, distances in The Netherlands are much smaller than for example in the United States. From Amsterdam you could almost reach whole the country in about 2,5 hours. Therefore, travel distance plays a much smaller role in The Netherlands.

For the third hypothesis no good results were found. Due to the limited data a too small group of firms with Big 4 auditors remained. Therefore the variable AUDITOR was kept outside the final regression. About the fact that there are not much firms with a Big 4 auditor in the sample group, an ambiguous answer can be given. It could be that there are really less companies with a Big 4 auditor who conduct in misbehavior. Based on the literature of Francis & Yu (2009), they conclude that larger audit firms provide audits of higher quality. Therefore you could expect that they find a major part of the misstatement and misbehavior. In that way, fewer companies with Big 4 auditors will be punished by the AFM, because the auditor already corrected the misstatements. So, therefore more companies who have a smaller auditor will get enforcement actions from the AFM. Another option is that the AFM does have a preference for smaller firms. They possibly might spare the larger companies for some reasons and prefer to go for smaller firms. However, for this claim no prove was found and about the Big 4 auditor no conclusions can be made in this study.

6 Conclusion

6.1 Conclusion

This study researches to what extent the Dutch AFM has some preferences in its enforcement actions. According to previous literature, what mainly has been done in the United States, the SEC has preferences in its actions. Kedia & Rajpogal (2011) found that the SEC has its preferences based on physical location of the firm. Firms closer to the SEC are less likely to conduct in misbehavior and also prior actions in the neighborhood are a factor in committing violations. Correia (2014) found additionally that political contributions could play an important factor for the preferences of the SEC. But where the SEC has been researched, the AFM is relatively undiscovered.

This study looks to the AFM and its preferences. The preferences of the AFM are operationalized by looking at the amount of penalties they give to firms who conduct in

misbehavior. Three hypotheses were created for testing the main question. First, it was expected that larger firms would receive lower penalties. Larger firms have greater resources and get more media attention (Kedia & Rajpogal, 2011) and Feroz et al. (2009) added that smaller audit firms receive more severe penalties. Heese (2015) argues additionally that larger firms are less likely to be a victim of the enforcement actions of the AFM. Second, firms closer to AFM are less likely to receive higher penalties. Firms closer to the institution are better informed about the activities and could therefore adjust their own actions towards the norms and values. At last, firms with a Big 4 firm as auditor receive lower penalties. Larger Big 4 firms perform audits of higher quality and are better able to detect misstatements and violations (Francis & Yu, 2009).

In order to test the three hypotheses, a statistical analysis is performed. A major part of the data is hand collected by registering the financial penalties from the register of the AFM. After that the needed data was collected from the Amadeus website. Due to several reasons a lot of penalties were lost during the collection period. Not all the penalties could be found in the register and not all the data from the needed companies could be retrieved. However, with a total number of 56 observations a regression model was made. Because the Big 4 firm variable puts its mark too much on the model, this variable was deleted from the model. With the

remaining two variables of firm size and distance, a proper model could be created. For firm size total assets of the company was taken as proxy, while distance was operationalized by measuring the kilometers from the company to the AFM office. A separation was made at 80 kilometers and a dummy was created to test this variable.

The main results of the study are first that firm size increases the amount of penalty a firm receives. With each 1% increase in total assets, the amount of penalty increases with 0.27%. Second, firms closer to the AFM are receiving higher penalties. With each kilometer further away from the AFM, firms receiving a 1.03% lower penalty. Both results are contradictory to the stated hypotheses, while the third hypothesis could not be tested in this study. A few reasons can be given to explain these results.

First, the AFM could have a kind of pity for smaller firms. Larger firms have more assets and more possibilities to pay for the penalty. The consequences for a big multinational when it gets a fine of €1 million are much smaller than for a smaller firm with less capital. In this way, the AFM could have a preference by punishing the larger company with a higher penalty than smaller firms. The AFM (2016) itself also declares that they take to capacity of the punished firm into account when determining the amount of the penalty. So, it can happen that two firms, who commit the same violation, do not receive the same amount of penalty. A big multinational can handle a large penalty often more easily than a smaller firm with less capital. About distance, two reasons can be given why the results differ from the hypothesis. Where previous literature mainly was about the information firms received, nowadays information is very easy accessible with the modern technologies. Earlier studies claim that information only could travel for 100 kilometers, but now everyone has access to the internet and could search for information. Secondly, travel costs in The Netherlands are not as high as in the United States. From the AFM office in Amsterdam, almost whole the country can be reached within 2,5 hours. The third hypothesis cannot be answered in this study, since the variable of Big 4 firm was dropped during the regression. However, most of the firms in the sample didn’t have a Big 4 auditor, so in that way you could say that it makes a difference to have a good auditor. But no conclusions can be made on this topic in this study.

Considering the limitation of the small number of observations, two final conclusions can be drawn from this study. First, larger firms receive higher penalties. Each 1% increase in total assets leads to a 0.27% higher penalty. Second, firms further away from the AMF receive lower penalties, where each kilometer decreases the penalty with 1.03%. For the Big 4 auditor variable no conclusion can be made in this study.

6.2 Limitations

The biggest limitation of this study is the total observations of the statistical part. Due to several reasons, the total amount of penalties of the AFM is reduced to a small number of observations.

punished companies were available. In this way, the conclusions that are made in this study are only generalizable to a certain extent. Because of the small sample, some major or minor observations could have a big impact on the empirical part and could therefore influence the model in a way which is not beneficial. Therefore, the conclusions made in this study are only valid to a certain extent and users of this information should keep this in mind.

6.3 Further research

Further research on this topic could be very interesting. While in the United States the SEC has been investigated very extensively, the Dutch AFM is relatively uninvestigated. Especially when more data becomes available, or all of the penalties in the period 2002-2015, further research would really help to understand the AFM better. Furthermore, this thesis only looked at companies; however the AFM also has the right to punish individuals. It could be interesting in which cases the AFM only decide to punish an individual rather than a whole company. What are the guidelines towards these cases and to what extent is an individual person really

responsible for the violation? At last, Correia (2014) investigated to what extent political connections influenced the enforcement preferences of the SEC. In The Netherlands the government has also some influence on the AFM. A topic for further research can be to what extent the Dutch government has influence on the actions of the AFM. By appointing directors and giving the budgets the AFM could indirect make its statement, but is that the point where the influence stops or reaches it further?

References

AFM website (2016), obtained on different dates, via https://www.afm.nl/nl-nl/over-afm AFM annual reports from 2002 – 2015, obtained on 10-06-2016 via

https://www.afm.nl/nl-nl/verslaglegging/jaarverslag-archief

Correia, M. M. (2014). Political connections and SEC enforcement. Journal of Accounting and

Economics, 57(2), 241-262.

Coval, J. & Moskowitz, T. (2001). Geography of investment: informed trading and asset prices.

Journal of Political Economy 109 (4), 811–841.

Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1996). Causes and consequences of earnings manipulation: An analysis of firms subject to enforcement actions by the

sec*. Contemporary accounting research, 13(1), 1-36.

Heese, J. (2015). Government preferences and SEC enforcement. Harvard Business School

Accounting & Management Unit Working Paper, (15-054).

DeHaan, E., Koh, K., Kedia, S., & Rajgopal, S. (2014). Does the revolving door affect the SEC’s enforcement outcomes?. Rock Center for Corporate Governance at Stanford University

Working Paper, (187), 14-14.

DeFond, M. L., Francis, J. R., & Hu, X. (2011). The geography of SEC enforcement and auditor reporting for financially distressed clients. Available at SSRN 1132885.

Feroz, E. H., Park, K. J., & Pastena, V. (2008). The financial and market effects of the SEC's accounting and auditing enforcement releases. Journal of Accounting Research, 29, 107-142. Files, R. (2012). SEC enforcement: Does forthright disclosure and cooperation really matter?

Journal of Accounting and Economics, 53(1), 353-374.

Financieele Dagblad (2015). AFM beboet fors minder in 2015. Obtained 06-01-2015, via http://fd.nl/economie-politiek/1133495/afm-beboet-fors-minder-in-2015

Francis, J. R., & Yu, M. D. (2009). Big 4 office size and audit quality. The Accounting Review, 84(5), 1521-1552.

Glaeser, E. L., Sacerdote, B., & Scheinkman, J. A. (1995). Crime and social interactions (No. w5026). National Bureau of Economic Research.

Karpoff, J. M., Lee, D. S., & Martin, G. S. (2008). The cost to firms of cooking the books. Journal

of Financial and Quantitative Analysis, 43(03), 581-611.

Kedia, S., & Rajgopal, S. (2011). Do the SEC's enforcement preferences affect corporate misconduct?. Journal of Accounting and Economics, 51(3), 259-278.

NRC Handelsblasd, Aantal klokkenluiders blijft toenemen. (11-04-2016) Obtained on 15-06-2016, via http://www.nrc.nl/nieuws/2016/04/11/aantal-klokkenluiders-blijft-toenemen

Sah, R. K. (1991). Social osmosis and patterns of crime: A dynamic economic analysis. Journal of

political Economy, 99(6).

Wet op het financieel toezicht (2006). Article 1:25, obtained on 09-05-2016 via http://maxius.nl/wet-op-het-financieel-toezicht/artikel1:25