• No results found

The mandatory audit rotation rule and its effect on audit fees : based on data from the Brazilian market

N/A
N/A
Protected

Academic year: 2021

Share "The mandatory audit rotation rule and its effect on audit fees : based on data from the Brazilian market"

Copied!
34
0
0

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

Hele tekst

(1)

The mandatory audit rotation rule and its effect on audit fees

Based on data from the Brazilian market

Name: A.M. (Alice) van Helden Student number: 10149481

Supervisor: dr. S.W. (Sanjay) Bissessur Date: 18 January 2018

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

(2)

Statement of Originality

This document is written by Alice van Helden 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.

(3)

Preface

Before you lies the dissertation ‘The mandatory audit rotation rule and its effect on audit fees’, which is based on data from the Brazilian market. It has been written to fulfill the graduation requirements of the Accountancy & Control master program at the University of Amsterdam. I was engaged in research and writing the dissertation from October 2017 to January 2018. Formulating the right research question and collecting the related data was quite laborious, but conducting extensive investigation has allowed me to answer the question that have been identified. My supervisor from the University of Amsterdam, dr. Sanjay Bissessur, was always available and willing to answer my queries. I would like to thank my supervisor for his guidance and support during this process.

I hope you enjoy your reading.

Alice van Helden

(4)

Abstract

This study examines the effect of the mandatory audit firm rotation rule (MFRR) on audit fees in the Brazilian market in which this regulation is required since 1999. The reason for this

requirement were large accounting scandals at Brazilian banks during the 1990s (Martinez & Reis 2010). This study takes into account data from Brazilian listed firms in the period from 2010 to 2016. The aim is to provide insights in the true costs of the MFRR and to make a contribution to the existing literature about this topic since there has not been written a lot about the effect of the MFRR on audit fees. It also provides insights in the consequences of the implementation of the MFRR, like opportunistic pricing and the low-balling effect. Results do not support the hypothesis that audit firms behave opportunistically in the final engagement year (hypothesis 1). Further, the results show that in the year the MFRR takes place, audit fees are significantly lower by 21% (hypothesis 2). The results also show that audit firms ask for a fee premium in second and subsequent years of 2.49% per year (hypothesis 3). In summary, audit firms offer a discount in the first place which will be offset by fee premiums in about eight years.

The findings of this study should be of interest to legislators, such as the European Union (EU) which requires the use of the MFRR for public-interest companies only since 2016. It provides practical information to European audit firms and client firms that are about to encounter a mandatory audit firm rotation.

(5)

Table of contents

1  Introduction 6 

1.1  Study relevance 7 

2  Literature review and hypothesis/theory development 10 

2.1   Legislation 10  2.2   Auditor tenure 11  2.3   MFRR 12  2.4   Audit Fees 14  2.5   Hypotheses 17  3  Research method 18  3.1  Empirical model 18  3.2  Variables 18 

3.2.1  Dependent and independent variables 18 

3.2.2  Dummy variables 19 

3.2.3  Control variables 19 

3.3  Sample- and data selection 21 

3.4  Inherent variables 22 

4  Results 23 

4.1  Analysis 23 

4.1.1  Descriptive statistics 23 

4.1.2  Testing for parametric assumptions 24 

4.1.3  Results regression analysis 25 

4.1.4  Hypothesis testing 27 

5  Conclusion 29 

(6)

1 Introduction

In 2014, the European Union (EU) published a new regulation on statutory audits which became applicable in June 2016. This regulation includes the new EU audit rules which require rotation of statutory auditors. Mandatory audit rotation imposes periodical breaks to audit engagements and is intended to avoid excessively long relationships between the auditor and the client (Cameran et al., 2015). This primarily aims at improving audit quality (Regulation EU No 537, 2014). Amongst researchers and others involved in the field, there is some disunity about the consequences of the new regulation. Some argue that an increase in audit tenure between audit firm and the client firm will negatively influence the independence of this relationship (Ghosh and Moon, 2005). Others state that the tenure of the relation between the audit firm and the client is positively related to earnings quality (Myers et al., 2003).

Most of the research focuses on the effects of mandatory rotation on audit quality. This has largely been the case due to the fact that there has been a lot of attention to rotation of audit firms and the claim that it will improve quality. However, the newly released rule also affects other issues, like audit fees. There are a couple of studies which focus on audit fees in the first years of an audit engagement (Simon and Francis, 1988; Pong and Whittington, 1994). In addition, Collier and Gregory (1996) are the first who did research on the influence of auditor change on audit fees. A general presumption that prevails is that audit fees will decrease due to rotation because of the reciprocally competition between audit firms in order to acquire audit clients. However, Simunic (1980) states that the learning curve effect within firms will decrease which subsequently leads to higher audit fees.

Opponents to mandatory audit firm rotation argue that the potential costs of mandatory rotation exceed its benefits (Hussey and Lan, 2001), and that the likelihood of audit failures might be greater in the initial period of an auditor-client relationship because of lack of auditor knowledge about client specific risks, processes and operations (PricewaterHouseCoopers, 2007).

Overall, it is expected that audit fees will lower due to auditor rotation because of

competitiveness between audit firms. However, on the other hand, auditing new firms is more costly because the auditor is relatively unfamiliar with the client.

To be able to make a proper consideration and to provide a useful conclusion, it is necessary to do research on data in a setting where compulsory rotation has already set in place. For example, rotation of audit firms every nine years was mandatory in Spain from 1988 to 1995 (Carrera et al., 2009). Another comparable example is South Korea in which mandatory audit

(7)

firm rotation was required from 2006 to 2010. This study provides empirical evidence of the economic impact of this policy initiative on audit quality, and the associated implication for audit fees (Kwon et al., 2014). Spain and South Korea, however, have been studied before in relation to the MFRR and audit fees.

This study will focus on the Brazilian market, since the effect of mandatory audit firm rotation on audit fees has not been studied before in this setting. Specially the Western markets are overrepresented in the theoretical literature. The choice for this country will be further substantiated in ‘1.1 Study relevance’. Brazil requires a five-year rotation for listed companies (except banks) extended to a rotation period of ten years if the company has a statutory audit committee (Martinez & Reis, 2010). Brazil has no requirements on audit partner rotation. The research question is as follows:

“What is the impact of mandatory audit firm rotation on audit fees in the Brazilian market?’’

This research question will be elaborated on the basis of a model provided by Cameran et al. (2015). In addition, three hypotheses will be addressed and in order to make these testable, some adjustments have been made to the model. This will be further explained in ‘3.1 Empirical model’.

1.1 Study relevance

Despite the relevance of the effects of the MFRR on audit fees, it has not been widely studied. Therefore, this study contributes to academic literature by providing insights from Brazil. Researchers are able to use these results as a benchmark for other countries around the world. Studies that have examined the effects of the MFRR on audit quality in combination with studies that have examined the effects on audit fees provides standard setters and legislators with valuable information. Next to standard setters and legislators, audit firms as well can derive value from the results in this study. As already stated, from 2016 the MFRR has become applicable in the EU. Audit- and client firms in countries affiliated with the EU can prepare for the upcoming rotation period and thereby consider the findings on audit fees in the MFRR setting. Based on the findings in this study, new pricing techniques may be developed. Besides, companies that are required to perform a statutory audit in the MFRR setting are able to objectively analyze offers by audit firms.

This study shows that implementation of the MFRR has far-reaching consequences and influences behaviors that cause for example opportunistic pricing and the low-balling effect. With this study, potential clients are able to take into account that audit firms initially offer a

(8)

discount to later ask a fee premium to compensate for this discount. Additionally, it will contribute to the existing literature and it can serve as a basis for further research. Ewelt-Knauer et al. (2013) have been written a review of stakeholders perspective about mandatory audit firm rotation. They have reviewed archival research, experimental and survey studies. Numerous archival studies have investigated the effect of audit tenure on audit quality and incur the (potential) effect of rotation from the findings.

Their findings of archival studies are mixed, but overall, there is limited evidence to suggest beneficial effect of mandatory rotation. This paper can be seen as one of the most comprehensive ones about this topic in de field due to the fact that they reviewed around 100 papers from different studies and methodologies. In their paper, however, the effect of mandatory audit firm rotation on audit fees is only discussed briefly. In addition, nothing is stated about the consequences of mandatory audit firm rotation on audit fees in the Brazilian market. This is consistent with most of the literature on this topic which particularly focus on the consequences for audit quality.

Another paper written by Ewelt-Knauer et al. (2012) discusses mandatory audit firm rotation and refer to more than 130 papers about the MFRR and the consequences of it. This paper also shows that most of the studies analyze the effect of mandatory audit firm rotation (and tenure) on audit quality and auditor independence. The effect on audit fees is only discussed in the following settings. From the Italian mandatory audit firm rotation environment, Barton (2002) states that rotations are often used for negotiating lower average costs per hour of audit work (method: narrative). Such price competition and the subsequent downward pressure on audit fees are particularly feared by auditors (KPMG International, 2010; IDW, 2012b; Ernst & Young, 2011).

Evidence from the US on audit fees (method: survey) suggest that initial year audit costs under mandatory audit firm rotation would increase by more than 20% combined auditor selection costs and additional auditor support costs totaling at least 17% of initial year audit fees.

Since 2006 mandatory audit firm rotation is required in South Korea, showing that audit hours as well as audit fees increased, whereas audit quality (measured as abnormal discretionary accruals) remained unchanged or decreased slightly (Kwon, Kim & Simnett, 2010).

From the known literature, it can be concluded that there is not a lot of research done about the effect of the MFRR on audit fees. With regard to the Brazilian market, there has not been a study before that has examined this. This study, along with the corresponding research question, can be seen as new and therefore relevant.

(9)

This paper will continue as follows. Chapter two provides a literature review and the

hypothesis/theory development. Besides, background information will be provided on the topics discussed in this paper. Chapter three describes the research method, model, variables and data sample used to conduct this research. Chapter four describes the tests that have been carried out and the results of the study. Chapter five includes the conclusion and answer to the research question. Finally the references used in this study will be presented.

(10)

2 Literature review and hypothesis/theory development

2.1 Legislation

The new legislation applies to the 28 member states of the EU (and to Liechtenstein, Iceland and Norway). This European rule sets an audit tenure for public-interest entities to a maximum of 10 years. After a 10 year audit tenure, listed firms are required to conduct auditor rotation and these firms may not engage with their previous auditor for a period of 4 years. This period is often called the ‘cooling-down period’. The new law applies to the 28 member states of the EU and to Iceland, Liechtenstein and Norway (Regulation (EU) No 537/2014, 2014). Member states are allowed to reduce this audit tenure.

Next to the EU, the United States is also currently considering the MFRR. The Government Accounting Office (GAO) was assigned to do research on a possible implementation of the MFRR. GAO conclude (2008) that there seems to be a lack of clear evidence regarding the benefits of the MFRR. The chairman of the Public Company Accounting Oversight Board (PCAOB), Doty, suggests rotation as a desirable way to improve auditor independence and objectivity (Doty 2011). However, there is still no law on MFR in the US to this day. Before the EU legislation on MFRR, some countries already accepted it, like Brazil which

implemented the MFRR in 1999. The Brazilian Securities Commission (CVM) issued an article in 1999 (Article 31 of Instruction 308) which states that an auditor should not provide services to a client for more than five years. At least three years must pass before the auditor can have another engagement with this client. Later, in 2011, the CVM softened this rule for companies with a statutory audit committee. From this year on, companies with an audit committee may extend the maximum tenure with the audit firm to ten years. Therefore, after this change, the maximum audit tenure in Brazil is the same as it is for the European audit tenure. For this reason, results of this study can be compared to countries in the EU. The General Accounting Office (2003) states that the primary reason for Brazil to pass this law was to improve the audit supervision after the accounting scandals at Banco Nacional and Banco Economico in the 1990s (Martinez & Reis, 2010).

The EU supports the MFRR (European Commission 2011, p. 3) mainly on the basis of two arguments. First of all, a long auditor tenure can impair the ability to be independent and thus lower audit quality. Second, a periodic rotation of auditors will lead to a greater objectivity, a new perspective and an improvement of audit quality.

The known research about the MFRR is two folded, some researchers are presenting supportive evidence and others providing opposing evidence about the rules standard setters create.

(11)

2.2 Auditor tenure

Independence in the auditor-client relationship is of great importance (Simunic, 1984) and the length of the auditor-client relationship can influence this independence. A situation that can influence the independence of an auditor is the length of the auditor-client relationship, auditor tenure.

Proponents of the MFRR believe that a long audit tenure results in a reduction of auditor independence. However, the arguments given by standard setters and governments are mainly supported by anecdotes. This section shows an overview of existing literature regarding auditor tenure.

As already mentioned, mainly two conflicting views exist regarding audit tenure. On the one hand, it is stated that a longer tenure will increase the chance that auditors will yield to pressures from management and thereby decrease audit quality. On the other hand, auditing firms argue that the quality of auditing partly depends on the company-specific knowledge about the audited company that has been built up over the years. This is known in the literature as the ‘learning curve effect’ and is supported by a lot of researchers (DeAngelo, 1981; Nelson et al, 2002). Johnson et al. (2002) find a lower audit quality during the first engagement years. A study performed by Cameran and Pettinicchio (2011) shows evidence supporting the existence of a learning curve effect.

The first body of literature basically supports a longer audit tenure and the potential benefits of it. It, however, has to be taken into account that most of the following literature is based on auditor tenure effects in a voluntary setting. This obviously is contradicting to the setting used in this study, which is mandatory. The outcome of the literature, however, is interesting and

contributes to more insight in the concept auditor tenure.

An extensive literature review (Lennox, 2012) shows that the general finding in existing literature proves to be that a longer audit tenure does not seem to have a harmful effect on audit quality. Another study (Summer, 1997) shows that auditors are less independent in short tenures than in longer ones. It is argued that this outcome is caused because a shorter audit tenure should undermine the incentives for building up a reputation.

A study by Gomez Aguilar et al. (2002) shows that auditors are most dependent in the first engagement year with the client because they want to recover the investment they made to get familiar with the firms client in the first year. In addition, the auditors show more independence in the following years.

(12)

A study by Nagy et al. (2004) shows a US-setting in which fraudulent financial reporting is more likely if the audit tenure is not more than three years. Geiger and Raghunandan (2002) show a comparable outcome. The literature also focuses on earnings quality (Myers et al. 2003; Chung & Kallapur, 2003) whereby a negative relation is found between discretionary accruals and auditor tenure. Gul et al. (2007) find higher discretionary accruals in the early years of a tenure, at a significant level. Johnson et al. (2002) find a decrease in audit quality during the first engagement year, this is consistent with the described learning curve effect. The existence of this effect is also supported by Cameran and Pettinicchio (2011). The study is about poor quality work in the first years of an audit which results in partner suspensions.

The second body of research shows the potential disadvantages of a longer auditor tenure. There is hardly any literature that uncover the disadvantages of a long auditor tenure. This is coherent with the previous described arguments which have an anecdotal basis. Ghosh and Moon (2005) argue that an increase in audit tenure between the auditor and audit firm negatively influences the independence between these firms. The study of Davis et al. (2009) shows that in the pre-SOX period, during audits with long tenures (>15 years), the auditor showed greater tolerance for earnings management.

As already stated, one should be careful to draw conclusions from this research because they focus on tenure in a voluntary setting instead of a mandatory setting.

According to the PCAOB (2011), the negative effects of these rotation circumstances may exaggerate the empirical results that have been found so far. Besides, there is no clear negative relationship between auditor tenure and audit quality that can be distinguished. Despite the lack of a dominant relation, regulators have decided to implement the MFRR based on the two main arguments stated above. The MFRR concept will be discussed in the following section.

2.3 MFRR

Despite the importance and relevance, the existing literature on the consequences of the MFRR is very broad in nature. This could be due to the fact that the MFRR has not been implemented in many settings. Countries where this has been the case are Brazil, Japan, Singapore, Korea, Italy and Spain. When looking at Italy, the following outcome have been found. The relationship between the MFRR and earnings quality has been mainly studied (Livne & Pettinicchio, 2012; Cameran et al. 2016). Cameran et al. (2016) find that earnings quality is higher in the final appointment term compared to previous appointment terms. Livne and Pettichino (2012) argue that earnings quality does not improve due to the MFRR. Corbella et al. (2015) argue the MFRR

(13)

causes improvement of audit quality for firms audited by non-Big 4 auditors. Cameran et al. (2015) provide a more recent study on the consequences of the MFRR on audit fees. The results of this study show that audit fees are significantly higher than normal. On the opposite, the paid fees to auditors are significantly lower. This discount will be compensated by higher audit fees in the upcoming years.

When looking at Korea, Kim & Yi (2009) argue that, in this particular country, there is

conducted less earnings management following a MFRR. However, here too caution has to be taken into account because results from comparable studies show inconsistencies.

A study from Firth et al. (2012) show a positive relation between mandatory partner rotation and audit quality. Carrera et al. (2009) have studied audit quality in a voluntary setting as well as in a mandatory setting in Spain and found no significant differences. This outcome, however, proves to be incomplete due to the fact that the MFRR in Spain has been withdrawn before the first rotation has taken place.

Results of the different studies are conflicting. For example, a study from Nagy (2005) show a positive relationship between rotation and audit quality and Krishnan et al. (2007) describe opposite findings. The setting in their studies is based on the collapse of Arthur Andersen in 2002 in the US, which is different from a MFRR-setting. Arthur Andersen has been known as one of the Big 5 accounting firms in the US. In 2002, the firm surrendered its licenses to practice as Certified Public Accountants after being found guilty of criminal charges relating to the firm's handling of the auditing of Enron, which had filed for bankruptcy in 2001 (Brown & Dugan, 2002). This ‘collapse-setting’ differs from the MFRR-setting because there is no limit of tenure of an engagement. Besides, the new auditors want to prevent another collapse comparable to Arthur Anderson and therefore put more effort in the audits.

Gietzmann and Senn (2002) provide a study which is based on two audit games. The first game limits a tenure to a particular number of years set by law. In the second game there is no MFRR rule taken into account. The results show that the MFRR can cause unnecessary costs in certain situations. Firms are motivated to keep their reputation if the audit market has matured and developed, rather than conspiring with management to retain the client. Therefore, according to this study, in a matured audit market it is not favorable to implement a MFRR and it will lead to unnecessary costs.

In summary, Cameran et al. (2005) provides a comprehensive literature review on the MFRR. Their study reviews 26 reports from regulators and other representative bodies from across the world and 33 academic studies. Out of the 26 regulator reports, 22 conclude against the MFRR,

(14)

furthermore the majority of the academic research did not support the MFRR as well. Cameran et al. (2015) and Gietzmann and Sen (2002) provided important findings from their research which suggests that the MFRR affect audit fees. Audit fees and related research will be discussed next.

2.4 Audit Fees

Studies on the determination of audit fees have been conducted since 1980 (Simunic, 1980), with an emphasis on English speaking countries. Some of these studies show convergence in the sense that clients’ size and complexity are the main determinants of fees charged (Köhler & Ratzinger-Sakel, 2012). Others indicate that the market pays higher values for large companies in the industry (Palmrose, 1986; Thinggaard & Kiertzner, 2008), perhaps due to the good

reputation of firms and market concentration.

An extensive literature review by Castro et al. (2015) show the determinants of audit fees paid by listed firms on the BM&F (Bovespa). Results show a positive relationship between audit fees and size, complexity of the client and Big 4 firms. Larger clients pay less in the first year of audit. It, however, has to be noticed that this research field is still in its infancy in Brazil and little has been studied about the determinants of fees charged by firms. The Brazilian Accounting Standard (NBC) P1, approved by the CFC Resolution 976/2003, has established general aspects and criteria that the auditor should consider when budgeting fees.

In December 2000, the BM&F (Bovespa) created the corporate governance levels in Brazil, whose purpose, among others, was distinguishing the internal control levels of listed companies. Firms that adhere to corporate governance levels are more profitable and safer for investors (Macedo & Siqueira, 2006). So, they are expected to have higher requirement levels in board of directors’ organization and structure, enabling the use of various levels created by the BM&F as a proxy to indicate companies with better internal controls and observe whether they contribute to reduce fees.

Since the first paper (Simunic, 1980), ‘total assets’ is the most widely used variable to measure company size. Joshi and Al-Bastaki (2000) mention several authors who concluded that client size is the most significant variable to explain audit fees. Some studies show that the Big 4 firms charge premium fees, when compared to smaller firms in the sector (Francis, 1984; Palmrose, 1986; Whisenant et al. 2003; André et al., 2011; Kwon et al., 2014). Waresul et al. (1996) show that premium fees are justified because Big 4 firms have higher quality teams and they apply better procedures, so it is expected that they better identify errors. Other studies notice that

(15)

premium fees paid to Big 4 firms might be related to the fact that the market reacts more favorably when a client chooses a large firm (Nichols & Smith, 1983; Lennox, 1999). Braunbeck (2010) concluded that, in Brazil, Big 4 firms provide higher quality services. Lawrence et al. (2011) analyzed the differences between audit quality of Big 4 firms and non-Big 4 firms, concluding that such a difference is insignificant.

Another important factor is changing the auditor. When a client makes the decision to change the auditor, the new auditor is selected by taking into account firms with better prices and conditions. Köhler and Ratzinger-Sakel (2012) found strong reductions in fees when auditors changed. Deis and Giroux (1996) and Simon and Francis (1988) highlighted that change is associated with significant reductions in fees; this behavior might be justified by firms on the grounds that, to establish new clients, they could charge lower initial values, adjusting them later. This is called ‘low balling’ which will be discussed in this section. DeAngelo (1981) stressed that such behavior is a competitive response.

Simunic (1980) presents a model to identify the process in which audit fees are determined, this is done from a production point of view. The study suggests that audit fees vary with the amount of work that an audit firm performs during the audit. An interesting study is provided by Hay et al. (2006) which provide an overview of audit fee drivers over the last 25 years. In short, they have identified client attributes which affect fees, such as size attributes (e.g. sales) and complexity attributes (e.g. number of subsidiaries). They have also identified auditor attributes which is particularly relevant for this study with regard to auditor tenure. Audit tenure is added as important attribute for this study. Besides, Simunic (1980) controls for this independent variable.

The use of this control variable is justified by referring to the learning curve effect (Simunici, 1980; Francis, 1984). They argue that a longer audit tenure relates negative to the amount of work, and therefore audit fees. In a MFRR-setting, where there takes place a rotation, the learning curve effect disappear. This is due to the fact that the auditor should perform more work because of the loss of the learning curve effect which results in higher audit fees.

At the same time, the concept of low-balling is often seen in the existing literature (Hay et al., 2006). Hay et al. (2006) defines low-balling as follows: ‘audit firms intentionally offering services at a discount in order to win new business, justifying a fee reduction’. The authors use a meta-analysis, including 23 studies in which a distinction is made between studies using an auditor change dummy and an auditor tenure variable. The results shows that audit fees are reduced in the first year after rotation only in a setting where an auditor change dummy is used as proxy.

(16)

This finding is consistent with the concept of low-balling. However, it has to be taken into account that from the 23 studies used, 11 of them provide insignificant results.

Low-balling is also identified on audit engagements as regular practice and DeAngelo (1981) argues that low-balling is present in a setting where auditors can earn future rents on an audit engagement.

The consequences of the MFRR using proprietary data has been studied by Cameran et al. (2015) and the results show three effects on audit fees.

1. The final-year audit fees are higher than normal, which is an indication for opportunistic pricing. Analysis using public fee data shows that the opportunistic pricing has an average of 29%.

2. The fees of the newly signed auditors in the first year are significantly lower due to higher engagement hours in the first year which is an indication for the low-balling effect. 3. The fees in the following years are significantly higher and even exceed the fee discount

given in the first year. This corresponds with the low-balling theory by DeAngelo (1981) which states that auditors rely on future rents on an audit engagement.

The low-balling effect is explained by DeAngelo (1981) as charging fees below the marginal cost of an audit and auditors rely on future rents on an audit engagement. Auditors compensate for their given discounts by setting higher fees in upcoming years. However, the learning curve effect suggests the contrary result. The learning curve effect should result in a decrease in audit fees according with the audit tenure. It is stated that an increase in audit tenure would negatively relate to the amount of work, and thus to audit fees. This, in turn means that when the MFRR is present, the learning curve effects is not present anymore. Therefore, the new auditor should perform more work due to the loss of the learning effect, resulting in an increase in audit fees. Corbella et al. (2015) used data which is publicly available and studied the influence of the MFRR on audit quality and audit fees. A distinction has been made between Big 4 audit firms and non-Big 4 audit firms. The results show that audit fees decrease following a MFRR for firms audited by a Big 4 audit firm.

(17)

2.5 Hypotheses

In summary, the existing literature as described in the above sections makes a distinction between three main effects of the MFRR on audit fees. The above effects one and two are directional, the third effect is non-directional and is therefore stated in the null-form. Combining the described theories above results in the following hypotheses:

 Hypothesis 1: The MFRR increases audit fees in the last engagement year (opportunistic pricing)

 Hypothesis 2: The MFRR decreases audit fees in the first year after rotation (low balling effect)

 Hypothesis 3: The MFRR does not influence audit fees in the second and subsequent years after rotation (combination of low balling and learning curve effect)

This is shown in the following figure, which shows the hypothesis timeline.

(18)

3 Research method

3.1 Empirical model

This study is using a quantitative approach to examine what the effect of MFRR will be on audit fees. In order to test this effect, three linear regression models will be used corresponding with the three hypotheses as stated in paragraph 2.5.

1. LnFEEi,t = β0 + β1PRE-ROTATEi,t + β2LnSIZEi,t+ β3ATRi,t + β4ROAi,t + Firm Fixed Effects + Year Fixed Effects

2. LnFEEi,t = β0 + β1ROTATEi,t + β2LnSIZEi,t+ β3ATRi,t + β4ROAi,t + Firm Fixed Effects + Year Fixed Effects

3. LnFEEi,t = β0 + β1TENUREi,t + β2LnSIZEi,t + β3ATRi,t + β4ROAi,t + Firm Fixed Effects + Year Fixed Effects

The main model which will be used in this study is based on the work of Cameran et al. (2015). In order to make all the hypotheses testable, some adjustments have been made to the model. The control variables that have been chosen are based on the model and are audit fee drivers that were significant in prior studies. The adjustments made to the model and the variables which are used in this study will be described throughout this chapter. The models are controlled for year fixed effects, in order to remove the time series effect. Additionally, all models are controlled for firm fixed effects, in order to remove the firm-specific part that stays the same in time.

3.2 Variables

3.2.1 Dependent and independent variables

The dependent variable is ‘audit fees’ which are fees for services in order to perform a statutory audit, including services that only the independent accountant can perform. To improve the model fit and linearity, the dependent variable is measured as the natural logarithm of audit fees, LnAuditFees which is consistent with prior research. To operationalize this dependent variable, it is put in the model as ‘audit fees for a certain firm in a certain year’. Data about audit fees is obtained from the database Audit Analytics. When the data was not fully available, audit fees have been subtracted from reference forms which companies are required to disclose to the CVM (Regulation of Corporate Disclosure). The independent variable is MFRR.

(19)

3.2.2 Dummy variables

The dummy variables in this study will serve as indicator variables and will take the value 0 or 1 to indicate the absence or presence of the effect that may be expected to shift the outcome. In order to test the hypotheses, the variables ‘pre-rotate’, ‘rotate’ and ‘tenure’ will be taken into account as dummy variables.

The dummy variable ‘pre-rotate’ will take the value 1 if the firm is subject to a rotation one year later (t+1). It is expected that, based on the literature and first hypothesis, firms with a pre-rotate value of 1 have a positive outcome for that variable.

The dummy variable ‘rotate’ will take the value 1 if the firm has rotated the auditor firm in year (t). It is expected that, based on the literature and second hypothesis, this variable will have a negative outcome for a value of 1.

The dummy variable ‘tenure’ will be used to test how auditor tenure influences audit fees after the MFRR. Therefore, the number of years that an audit firm has an engagement with the client is taken into account. Prior research (Johnson et al., 2002) grouped audit tenure in certain periods, however, because of the limited tenure, this is not done in this study. As can be seen in figure 1, hypothesis three takes the second and subsequent years into account after rotation. It is expected that, based on the literature and the third hypothesis, this variable will take the value 1, in the year after rotation. The model will include the need for compensation for the initial discount given in the first year after rotation by using this method, as already described the low balling effect. Besides the variable tenure also includes the learning effect. Because prior research show these two effects are opposing, it is not possible to provide a clear expectation.

3.2.3 Control variables

By now there has been published more than 100 studies about models with regard to audit fees. The model used for this study is mainly based on Cameran et al. (2015) which is based on literature by Hay et al. (2006). The study of Hay et al. (2006) serves therefore as a basis in order to make a selection of audit fee drivers. The results show, using a meta-analysis, that some fee drivers are consistent throughout the years, studies and countries. A selection of the most important variables has been made to make the data collection feasible for this study. The control variables used in this model will be described in this part. Table 1 provides an overview of the variables, including measurement and expected outcome.

(20)

3.2.3.1 Size

The first control variable in the model is the most dominant audit fee driver in existing literature: ‘size’. This variable is measured by total assets. To improve the model fit and linearity, the variable is measured as the natural logarithm of size, LnSize. Hay et al. (2006) shows that LnSize generally explains over 70% of audit fee variation. LnSize is identified by Simunic (1980) as a very significant determinant in his seminal study. Simunic (1980) provides a production view which takes into account the amount of work that an auditor spend on an audit. A large client requires more work for an auditor and therefore it is expected that size is positively related to audit fees.

3.2.3.2 Asset turnover

Next to size, the model will also control for asset turnover which is measured as sales divided by total assets and denoted as ATR (Asset Turnover Ratio). It is expected that ATR is positively related to audit fees.

3.2.3.3 Profitability

The third control variable is ‘profitability’ which is calculated as net income divided by total assets which will be denoted asROA. Hay et al. (2006) show that a lower ROA will most probably lead to a higher risk for the auditor and a higher audit fee. It is therefore suggested that profitability of the client reflects a measure of risk. The auditor may be exposed to a loss when the client is not financially profitable (Simunic, 1980). Results are mixed for including ROA as a control variable in an audit fee model. However, the meta-analysis by Hay et al. (2006) shows an overall negative significant result. Therefore, it is expected ROA will have a negative impact on audit fees.

(21)

Variable Dependent/independent Measurement Expected outcome

Audit fees Dependent Based on data from Audit Analytics or CVM reports

n/a

Pre-rotate Test variable (Dummy) Value 1 if firm is subject to a MFRR (t+1)

+

Rotate Test variable (Dummy) Value 1 if firm rotated audit firm that year (t)

-

Tenure Test variable (Dummy) Amount of years the audit firm is engaged to client

No influence

Size Control variable Based on total assets +

ATR Control variable Based on sales/total assets +

Profitability Control variable Based on ROA -

Table 1: Variables description

3.3 Sample- and data selection

The sample consists of 186 observations of 27 listed firms in Brazil on the Ibovespa (BM&F Bovespa) between 2010 and 2016, and therefore takes seven sample years into account. This period is chosen because Brazil requires mandatory audit fee disclosure since 2009 (de Lima et al., 2015) and not all data on 2017 is available yet. The data that will be used for this study is primarily obtained from the database Audit Analytics. When data is not available, reference files are used which companies are required to file each year to the CVM. In these files, data from audit fees and auditor rotation can be found. The analysis is limited to US data from firms with a registration at the SEC because they have a stock exchange listing in the US. Therefore, this study is limited to Brazilian listed firms which also have a stock exchange listing in the US. On this basis, out of the 193 companies which are listed on the BM&F Bovespa in 2016, a selection is made. This study therefore is based and conducted on a subsample from Brazilian firms. All data has been collected in the Brazilian currency (Raeis) and, when necessary the American currency (US dollars) has been translated in Raeis using the period-end exchange rate given by the Central Bank of Brazil.

As already stated, 193 companies were listed on the BM&F Bovespa in 2016, only firms that have conducted at least one auditor rotation have been taken into account. From these 193 companies, only 27 companies have rotated auditor firms between 2010 and 2016 and have a stock exchange listing in the US. Three firms observations have been excluded, because those

(22)

companies changed auditor more than once a year. Due to multiple switches of accountants in the research period for three companies, most of which happened within one year, the data consisted of data that had double values for some of the years. This lead the data to be useless for the fixed effects analysis. These companies were excluded from the analysis. Therefore, the total of observations is 186 (27 firms * 7 years – 3 double-year rotations). The total sample therefore consists of 27 listed firms, most of them have been audited by Big-4 audit firms.

3.4 Inherent variables

There are identified certain variables in prior research that are inherent to the sample used for this study, for example auditor quality. This variable is mainly measured as dummy variable which takes value 1 if the audit firm is part of one of the Big 4 audit firms. Since the vast majority of firms in this data sample are audited by auditors from a Big 4 firm, this control variable will not be taken into account in this study.

Another potential control variable suggested by prior research is the firm’s form of ownership which could be a driver of audit fees. This variable can be measured using a dummy variable which distinguish between private firms and public firms. Existing literature argue that 9 out of 14 studies identified a positive and significant relation between the form of ownership and audit fees. In these studies, value 1 has been taken for public companies. An explanation for this relationship between audit fees and the form of ownership can be given by potential agency costs for the auditor or firm. This study is based on listed firms on the BM&F Bovespa and therefore, this control variable is excluded.

(23)

4 Results

In this chapter the results of the empirical research are presented. After presenting the data and regression models, the hypotheses are tested.

4.1 Analysis

In the following section descriptive statistics, parametric assumptions and the results of the regression analysis will be explained. Finally, the hypotheses will be tested.

4.1.1 Descriptive statistics

Table 2 provides descriptive statistics on the dependent, test and control variables. All currency amounts in the sample have either been collected in Brazilian Reais (R$) or converted into R$. The mean of LnAuditFees is 14.61. The lowest LnAuditFees charged in a certain firm year observation is 8.923. The highest LnAuditFees charged is 16.98. This shows a strong dispersion in the fees levied. The fees are correlated with the size and complexity of the organization.

With regards to the test variables, the following statistics are reported in Table 2. The dummy variable Pre-rotate has a mean value of 0.183, meaning that in 18.3% of firm-year

observations an auditor rotation will occur in the year after the observation year. The mean value for the dummy Rotate (ROTATE) of 0.113 implies that in 11.3% of the observations, firms rotated auditor in that observation year. The average relationship between a firm and auditor in the sample is, denoted by Tenure, 2.462 years. This variable indicates the length of the auditor relationship after rotation in the sample.

With regard to the control variables, the following statistics are reported in Table 2. The mean of Size (LnTotalAssets) for the firm-years is 23.78, measured by total assets. Size shows a comparative spread like the audit fees.

The mean for ATR (Asset Turnover Ratio) is 0.592, which corresponds with sales being 59,2% of total assets. In this sample, firms have a mean ROA of 2,57%. Furthermore, for each variable the mean, standard deviation, minimum value and maximum value, skewness and kurtosis are given in Table 2. ATR and ROA show a strong kurtosis, indicating a strong

concentration of values around the average. The sample period could help explain these values, as this in the middle of a worldwide financial crisis, that also hit Brazil in that period.

Based on the regression coefficient, calculated with a T-test, it can be determined whether there are significant effects of ATR, ROA, Size, Pre-rotate, Rotate and Tenure on audit fees. An alfa of 0,05 is used.

(24)

(1) (2) (3) (4) (5) (8) (9) (15)

VARIABLES N mean sd min max skewness kurtosis median

Prerotate 186 0.183 0.388 0 1 1.641 3.694 0 LnTotalAssets 180 23.78 1.546 19.26 26.87 0.0495 3.684 23.48 ATR 180 0.592 0.647 0 3.721 3.515 16.17 0.432 ROA 180 0.0257 0.0650 -0.414 0.180 -1.960 15.23 0.0252 ROTATE 186 0.113 0.317 0 1 2.446 6.984 0 Tenure 91 2.462 1.463 1 6 0.729 2.483 2 lnauditfees 186 14.61 1.335 8.923 16.98 -1.288 6.264 14.68 TENURE 185 0.222 0.416 0 1 1.340 2.797 0 Number of comp 26 26 26 26 26 26 26 26

Table 2: Descriptive statistics. Notes:

Prerotate is a dummy variable with a value of “1” that indicates the year before an auditor rotation. LnTotalAssets is the natural logarithm of the value of Total Assets of each company in the sample for each company year. ATR refers to Asset Turnover ratio which is measured as sales divided by total assets. ROA indicates the Return on Assets. This variable was calculated by dividing Net Income by Total Assets for each company year. ROTATE is a dummy variable with a value of “1” for the year in which an auditor rotation takes place. Tenure indicates the length in client relationship for an auditor. Depending on the moment of rotation in the sample this can lie between 1 and 6 years, lnauditfees is the natural logarithm of the audit fees for each company year. TENURE is a dummy value indicating the following year(s) of an auditor client relationship. The dummy gets a value “1” for the first and subsequent years after an auditor rotation.

4.1.2 Testing for parametric assumptions

To make sure the regression model can be used for testing the relationship between rotation, tenure and audit fees the parametric assumptions were tested. Linearity was tested by checking scatter plots of predictor and the dependent variable. A clear linear relationship was determined between the dependent and predictor variables.

Impendence of variables was tested by using a test for multicollinearity (vif) to determine any multicollinearity between predictor variables. None of the predictor variables showed a vif score higher than 1,5 indicating no disturbing multicollinearity. Due to the time series nature of the

(25)

data the fixed effects regression models help correct for the non-independence of the data points for the observations.

Normalcy was tested by checking for usual skewness and kurtosis in the data and apply the Shapiro Wilk test. The data for each company in the dataset is limited to an average of 10 data points. These tests showed that the data was not normally distributed. Due to the fact that the data showed no extreme outliers, no data manipulation was applied.

Equality of variance was tested using the Modified Wald test for fixed effects time series regression. The test showed a significant inequality of variance in the data. Adding robust standard errors in the model helps remedy misinterpretation of the results.

The table below shows the correlation between the different variables.

Ln audit fees ATR ROA Ln size Pre-rotate

Ln audit fees 1.000

ATR -0.0936 1.000

ROA -0.1319 0.1201 1.000

Ln size 0.7685 -0.2684 0.0384 1.000

Pre-rotate -0.1628 -0.0293 0.0566 -0.1246 1.000

Table 3: Correlation table. Notes:

This table shows the bivariate Pearson correlations between the dependent and the non-dichotomous independent values. Significant relationships have indicated with an asterisk. *: p < .10, **: p < .05, ***: p < .01. The table shows an indication of a linear relationship between some of the predictors and the dependent variable. No signs possible multicollinearity are present as none of the correlation coefficients between the predictor variables reach the limit of .80 (Field, 2013).

4.1.3 Results regression analysis

To test the hypotheses a time fixed effects regression model for each of the hypotheses was designed and run in Stata. The following models were tested:

1. LnFEEi,t = β0 + β1PRE-ROTATEi,t + β2LnSIZEi,t+ β3ATRi,t + β4ROAi,t + Firm Fixed Effects + Year Fixed Effects

2. LnFEEi,t = β0 + β1ROTATEi,t + β2LnSIZEi,t+ β3ATRi,t + β4ROAi,t + Firm Fixed Effects + Year Fixed Effects

3. LnFEEi,t = β0 + β1TENUREi,t + β2LnSIZEi,t + β3ATRi,t + β4ROAi,t + Firm Fixed Effects + Year Fixed Effects

(26)

The results of the regression analyses are presented in Table 4. To test the hypotheses, 3 regression analyses were performed. The adjusted R-squares are of these models are 24.1%, 26.9% and 23.7%, which indicates the amount of variation in the audit fees is predicted by these models.

(1) (2) (3)

VARIABLES Model 1 Model 2 Model 3

ATR 0.381 0.336 0.396 (0.287) (0.251) (0.276) ROA 0.0648 0.102 0.0649 (0.355) (0.350) (0.395) LNTotalAssets 0.345** 0.298* 0.353** (0.138) (0.148) (0.141) prerotate -0.0571 (0.0556) ROTATE -0.210** (0.0947) TENURE 0.0249 (0.0824) Constant 6.291* 7.455** 6.076* (3.344) (3.561) (3.403) Observations 177 177 176 R-squared 0.241 0.269 0.237 Number of comp 26 26 26

Company FE YES YES YES

Year FE YES YES YES

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Table 4: Results of regression models. Notes:

This table shows the results of the three models that were tested. Model 1 is the basic model, used to test hypothesis 1. This contains the predictor and control variables as well as Prerotate as the dummy variable indicating the year before rotation. The model shows no influence of a pre-rotation year on the size of the audit fees. Model 2 replaces Prerotate for ROTATE indicating the year of rotation. This is used to determine if the year of rotation influences the amount of audit fees paid. The final model, Model 3 uses TENURE to determine an effect of a longer relationship on the audit fees. As the models substitute predictor variables, a limited effect is noticeable on the predictive value of the model, as the R2 of the models differs marginally.

(27)

4.1.4 Hypothesis testing

In model 1, the test variable is Prerotate. This dummy variable takes the value 1 if firms switch auditor in the year following that firm-year observation. Cameran et al. (2015) found that Italian audit-firms tend to behave opportunistic in pricing their audit services in this situation. The result of that study show that final engagement fees are 7% higher than normal.

In the current regression analysis, these expectations do not hold. First, the expected sign of the variable is positive. The coefficient in the regression analysis is negative. Second, the coefficient is not significant. Hence, in the Brazilian setting there is no evidence for

opportunistic pricing. A potential explanation for this finding is the fact that client firms that rotate auditor may become a client again after 5 or 10 years. If an auditor behaves

opportunistically in the final engagement year, the client firm may see this as a reason not to choose (the firms have to make a choice due to the MFR rule) this audit firm anymore. This process may lead the auditor to not behave opportunistically. Analyzing the control variables, ATR (β = 0.381) and ROA (β=0.0648) are not significant and their signs are not as expected based on prior literature. However, in the sample SIZE (LnTotalAssets) (β = 0.345) is

significant. According to Hay et al (2006), this control variable mainly shows significant positive results but based on their meta-analysis these results may vary from study to study.

H1 is rejected. No statistical significant proof for relationship between the rotation and opportunistic pricing was found.

In model 2, the test variable is ROTATE. This dummy variable takes the value 1 if the firm is switching auditor that year. Based on prior research by Cameran et al. (2015), it is expected that audit firms tend to give a discount to new audit customers in the first year of the engagement. DeAngelo (1981) identifies this behavior as common practice and refers to it as

low-balling. Model 2 shows evidence of low-balling by Brazilian audit firms in the period 2010-2016.

The test variable Rotate shows a coefficient of β = -0.210, which is significant at the 5% level. This outcome is as expected.

Since the dependent variable is the natural logarithm of AUDIT FEE, procedures have to be applied to the coefficients to transform them into numbers of economic value. This procedure calculates the percentage change of AUDIT FEE following a value of 1 for

ROTATE. The AUDIT FEE discount given in the first year after rotation corresponds to 21%. Similar to model 1, the control variable SIZE is significant.

H2 is Accepted. A negative relationship between the rotation and the audit fees, showing proof of a discount given by the new audit firm.

(28)

Model 3 is regressed to test whether TENURE has an effect on AUDIT FEE.

TENURE is expected to have no effect on AUDIT FEE, due to two contradicting findings in prior literature. The learning-curve effect opposes the finding by Cameran et al. (2015), who find that in years after rotation auditors increase fees to earn back their initial discount. The test variable in model 3 has the coefficient β = 0.0249 and is not significant. TENURE shows a positive but non-significant effect on audit fees.

The economic value of this coefficient is derived in the same way as for ROTATE. The coefficient corresponds to a fee premium of 2.49%. This means that an increase in TENURE by one year increases the AUDIT FEE by 2.49%. Relating this to the initial discount of 21%, it takes approximately 8 years for an audit firm to earn back their initial discount given. The same control variables are significant as in the previous models.

H3 is accepted. No significant effect of the rotation on audit fees in the second year and later years indicating no price effects in the development of the new auditor client relationship was found.

Putting the results together, it seems that auditors do not behave opportunistically in their final-engagement year with a client. Audit-firms do tend to low-ball (i.e. giving a discount), on audit fees for their services in the first year of an audit-engagement. This discount is 21%. However, this initial discount is “offset” by fee premiums in later years. The fee premium is 2,49% per year. So, the initial discount is offset after approximately eight years.

(29)

5 Conclusion

The new EU regulation requires mandatory firm rotation of statutory audits. This study has examined the relationship between the MFRR and audit fees. Advocates of this rule suggest that a longer audit tenure disadvantages the auditor independence. By periodically rotating the concerning audit firm, standard setters are trying to improve audit independence and therefore quality. This study provides practical and academic contribution on the effects of the MFRR on audit fees.

This study has used data from 27 Brazilian listed firms in the time-period of 2010 and 2016. Data from a sample of 186 firm-year observations is collected.

From prior literature, three main effects of the MFRR on audit fees are known and

hypothesized. First, it is expected that auditors engage in opportunistic pricing of the audit in the final year before the MFRR. Second, it is expected that in the year of an auditor rotation the audit fees are lower, due to low-balling. Third, it is expected that audit fees are not influenced by an audit firm rotation in second and subsequent years after the MFRR.

Results do not support the hypothesis that audit firms behave opportunistically in the final engagement year (hypothesis 1). Further, the results show that in the year the MFRR takes place, audit fees are significantly lower by 21% (hypothesis 2). The results also show that audit firms ask for a fee premium in second and subsequent years of 2.49% per year (hypothesis 3). In summary, audit firms offer a discount in the first place which will be offset by fee premiums in about eight years.

Despite the relevance of the effects of the MFRR on audit fees, it has not been widely studied. This study contributes to academic literature by providing an insight in the Brazilian setting. The results can be used as a benchmark for other countries around the world who also implement the MFRR. Studies that have examined the effects of the MFRR on audit quality in combination with studies that have examined the effects on audit fees provides standard setters and legislators with valuable information. Next to standard setters and legislators, audit firms as well can obtain value from the results in this study. In 2016 new regulations on the MFRR has become applicable in the EU. Audit- and client firms in countries affiliated with the EU can prepare for the upcoming rotation period and thereby consider the findings on audit fees in the MFRR setting. Based on the findings in this study, new pricing techniques may be developed. Besides, companies that are required to perform a statutory audit with regard to the MFRR are able to analyze in an objective way offers by audit firms. Potential clients are able to take into account that audit firms initially offer a discount to later ask a fee premium to compensate for this discount.

(30)

The findings and conclusions of this study may be generalizable to other countries where an MFRR is applicable, however, some caution has to be taken into account. First of all, the sample consists of 27 firms which may lead to low statistical power due to the size. Secondly, considering the period (2010-2016) in which data was collected, it might be possible that firms of the client were already familiar to the pricing techniques applied by the audit firms. The MFRR is required since 1999 and therefore, regression results in this period may be weaker than in the first years of the concerning rule. Thirdly, this study does not distinct between mandatory audit firm rotation and voluntary audit firm rotation. Due to data limitations, it was not feasible to distinct between those two types of rotations. The possibility therefore exists that certain audit firm rotations in the sample are not due to mandatory requirements but due to a voluntary choice. The fourth limitation has to do with audit markets which are country specific. Firms handle different legal practices in different countries, despite the fact that Big-4 companies operate as a global network. These practices differ among countries and markets.

Due to these four limitations it could be difficult to make a generalization to other countries on the basis of these findings which could impair external validity.

The relevance of the MFRR, also after being applicable in the EU only since 2016, challenges researchers to do research on the true benefits and costs of the MFRR. In a few years, there may be sufficient data available about the consequences of the MFRR in the European setting. This gives the opportunity for researchers and practitioners to do research on the effects of the MFRR on audit fees. Researchers might be able to motivate or discourage the

(31)

6 References

Brown, K. & Dugan, J. (2002). Arthur Anderson’s fall from grace is a sad tale of greed and miscues’. Wall Street Journal.

Cameran, M., Di Vincenzo, D. & Merlotti, E. (2005). The audit firm rotation rule: A review of the literature. Doctoral dissertation. Bocconi University.

Cameran, M., Negri, G., & Pettinicchio, A. (2015). The audit mandatory rotation rule: the state of the art. The journal of financial perspectives, 3 (2). R. Cont. Fin – USP Sao Paulo

Carrera, N., Gomez Aguilar, N., & Ruiz-Barbadillo, E. (2009). Does mandatory audit firm rotation enhance auditor independence? A journal of practice & theory 28 (1): 113-135

Castro, W., Peleias, I, R. & Silva, G. P. (2015). Determinants of audit fees: a study in the companies listed on the BM&FBovespa, Brazil.

Chung, H., & Kallapur, S. (2003). Client importance, non-audit services, and abnormal accruals. The Accounting Review 78(4): 931-955.

Collier, P., & A. Gregory. (1996). Audit committee effectiveness and the audit fee. European

Accounting Review 5 (2): 177-198.

Corbella, S., Florio, C., Gotti, G., & Mastrolia, S. A. (2015). Audit firm rotation, audit fees and audit quality: The experience of Italian public companies. Journal of International Accounting,

Auditing and Taxation 25: 46-66

Davis, L. R., Soo, B. S., & Trompeter, G. M. (2009). Auditor tenure and the ability to meet or beat earnings forecasts. Contemporary Accounting Research 26(2): 517-548.

DeAngelo, L. E. (1981). Auditor size and audit quality. Journal of Accounting and Economics 3: 183-199.

DeAngelo, L. E. (1981). Auditor independence, 'low balling', and disclosure regulation.

(32)

Deis, D. & Giroux, G. (1996). The effect of auditor changes on audit fees, audit hours and audit quality. Journal of Accounting and Public Policy 15(1): 55-76.

Ewelt-Knauer, C., Gold, A. & Pott, C. (2013). Mandatory audit firm rotation: a review of stakeholder perspectives and prior research. European Accounting Review 10 (1): 27-41.

Ewelt-Knauer, C., Gold, A. & Pott, C. (2012). What do we know about mandatory audit firm rotation? ICAS.

Field, A. (2013). Discovering statistics using IBM SPSS Statistics. Sage Publications Ltd.

Firth, M., Rui, O. M., & Wu, X. (2012). How do various form of auditor rotation affect audit quality? Evidence from China. The International Journal of Accounting 47(1): 109-138.

Ghosh, A., & Doocheol Moon. (2005). Auditor tenure and perceptions of audit quality. The

Accounting Review, 80(2): 585-612. Retrieved from http://www.jstor.org/stable/4093070

Gietzmann, M. B., & Sen, P. K. (2002). Improving auditor independence through selective mandatory rotation. International Journal of Auditing 6(2): 183-201.

Gul, F. A., Jaggi, B. L., & Krishnan, G. V. (2007). Auditor independence: evidence on the joint effects of auditor tenure and non-audit fees. Auditing: A Journal of Practice & Theory 26(2): 117-142.

Gomez Aguilar, N. & Ruiz-Barbadillo, E. (2002). Does auditor tenure improve audit quality? Mandatory auditor rotation versus long term auditing. An empirical analysis. Working paper, University of Cadiz Spain.

Hay, D. C., Knechel, W. R., & Wong, N. 2006. Audit fees: A meta-analysis of the effect of supply and demand attributes. Contemporary Accounting Research 23(1): 141-191.

Johnson, V. E., Khurana, I. K., & Reynolds, J. K. (2002). Audit-Firm tenure and the quality of financial reports. Contemporary Accounting Research 19(4): 637-660.

(33)

Joshi, P. L. & Al-Bastaki, H. (2002). Determinants of audit fees: evidence from the companies listed in Bahrain. International Journal of Auditing 4(2): 129-138.

Kim, J. B., & Yi, C. H. (2009). Does auditor designation by the regulatory authority improve audit quality? Evidence from Korea. Journal of Accounting and Public Policy 28(3): 207-230.

Köhler, A. G. & Ratzinger-Sakel, N. V. S. (2012). Audit and non-audit fees in Germany – the impact of audit market characteristics. Schmalenbach Business Review 64: 281-307.

Krishnan, J., Raghunandan, K., & Yang, J. S. (2007). Were former Andersen clients treated more leniently than other clients? Evidence from Going‐Concern Modified Audit Opinions.

Accounting Horizons 21(4): 423-435.

Kwon, S.Y., Lim, Y. & Simnett, R. (2014). The effect of mandatory audit firm rotation on audit quality and audit fees: empirical evidence from the Korean audit market. A journal of practice

& theory 33 (4): 167-196.

Lawrence, A., Minutti-Meza, M. & Zhang, P. (2011). Can Big 4 versus non-Big 4 differences in audit-quality proxies be attributed to client characteristics? The Accounting Review 86(1): 259-286.

Livne, G., & Pettinicchio, A. (2012). Empirical evidence on reporting quality under dual mandatory rotation rule. Working Paper, Bocconi University.

Martinez, A. L., & Reis, G. M. R. (2010). Audit firm rotation and earnings management in Brazil. Working Paper, FUCAPE Business School and Federal University of Bahia.

Myers, J., Myers, L. & Omer, T. (2003) Exploring the term of the auditor: client relationship and the quality of earnings: A case for mandatory auditor rotation? The Accounting Review 78 (3): 779-799.

Nagy, A. L. & Carcello, J. V. (2005). Mandatory audit firm turnover, financial reporting quality, and client bargaining power: The case of Arthur Andersen. Accounting Horizons 19(2): 51-68.

(34)

Nelson, M. W., Elliott, J. A., & Tarpley, L. (2002). Evidence from auditors about managers' and auditors' earnings management decisions. The Accounting Review 77(1): 175-202.

Lennox, C. S. (2012). Auditor tenure and rotation. Working Paper, University of Southern California.

Pong, C.M. & Whittington, G. (1994), The determinants of audit fees: some empirical models. Journal of Business Finance & Accounting, 21: 1071–1095.

doi:10.1111/j.1468-5957.1994.tb00365.x

Regulation (EU) No 537/2014 of the European Parliament and of the Council of 16 April 2014 on specific requirements regarding statutory audit of public-interest entities and repealing Commission Decision 2005/909/EC (OJL 158, 27-5-2014)

Simon, D. & Francis, J. (1988). The effects of auditor change on audit fees: tests of price cutting and price recovery. The Accounting Review 63(2): 255-269.

Simunic, D. (1980). The pricing of audit services: theory and evidence. Journal of Accounting

Research 18(1): 161-190.

Summer, M. (1997). Does mandatory rotation enhance auditor independence? Regensburger

Diskussionsbeiträge zur Wirtschaftswissenschaft 301.

Stuber, W., & Stuber, A. (2011). Brazil: The Brazilian Regulator Softens The Rule Of Rotation Of

Audit Firms. Retrieved from

http://www.mondaq.com/brazil/x/154116/Compliance/The+Brazilian+regulator+softens+th e+rule+of+rotation+of+audit+firms

Thinggaard, F. & Kiertzner, L. (2008). Determinants of audit fees: evidence from a small capital market with a joint audit requirement. International Journal of Auditing 12(2): 141-158.

Referenties

GERELATEERDE DOCUMENTEN

During an internship at Neopost Inc., of 14 weeks, we developed the server component of a software bus, called the XBus, using formal methods during the design, validation and

affordable, reliable, clean, high-quality, safe and benign energy services to support economic and human

Het verschil in effect tussen live en uitgesteld kijken op merkbekendheid en merkreputatie van de sponsor kan mogelijk worden gemedieerd door de mate van betrokkenheid bij

For the right to develop or maintain relationships to be operational, five requirements are identified for LGBT-people and -relationships to be fully accepted and fully enjoy

The data contains the total revenue, the revenue of different product groups, the revenue in cash, the revenue in card and the cash/total payment ratio.. The product groups

This chapter described the running-in of rolling-sliding contacts on macroscopic and microscopic level. 1) On macro-scale, the geometrical change of the contacting

The mega-sporting events taken into account within this paper will be the summer and winter Olympics, the FIFA World Championships, and the UEFA European Championships

All na- tional reports mention cases of good practices, where higher education institutions have thoughtfully considered which external stakeholders are most relevant to them, and