CHAPTER I

Introduction

The auditing profession is quite old. With the passage of time, like others, audit process has also been improved and transformed. Accounting and auditing must carried out simultaneously. To maintain the accounts, a group of employees is required all twelve months of the year. In the same way for audit number of staff is required for the whole year. But now the profession has adopted many techniques that the same job can be done in much lesser time with lesser staff. For example, it was recommended by the Panel of 2000 on audit effectiveness that the auditors in the US will soon be using forensic techniques in every audit (it would increase audit fee).

Under the corporate law, audit of financial statement is mandatory for the public limited companies as well as private limited companies. Every organization needs qualified accountants for their financial matters. The members' directory of the ICAP 2005-06 (data for the year 2007-2008 was not yet published) shows that there are 275 sole proprietorships, 101 firms with five or less partners, 6 firms with more than five but less than 10 partners and 5 firms with more than 10 partners operating in Pakistan. Four of the five large firms and at least two of the six firms with

more than five but less than 10 partners are associates of well known international networks. All these firms employ a total of 850 qualified chartered accountants constituting 23.67% of the total membership.

The importance of quality control over audit and financial reporting has always been a major concern. For this all auditors of the listed companies needs to maintain a satisfactory quality control rating provided by the ICAP. To ensure the high quality performance by the auditors, the International Auditing and Assurance Standard Board (IAASB) of the International Federation of Accountants (IFAC) has issued International Standard on Quality Control ISQC1, a revised International Standard on Auditing ISA 220. In Pakistan, we are still in the process of adopting ISQC1. Every firm conducting audits need to undergo a QCR organized by board in every two and half years. However, under special circumstances earlier review may be asked.

The International Accounting Standards Committee (IASC) was shaped as an independent private-sector in 1973 through an agreement which was made by professional accountancy bodies from countries like Australia, Canada, France, Germany, Japan, Mexico, the Netherlands, the United Kingdom and Ireland and the United States of America. Ten years later in 1983, the IASC's members included all the professional members of the International Federation of Accountants (IFAC). At the time when the Board of IASC was dissolved in 2001, there were 153 members in 112 countries. The IASC was founded to originate and circulate, in the public interest, International Accounting Standards (IAS) to be observed in the presentation of published financial statements and to encourage their worldwide recognition and execution. It was visualized that IAS would be capable of worldwide acceptance and contribute to a significant improvement in the quality and comparability of financial statements. Over the years the role of the IASC in standard setting has changed dramatically in terms of its influence and a much wider acceptability by entities all over the world. In March 2001, the IASC Foundation, being the parent entity of International Accounting Standards Board (IASB) was formed as a non-profit organization. The IASB, on 1 April 2001, assumed accounting standard setting responsibilities from the IASC Board, which was dissolved as stated above. Although the IAS in themselves does not the force of law, they have become internationally accepted and are applied by a growing number of companies. More importantly, a great number of countries, including Pakistan, have recognized the IAS as an acceptable financial reporting framework in their legislation for preparing primary financial statements.

An issue relating to Audit committee needs attention now-a-days. Audit committees are an integral part of code. The code is mandatory for listed companies but it is observed that a large number of institutions and corporate entities that are under no obligation to follow the said code are also implementing it this, of course, is a very healthy practice which needs further encouragement. The audit committee looks at the internal audit reports and other related activities.

The International Auditing and Assurance Standards Board (IAASB) have completed its 100th meeting in its 29 year history in July 2007. Since March 1978, the group, originally named the International Auditing Practices Committee (IAPC), has gained growing gratitude for the quality of its standards. Its first meeting was held in 1978. The IAASB has published 32 International Standards on Auditing (ISAs) and approximately 20 International Auditing Practice Statements (IAPSs) and other pronouncements on topics such as quality control. Over 100 countries including Pakistan are now using or in the process of adopting ISAs into their national auditing standards.

The issue regarding the audit services in Pakistan has received little attention. This is important because of the increase in competition in the audit service market. There are big four audit companies operating in Pakistan. Those big four audit firms are: A.F.Ferguson (Price Waterhouse Coopers), M. Yousuf Adil Saleem & Co. (Deloitte), Ford Rhodes Sidat Hyder (Ernst & Young), Taseer Hadi & Co. (KPMG). They have reported revenue of $ 28.2bn, $27.4bn, $24.5bn, $22.7bn respectively for the year 2008. There are approx. 400-500 audit firms operating in Pakistan.

Different studies have been conducted on the audit services market in many developed countries. However countries, like Pakistan, have not been studied regarding market for audit service in detail. The last study was conducted in Pakistan in 1997 (Simon and Taylor). Therefore this research expands the previous studies into the market for audit services in Pakistan.

Due to significant importance in the demand for audit services, variables are analyzed in this study in order to predict the audit fee in the Pakistani audit market. This study will provide the information about the determinants of audit fee in Pakistan which was never provided in detail. This study will also increase the knowledge of the auditor and auditee regarding the audit fee.

This study will determine the relationship between the explanatory variables and the audit fee associated with it. Underlying variables are client size, audit risk, auditee complexity and profitability. Previous researchers (Simunic, 1980; Palmrose, 1986; Rubin, 1988; Simon and Francis, 1988; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Cameran, 2005; Besacier and Schatt, 2006; Carson, Fargher, Simon and Taylor, 2003; Ji-Hong, 2007) have used these variables in their studies. The major results of earlier studies have found that audit fee is positively associated with size, audit risk, audit complexity and profitability. In general, these researches have found that these variables explain a large proportion of audit fees.

The rest of the research is structured as follows: Chapter II provides a review of previous researches on audit fee. Chapter III presents the data and description, containing the sample and scope of data. The proposed regression model and equation is also presented in this section. Chapter IV describes the explanatory variables used in the study and develops the testable hypothesis for the research. Chapter V provides the results of the research, while conclusion and limitations of the study are given in Chapter VI.

CHAPTER II

Literature Review:

Beginning with the Simunic (1980), many studies have been conducted for audit in several countries. The majority of the work is done in the Unites states, the United Kingdom, and Australia. Cameran (2005) discussed in his research, “Limited work has been done in other countries like India (Simon et al., 1986), Singapore (Low et al., 1990), Hong Kong (Simon et al., 1992), Malaysia (Simon et al., 1992), South Africa (Simon, 1995), Bangladesh (Karim & Moizer, 1996), the Netherlands (Langendijk, 1997), Norway (Firth, 1997), Japan (Taylor, 1997), Pakistan (Simon & Taylor, 1997), South Korea (Taylor et al., 1999), Finland (Niemi, 2002), and Nigeria (Taylor & Simon, 2003) and few other countries (Taylor & Simon, 1999; Fargher et al., 2001)”. These studies on the audit fee have developed models that explain the reasons of variation in the audit fees paid by the auditee. As the audit fee is the product of unit price and the quantity of audit services demanded by the auditee, a difference in fees is found which represent either the effect of quantity differences or price differences (Simunic, 1980). Different researchers have tested models differently in order to explain the variables of audit fees.

In 1978, The Commission on Auditors' Responsibilities (Commission on Auditors' Responsibilities, Reports, Conclusions, and Recommendations, New York, 1978)reported that the relationship between audit fees and audit performance exists in the market. On this report many researchers had investigated audit pricing, audit quality, etc. (Simunic, 1980; Palmrose, 1986; Francis and Simon, 1988). They all worked on the private sector. Their studies are discussed below.

Beginning with the Simunic (1980) he provides evidence that the price competition prevails throughout the market for the audits of publicly held companies, irrespective of the Big Eight Firms. Simunic's (1980) study was conducted from a sample survey of publicly held corporation in the United States in 1977. The variables size, complexity, diversification, receivables and inventories, in the control variables for differences in loss exposure found to be significant determinants of audit fees. These variables found successful in explaining the cross-sectional variation in internal audit costs. The control variables for differences in the assessed loss-sharing ratio represents that the Profits found in significant however, another two variables Loss and subject to qualified opinion found significant. Time could not indicate price policies for learning effects in the audit fees. For the variable auditor the hypothesis that price competition cannot be rejected.

The result for Simunic (1980) suggests that the variable auditor was not significantly associated with higher fees. However, Simunic (1980) found number of variables significant, as well as the failure of rejecting the hypothesis that the price competition prevails throughout the market for financial audit services.

After Simunic (1980), Palmrose (1986) studied to find about a big eight auditor price premium in the small auditee segment in the U.S. audit market for publicly traded companies. Palmrose's (1986) major focus was on the audit firm's size and especially on the big eight firms. Hypothesis for Palmrose's (1986) study was that the big eight function as cartel and charge higher than competitively warranted fees. The same hypothesis was tested by Simunic (1980). According to Simunic (1980), the big eight charge lower fees than non-big eight firms. Variables used in this study are size, number of reports, and number of locations, client participation, ownership, report modifications, and client industry. Number of reports includes special reports (opinions) and additional reports. These reports affect the audit fees. Locations explained the complexity of the clients' business. The result indicated that there was a significant association between auditor size and audit fees based on a big eight/non-big eight firms. The industry specialization is found insignificant.

After Simunic (1980) and Palmrose (1986), Francis and Simon (1987) report their findings that a big eight price premium exists to both national firms and local/regional firms. The purpose of Francis and Simon's (1987) study was to find audit pricing in the small-client segment in the U.S. audit market for publicly traded companies. It was important because Simunic (1980) and Palmrose (1986) both did not find price differences between big eight and non-big eight auditors. The explanatory variables used in Francis and Simon's (1987) study were size, audit complexity, audit risk, AUDITOR. Francis and Simon's (1987) study strongly explained that the big eight price premium exists within both second-tier national auditors and local/regional auditors. Francis and Simon's (1987) study was the first study to prove the existence of big eight premium in both subgroups.

The above three studies adopted the private sector for their research. Therefore, their results shows audit pricing only in that sector of the economy. Private sector differs from public sector in both the environment and procedures of audit process. For example, the accounting systems and financial reports' procedures, financial statements, audit risk, are different in both the sectors. Rubin (1988) extended prior research by developing a model of audit fees for municipalities.

The purpose of Rubin's (1988) study was to provide evidence on the determinants of audit fees paid by municipalities to their external auditors. Many variables of Rubin's (1988) study were based on the variables used in the private sector. New variables were also added in the model to give unique results on municipal government firm. The variables used in Rubin (1988) study's model were: auditee size, loss exposure, entity complexity, report complexity, auditor retention and selection. Remaining variables were the current auditor's tenure time, audit production costs, and auditor size. Single year data was selected for the year 1982.

Results for Rubin's (1988) suggested a significant relation of auditee size and audit fee as found in prior studies. Loss exposure measured differently for municipalities found significant in the model. Entity complexity had a positive relation with audit fees. Variable number of report was significant where as opinion modification was not significantly related. Auditor tenure was not significantly related to audit fees. Remaining variable auditor size was not significant with the audit fee's model.

Ward, Elder, Kattelus (1994) extended the previous research on the determinants of municipal audit fees. The study found that audit fee is positively associated with number of audit adjustments. The results were consistent with that time's fee premium. The study adopted the model used by Rubin (1988). Few variables were added in the model which indicated the importance in the municipal environment. Auditor expertise, audit adjustment and audit qualifications were added. Auditor experience and audit adjustment were positively associated with audit fee in the study however the results for audit qualification was consistent with the previous studies (Rubin 1988). The results for revenue, number of significant funds, CAFR reports, audit bid, auditor tenure, Big Six auditor and bond rating were consistent with Rubin's (1988). Finance officer tenure, busy season audits and debt per capita were found insignificant. Few agency and political variables were added in Ward et al. (1994) study namely, REVPER, TAX, MGR, TYPE; TURNOVER, MINPAR. Only TAX, REVPER were significant. Remaining agency and political variables MGR, TYPE, TURNOVER, and MINPAR were insignificant in Ward et al. (1994) study.

Simon and Francis (1988) extended prior research on price cutting. According to Simon and Francis (1988) previous researcher (Simunic, 1980; Palmrose, 1986) failed to provide evidence on price cutting due to very small sample size. Simon and Francis (1988) took a large sample of 214 public companies that change auditors and 226 firms that did not change auditors during the period of six years (1979-1984) to find the existence of price cutting. The objective of Simon and Francis's (1988) study was to determine the price cutting system on initial audit engagements and if it does exists then when return to normal levels. Results for Simon and Francis's (1988) study showed that there was an average discount of 24 percent in the initial year and 15 percent for each next two years. In the fourth year, no significant difference was found from normal levels. The explanatory variables used in Simon and Francis's (1988) study were same used by previous researchers (Simunic, 1980; Palmrose, 1986; Rubin, 1988). These variables are auditee size, audit risk and complexity and auditor size.

Previously researchers have used word brand name for big eight firms. This brand name affects the audit pricing and auditor selection. Brand name reflects the reputation of the auditor and the reputation comes from the quality of services provided by the auditor to their clients. In 1970's it was found that the firms are adopting poor financial practices. To monitor the audit quality and to improve these financial practices federal government issued OMB Circular in 1979 and 1984. On the implication of these rules audit quality became the continual process.

Deis and Girouz (1992) studied on the determinants of audit quality of small and independent CPA firms in Texas for the period 1984-1989. Quality control review (QCR) was analyzed for the measure of quality. Dei and Girouz's (1992) study was primarily based on private sector however; it determined the merits of private sector quality to the public sector. Two issues were discussed in the study: auditor reputation and power conflict. These two were found to be significant factors of audit quality.

Maher and Tiessen (1992) studied the behavior of audit fees during a period when there was a increase in competition in the market for independent audit services. In late 1970's and early 1980's economy downturn was observed because of the federal investigation in the accounting profession. According to Maher and Teissen (1992) the business press (e.g., Journal of Accountancy 1984; The Wall Street Journal 1985, 1987; Work 1985) reported the increase in competition in the market for audit services among public accounting firms, but there was no evidence of decrease in audit fees. Therefore, Maher and Teissen (1992) studied to determine whether real audit fees decreased in 1977 and 1981. Maher and Teissen's (1992) study was done on the publicly traded companies reporting external audit fees in a University of Michigan database.

The results of Maher and Teissen's (1992) study provide evidence that there was an increase in the competition in the market for audit services and audit fees decreased between 1977 and 1981. Three controlled variables were used in the study: auditee size, auditee complexity and risky assets. The variables used in the model were, REV, DIVERS, SUBS, FORGN, INV, and RECV. The results were found to be significant in Maher and Teissen's (1992) study.

Davis, Ricchiute, Trompeter (1993) raised a question in their study, “Audit clients, provided with non audit services will result in knowledge spillovers and audit efficiencies that can produce economic rents for auditors?” Past studies have addressed that non audit services may give rise in knowledge spillovers which would lead in rise of audit fees but they did not provide any direct evidence on this issue.

The objective of Davis et al. (1993) study was to provide direct evidence on this issue which was not provided previously. Public accounting firms were selected as the sample of Davis et al. (1993) study. The variables used were FEES, HRS, RATIOHRS, RATHRS, HRISK, MRISK, TAX, ACC and OTHER. TAX, ACC, OTHER were not significant. Results were consistent with the previous research (Simunic, 1980; Palmrose, 1986) that there is a positive relation between audit and non audit fees.

Poitras, Young, Talib (1995) focused on the determinants of audit fees from the sample of Singaporean and Malaysian firms which were publicly traded on the Stock Exchange of Singapore (SES) for the period 1986-1990. Only limited numbers of variables were analyzed in Poitras et al. (1995) fee model. They were total assets, sales, previous profit performance, current assets and industry classification.

For the first time Beattie, Goodacre, Pratt, Stevenson (2000) worked on the determinants of audit fee in the charity sector. Beattie et al. (2000) model was based on the 210 charity institutions with incoming resources of £27 million. The important variables used in the previous research were size, complexity and audit firm location. Variables used in Beattie et al. (2000) study were auditee size, auditee complexity, audit production costs, non-audit services, audit difficulties and big six firms. Auditee size, auditee complexity, audit production costs, non-audit services were positively associated with audit fees. Audit difficulties were not significant and results a delay. Beattie et al. (2000) also observed an affiliation between audit fees and non-audit services. Model was developed for grant-making and fund-raising charities.

The results for Beattie et al.'s (2000) study showed the complexity for the audit of fund-raising charities. It was observed that lower concentration was given by auditors to the charity sector as compared to private sector. The results for Beattie et al. (2000) showed that big six firms receive higher audit fees than non-big six firms. Fee premium was also observed in Beattie et al. (2000) study. Beattie et al. (2000) study also demonstrated that the rate of the audit fee for the charity is almost the half of the private sector's rate. At that time public confidence on charity sector was very low because of the frauds and poor quality of reporting (Accountancy Age, 1998).

The audit can only provide accountability in charity sector's performance. Under the Companies Act 1985, incorporated charities whose annual gross income exceeds £250,000 or whose total of balance sheet exceeds £1.4 million must perform full statutory audit. Whereas, unincorporated charities in England whose annual income exceeds £250,000 (Charities Act 1993), in Scotland £ 100,000 (Law Reform Ac 1990). Charities below these may required to perform an independent audit or can appoint an accountant.

Joshi and Al-Bastaki (2000) focused on the Middle Eastern Country, Bahrain which was not been studied before in detail. Joshi and Al-Bastaki (2000) examined the audit fee structure in Bahrain for 38 listed companies on Bahrain Stock Exchange (BSE). Variables used by Joshi and Al-Bastaki (2000) were size, audit risk, audit complexity, timing of audit and profitability. According to Joshi and Al-Bastaki (2000) there was lack of information on the determinants of audit fee in developing countries. So Joshi and Al-Bastaki (2000) tried to fill that gap from their research.

The findings for Joshi and Al-Bastaki's (2000) study fully support the previous results on determinants of audit fee that size, audit risk, audit complexity and profitability are the important determinants of audit fee and had a positive association with audit fees. However, timing of audit was not a significant determinant of audit fee. Joshi and Al-Bastaki (2000) also identified further research topics: the importance of non-audit services and stability of audit fees, extend of market concentration, and audit services in unregulated market.

Zhou and Elder (2001) investigated the relationship between audit quality, and earning management in the initial public offering process (IPO). Initial public offering process (IPO) is characterized by information lying between management and investors and between informed and uninformed investors. This information creates a chance for management to connect in earnings management.

Zhou and Elder's (2001) study found that auditors (Big 5 auditors and industry specialist) were related to less earnings management in IPO process. Zhou and Elder's (2001) study demonstrated that the quality is an important constraint on earnings management in IPO process and industry specialization as a measure of quality, also constrains earnings management. Zhou and Elder (2001) used 1,048 observations during the period 1996-1998.

Zhou and Elder (2001) found out that discretionary accruals for IPO firms were lower when Big Five auditors were used. Zhou and Elder (2001) also found that the companies audited by specialist auditors, connect in less earnings management. In Zhou and Elder's (2001) study, they actually examined whether size and industry specialization was linked with lower earnings management in IPO firms. And Zhou and Elder's (2001) study found that both measures were associated with lower discretionary accruals, consistent with high quality auditors constraining earnings management and providing more precise information. In the end Zhou and Elder (2001) concluded that investors should be willing to pay more for issues audited by high quality audit firms.

Pae and Yoo (2001) presented a model in which a firm's holder, an auditor and outside investors deliberately interact. The owner's investment and the auditor's effort, jointly effect the auditor's report on the firm's Income statements. According to Pae and Yoo's (2001) study if the auditor's legal liability is large to the investor, then an efficiency loss arises because of the auditor overinvest efforts and owner underinvested in the internal control system. Improvement can be made by changing auditor's legal liability. Pae and Yoo's (2001) study also interpret the changes in the regulatory environment.

The results for the Pae and Yoo's (2001) study showed that an increase in the auditor's legal liability reduces the owner's investment which motivates the auditor to provide more efforts. The unique feature of Pae and Yoo's (2001) study was that, he proved that the in formativeness of audit reports depends not only on the auditor's efforts but also on the owner's heavy investment.

Simon and Taylor (2002) examined the audit services market in Ireland during 1990-1999. The accounting and auditing practices in Ireland are similar with practices adopted by other countries. Simon and Taylor's (2002) study is the first study to consider the market of audit services in Ireland in detail. Simon and Taylor's (2002) study indicated similarities in Ireland audit services market and other countries studied before like United States, United Kingdom, Australia etc. Previous researchers (Simunic, 1980; Palmrose, 1986, Rubin, 1988) showed that auditee size, audit risk, audit complexity and type of auditor relates to audit fees. Following the previous research, Simon and Taylor (2002) also proved that audit fees are positively related to auditee size, audit risk, audit complexity and type of auditor.

Simon and Taylor's (2002) study made an additional contribution by examining the large audit firm fee premium in audit market. The findings for Simon and Taylor's (2002) study indicates that fee premium was only attributable to only few large accounting firms (Price Waterhouse and Coopers and Lybrand) rather than to all large accounting firms. Simon and Taylor's (2002) study concluded that auditee size, audit risk, audit complexity was found to be important variables in explaining audit services in Ireland. The only difference found in the audit service market in Ireland and other countries is that there was no generalized fee premium observed in the audit service market for Ireland.

According to Cobbin (2002) there were fifty six studies drawn from seventeen countries during the period 1980-2000, on the determinants of audit fees. Cobbin's (2002) study tracked much of his work on the determinants of audit fees which were expanded on seventeen different markets. This approach was adopted to measure the level of consistency of application and results in the literature. Cobbin (2002) found UK, Australia, New Zealand, India and Ireland well placed as research environment because of the compulsory requirement for the disclosure of audit fees related information. Then Cobbin (2002) considered the developing market for Pakistan, Bangladesh, Singapore, Hong Kong, Malaysia and the developed markets for Japan, South Korea, South Africa. Cobbin (2002) gave individual consideration to the Netherlands and Norway.

The main objective of Cobbin's (2002) study was to demonstrate the literature which was spread to a number of international markets beyond US, to highlight the common work with respect to the determinants of audit fee. Many of the studies done on these countries replicated the model of Simunic (1980). Consistency in results were found where the variables were auditee size, complexity and risk with auditor size in measuring the determinants of audit fees regardless of differentiation adopted by different researchers in different markets studied.

Carson et al. (2003) examined the audit fee premium to the Big Six brand in the small and large companies in the audit service market for Australia during 1995-1999. The issue regarding fee premium was important at that time because of increasing competition in the audit service market. Carson et al. (2003) study found evidence that fee premium exist in large auditors in the small auditee market segment but not in large auditee segment. There is no linearity found between audit fees and client size in Carson et al. (2003) research. Industry specialization and other explanatory variables were not used in Carson et al. (2003) study because of the small sample size. Early researchers did not find any evidence of fee premium to large auditors in the large client segment. Carson et al. (2003) reexamined this issue using Big Six firms. But Carson et al. (2003) did not find any evidence of price premium to Big Six auditors in the large auditee segment of the market.

Whisenant, Sankaraguruswamy, Raghunandan (2003) examined the characteristics of clients, auditors and their relationship to determine audit and non-audit fees. Prior studies (Simunic, 1980; Palmrose, 1986, Davis et al., 1993) provide evidence on significant association between audit and non-audit fees. Variables used in Whisenant et al. (2003) study were auditor brand names, relative amounts of audit and non-audit fees, client size, the provision of financial information system services, financial stress and audit opinion and industry.

Whisenant et al. (2003) study found that audit fees and non-audit fees are determined mutually but fail to observe the simultaneous determination of audit and non-audit fees. Whisenant et al. (2003) study concluded that changes occur in trouble terms in audit and non-audit fee pricing models, the fees also change. Thus audit and non-audit fees are indirectly through the parameters determining each fee.

Cameran (2005) examined the determinants of audit fee in the Italian market for audit services. There was no published research on determinants of audit fee in Italian audit market. Cameran (2005) mentioned in his study that Italian audit market is not directly comparable with other countries' audit market where these types of studies have been conducted.

In 2000, Italian Antitrust Authority determined that the association of audit firms (Associazione Italiana Revisori Contabili) and its members had concluded agreements in order to standardize fees and co-ordinate client acquisitions. Associazione Italiana Revisori Contabili's (Assirevi) members include Big Six firms of that time. For this reason the Italian Antitrust Authority condemned the Big Italian audit firms for price fixing. This was the motivation of Cameran's (2005) study.

Following the previous studies, Cameran (2005) found that the size, complexity and risk have an impact on the audit fee paid by auditee in Italy. It was also found by Cameran (2005) that fee premium was only attributable to KPMG. Thus no general fee premium was observed in Italian audit market. Additional variables used in Cameran's (2005) study were audit firm tenure and an auditor size to verify the Big auditors fee premium in Italian audit market.

Results for the Cameran's (2005) study indicate that the Client size, audit complexity and audit risk are significant determinants of audit fees. However, audit tenure is not a significant determinant. Results also showed that the price premium was only attributable to KPMG not with other large firms. This was because of the better reputation than other big firms. Cameran (2005) also mentioned in his study that KPMG had a good reputation but it was not the market leader.

Leventis, Weetman, Caramanis (2005) examined the determinants of audit report lag from the Athens Stock Exchange (ASE) for the year 2000. Athens is a European audit market which was upgraded by International Investment Fund. It was never researched before. The audit report lag is a gap between the end of the financial year and the publication of the audited financial statements. The audit report lag is a critical factor in newly developed markets where the audited financial statements are the only source of information for the investors.

Variables considered in Leventis et al. (2005) study were the auditor type, audit fees, number of remarks in the audit report, and the uncertainty in the that audit report. It was found from the results of Leventis et al. (2005) study that audit report lag and type of auditor, audit fees, number of remarks, presence of extraordinary items and uncertainties, were significant variables in the model. All the companies listed in Athens Stock Exchange were audited under regulatory deadlines.

It was suggested by Leventis et al. (2005) that to regulate more closely the publication of audited company accounts must focus on audit specific issues rather than on audit client characteristics. It was observed by Leventis et al. (2005) that companies who choose the external auditor from Big Eight firms or pay fee premium, had a less audit report lag. Furthermore, companies who reported extraordinary items in their financial statement had a large audit report lag. No relationship was found with nature of ownership, profitability, gearing, complexity, industry type, opinion and a change of auditor in Leventis et al. (2005) study. Audit reporting was associated with fee premium.

Hence, to reduce the audit report gap, companies need to pay fee premiums to the auditor. Leventis et al. (2005) also identified that further research could be done on audit report lag from different approaches.

Thinggaard and Kiertzner (2005) examined the audit fees in European countries outside the UK, companies listed on the Copenhagen Stock Exchange (CSE) for the year 2002. 2002 was the first year of disclosure of audit fees by the companies. Thinggaard and Kiertzner (2005) indicated that having two independent auditors would reduce the audit fees but only for larger companies. Thinggaard and Kiertzner's (2005) study extended the Cobbin's (2002) research which was also on the US, UK and the Australian capital markets.

Variables used in Thinggaard and Kiertzner's (2005) study were client size, complexity and risk. Client size, complexity and audit risk found to be significant in Thinggaard and Kiertzner's (2005) study. However, no results found for price premiums charged by Big Four firms. Finally Thinggaard and Kiertzner's (2005) study found that a positive relationship exists between other fees and audit fees.

Basioudis, Geiger, Papanastasiou (2006) examined the assertion that auditors may act more favorably towards clients who pay higher NAS fees. Basioudis et al. (2006) also examined the audit reports of companies in the UK and the degree of non-audit fees paid by these companies to their auditors. It was observed by Basioudis et al. (2006) that companies with large audit fees, receives a going concern modified (GCM) audit opinion, whereas companies with large non-audit fees, receives GCM audit opinion. Both audit and non-audit fees are positively associated with GCM audit opinion.

The actual purpose of Basioudis et al. (2006) study was to examine the connection between audit and non-audit service (NAS) fees received by an auditor. The Basioudis et al. (2006) found a negative association between receiving a going concern (GC) modified audit report and the degree of NAS fees. Whereas, positive association between audit fees and GC opinions. Additionally, Basioudis et al. (2006) results were consistent with the audit pricing studies (Simunic, 1980; Palmrose, 1986; Francis and Simon, 1987). Result for Basioudis et al. (2006) were also consistent with previous research that found a positive link between audit fees, audit effort and the likelihood of receiving a modified audit opinion (Palmrose, 1986; Francis and Simon, 1987).

Besacier and Schatt (2006) examined the factors influence the audit fees in France. In the beginning of 2000, a variety of accounting frauds were observed in United States (Enron, WorldCom) and in Europe (Parmalat). Following these observations several steps were taken to improve the value of corporate governance, including external audit practices and independence.

United States required adopting the Sarbanes-Oxley law (SOX) with rules and procedures for audit firms through the implementation of the Public Company Accounting Oversight Board, which could reduce self-regulation in the profession of auditing. Following these issues, preventive measures were also taken in France since 2000.

It was mandatory for public companies to publish their audit fees. No data was available on this regard that is why Besacier and Schatt (2006) gave it a beginning. The audit process in France involves two separate auditors for the firms that need to publish consolidated financial statements. From 2003, it was compulsory for firms to disclose their audit fees. The Besacier and Schatt's (2006) study was based on the audit fees paid in the year 2002. This year was selected because many public companies publish their audit fees in 2002. Sample of 127 companies were analyzed. Variables used in Besacier and Schatt's (2006) study were firm size, firm risk and two Big four firms. Results for Besacier and Schatt's (2006) study concluded that audit fees rely on firm size, firm risk and two Big four firms.

Giroux and McLelland (2006) examined the municipal audit fees and then analyzed this using structural equation modeling (SEM). Sample used in Giroux and McLelland's (2006) was large cities for the year 1996. Variables used in Giroux and McLelland's (2006) study were client size, complexity of client operations, financial risk, auditing factors and governance structure. The result for Giroux and McLelland's (2006) study demonstrated that SEM modeling could be used in the context of explaining audit fees.

Carson and Fargher (2007) examined the impact of client size on audit fee premiums in the Australian audit services market. Carson and Fargher (2007) helped us understand the fee premium attributable to auditor characteristics, such as industry specialization with respect to client size. Following variables were used in Carson and Fargher's (2007) study: Client size, fees from large clients, bargaining power, industry specialization.

Results for Carson and Fargher's (2007) study showed that the client characteristics of large clients impact audit fees and the auditors of the largest client were designated the industry specialist. Carson and Fargher (2007) also highlighted that the higher fees received from larger clients in each industry. Carson and Fargher (2007) also found that there was a higher demand for more complex accounting and auditing issues that need to be considered before imposing fee premium to a lack of competitive pressure. Hence, Carson and Fargher (2007) concluded that the fee premium was recognized as industry specialist audit firms in the Australian audit services market. Reason for higher fees paid by large clients was the demand for additional audit services asked by the clients. Fee premium was attributable to industry leaders.

Clatworthy and Peel (2007) examined the determinants of external audit fees from UK companies, for the first time. Quoted sector (Main market, the Alternative Investment Market and Ofex) and unquoted market (public and private companies) were analyzed. Variables considered in Clatworthy and Peel's (2007) study was firm size, audit risk and complexity.

Oxera (2006), a recent report: Competition and Choice in the UK Audit Market, governed by the Department of Trade and Industry. The Financial Reporting Council highlighted some issues relating to audit competition, audit pricing, Big Four market saturation, etc. Motivation of the Clatworthy and Peel's (2007) study came from these issues. These issues were discussed in Clatworthy and Peel's (2007) study.

Findings for the Clatworthy and Peel's (2007) study provide evidence that corporate form is an significant determinant of audit fees. Firm size, risk and complexity were significant determinants of audit fees. After these, it was observed by Clatworthy and Peel (2007) that the quoted sector was charged with higher audit fees whereas, the unquoted sector were charged with lower audit fees. This difference was observed because of the high quality demanded by the quoted firms from the auditors. A positive association was found between audit and consultancy fees in Clatworthy and Peel's (2007) study.

Ji-hong (2007) examined the audit service pricing in China audit market. According to Ji-hong (2007) researchers commonly used total assets and sales as the proxy of auditee size to check the relationship to audit fees. Variables used in Ji-hong's (2007) study was auditee size, auditee complexity, audit risk, auditor size and audit tenure. In China in 2003, a temporary prescript was issued about how to measure off small and medium size enterprises. This prescript used assets and sales as the proxies of auditee size. Auditee size, auditee complexity and auditor size were important variables in Ji-hong's (2007) study. These variables were found significant. However, audit risk and audit tenure were insignificant.

Choi, Kim, Kim, Zang (2007) examined the size, quality and audit pricing in the U.S audit market for the period 2000-2005. The question which was studied was: how the size of a local office within an audit firm is significant, determining audit quality and audit fees above and outside firm size at the national level and auditor industry leadership at the city level.

Results for the Choi et al. (2007) study showed that office size was significant with the audit quality and audit fees. Choi et al. (2007) study also found that the Big Four firms provide higher quality to their clients than small audit firms. Choi et al. (2007) concluded that office size is one of the important determinants of audit quality and audit pricing. It was suggested by Choi et al. (2007) that Big Four firms should implement strategies to provide audit services to smaller firms too because smaller auditor can damage the reputation of the firm.

Griffin, Lont, Sun (2008) examined the audit and non-audit fees for New Zealand audit market for the period 2003-2006. Griffin et al. (2008) examined that whether New Zealand audit fees have increased in recent years and whether New Zealand non-audit fees have decreased.

According to Griffin et al. (2008) there was a strict rule was imposed by the government and the adoption of International Financial Reporting Standards (IFRS) by the companies in 2005. Question arises that because of these strictness audit fees would increase or non-audit fees will decreased.

It was observed that in response to the adoption of SOX by the US companies, audit fee was increased. But the case of New Zealand was unclear that is why Griffin et al. (2008) made a research on this issue. Griffin et al. (2008) expected an increase in New Zealand's audit fees. Variables used in Griffin et al. (2008) study were client size, complexity and risk. These variables were significant.

Results for Griffin et al. (2008) research indicated that the audit fees did not change in the period of 2003-2004. A change was observed in later years. Significant increase in audit fees and significant decrease in non-audit fees was observed in Big Four firms by Griffin et al. (2008). This change in fees reflected the efforts, time and quality provided by Big Four firms to their clients.

Al-Shammari, Al-Yaqout, Al-Hussaini (2008) examined the importance of audit fee in the Kuwait market for audit services. In Kuwait, the audit market was never studied before. Variables used in Al-Shammari et al.,'s (2008) study were client size, risk, complexity, profitability and auditor. Similarities as well as differences were found by Al-Shammari et al., (2008) in the Kuwait audit market and other countries studied before. No fee premium was observed in Kuwait audit market for Big Four auditing firms. Companies listed with Kuwait Stock Exchange (KSE) need to appoint two separate auditors to audit their company's accounts. Audit fees had a positive association with size and complexity in Al-Shammari et al.'s (2008) study. Other variables risk, profitability and auditor were insignificant in Al-Shammari et al.,'s (2008) study.

CHAPTER III

Research Method:

The target population selected for this study regarding audit fee are all non-banking sectors listed at Karachi Stock Exchange (KSE). The population selected for this study is consisted of more than 400 companies listed at the Karachi Stock Exchange (KSE). Companies were selected on the basis of similar calculation methods; methods considered were Inventory Calculation Method and Depreciation Method. On the basis of these methods, the initial sample was 103 companies listed at Karachi Stock Exchange (KSE). As stated earlier, companies who provide financial services were not considered in this study. The data was analyzed for the year 2008. Audit fees and other company-specific information were collected from annual reports of respective companies.

Collecting data for those companies was bit difficult. There were 47 companies whose data was missing. Therefore, those 47 companies were excluded from the study. The details for missing data are as follows: There were 6 companies whose amount for long-term debt from the balance sheet was missing. The amount for long-term debt was required for the measurement of audit risk. Therefore, those 6 companies were excluded from the study. After excluding those 6 companies, there were 41 companies whose overall data was not available. Balance sheet and profit and loss summary was missing for the required year. Therefore, those 41 companies were also excluded from the study. After excluded the companies whose data was missing, researcher came up with 56 companies.

The final sample used in this study consists of 56 companies listed at Karachi Stock Exchange (KSE). As audit fee is required to be disclosed by the companies in financial statement in Pakistan, it was not necessary to collect data for the audit fees by means of questionnaires. Audit fee paid annually were considered in this study.

The basic research approach for the study will be regression model of audit fees, as used in the majority of previous studies (Simunic, 1980; Palmrose, 1986; Simon and Francis, 1988; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Cameran, 2005; Besacier and Schatt, 2006) of audit fees. Audit fee was taken as a dependent variable which was explained by various explanatory variables. Those explanatory variables were client size, audit risk, audit complexity and profitability. These explanatory variables are important variable in determining audit fees. The regression model applied in this study is as follow:

AUDEE= b0+b1 CSIZE+b2 AUDRISK+b3AUDCOMP+PROFIT+e

AUDEE= Audit Fees

CSIZE=Client Size (Total Assets)

AUDRISK= Audit Risk (Long-term Debt Ratio)

AUDCOMP= Audit Complexity ((Receivable turnover+Inventory turnover+Current assets)/3)

ROA= Return on Assets (Profitability)

e= Error term

CHAPTER IV

DEPENDENET VARABLE:

Audit Fee:

Audit is an evaluation of an organization. Audits are performed to ascertain the validity and reliability of information and also provide an assessment of a system's internal control. The purpose of auditing is to verify the accuracy and completeness of the records. After analyzing the statements an auditor express an opinion on the organization under evaluation based on work done. On the other hand auditors charge an audit fee depending upon the client size, risks associated with the firm, complexity, returns and for the time required for the audit. May it be any company, the core requirement of financial audits stands still as it shows the clear picture of financial indicators of the company; it is now the mandatory policy for listed company. Large diversified companies with extensive receivable, inventories, returns, involved in foreign operations pay higher fees.

Audit outputs are not readily observed and audit services are unique to each organization, it is difficult to conclude exactly how a particular variable affects the final audit fee. Therefore, the audit fee model is based upon those factors which were used by the previous researchers (Simunic, 1980; Palmrose, 1986; Simon and Francis, 1988; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Cameran, 2005; Besacier and Schatt, 2006) that having an affect on the total audit fee.

All models used by previous researchers have included independent variables differently. Use of variables was different as well as their measures were also different. Explanatory variables used in this study are client size, audit risk, auditee complexity and profitability. Researcher in this research assumed a logarithm of audit fee in this study. The measures and explanations of explanatory variables are discussed as follows:

INDEPENDENT VARIABLES

Client Size:

Researchers have indicated that the most important variable in explaining the variation in the audit fee is client size. An audit involves the review of accounting and internal control system of the organization's financial transactions. Lot of time and efforts are required by the auditors of big companies to test and analyze the data (Simunic, 1980; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Cameran, 2005). Larger companies would usually involve in more transactions and have larger balance sheets, which need more efforts by the auditor. Therefore, higher audit fees are charged by the auditor for large size (Carson and Fargher, 2003; Ji-Hong, 2007).

It is also found that size of the firm is related with quality as well. So, large size should be consistent with higher quality as well for their better reputation. Therefore, auditee pays higher fees for quality services provided by the auditor. Prior researches (Simunic, 1980; Palmrose, 1986; Francis and Simon, 1987; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Carson and Fargher, 2003; Cameran, 2005; Ji-Hong, 2007) provide evidence that the size is the most vital variable in shaping the audit fees.

Client size is commonly measured by total assets, mainly because the auditing process itself may be influenced by inspection of the company's balance sheet, as found in the previous researches (Simunic, 1980; Palmrose, 1986; Besacier and Schatt, 2006). This explanatory variable is usually expressed in natural logarithmic form. A total sale (Ji-Hong, 2007) and population (Rubin, 1988) is also used in measuring client size.

As suggested in the early researches, we have assumed total assets to measure the client size (CSIZE). Hence in this research, researcher assumes a positive association audit fees and client size (CSIZE) and the first hypothesis developed, is as follows:

H1: Audit fee is positively associated with total assets at year end.

Audit Risk:

Audit risk is also considered an important variable in explaining audit fees. This importance has increased with the raise in the number of lawsuits against auditors' worldwide. Audit risk is an inappropriate opinion provided by the auditor on the financial statements particularly on those financial statements which contains material misstatements (Joshi and Al-Bastaki, 2000). The audit risk is the risk which is often cited in the literature is linked to the company's ability to pay. The auditor charge higher fees when he found the audited company facing difficulties due to risk of non-payment of fees. Companies involved in more financial risks are charged with higher fees because the auditor need to take a higher risk of liability for losses attributable to misrepresentations in the financial statements and extensive testing is involved in auditing the company's accounts.

Researchers used different methods for the measure of audit risk. Few researchers (Simunic, 1980; Francis and Simon, 1987; Maher, Tiessen) considered receivables and inventories as the measure of audit risk. These researchers considered receivables and inventories, both as the most risky items of the balance sheet. However Besacier and Schatt (2006) considered risk categories' evaluation a complex one. Besacier and Schatt (2006) used GROWTH as a measure of risk. Besacier and Schatt (2006) used receivables and inventories risk items of balance sheet, whose evaluation is complex. In Carson et al (2003) study risk is calculated as current ratio, quick ratio, leverage, the return on investment, audit opinion and a loss indicator variable. Ji-hong (2007) represented opinion, current ratio, quick ratio, total-debt-to-total assets, long-term liabilities to total assets, capital stock to total assets, loss, return on equity, EBIT as the risk of listed companies, thus influencing audit service pricing. Cameran (2005) used liquidity ratios as the measure of audit fees. Loss, opinion and deficit were used as the proxies of auditor risk in Simon and Taylor's (2002) study. Debt ratio was used as audit risk by Joshi and Al-Bastaki (2000) and Rubin (1988).

Hence, researcher in this research assumes that a company with higher risks is charged with higher audit fees. The formula for the calculation of audit risk (AUDRISK) is long-term debt divided by total assets. They hypothesis developed for the measurement of audit risk (AUDRISK) is as follows:

H2: Audit feeis positively associated with the auditee's debt ratio

Audit Complexity:

Audit complexity is also considered important variable in determining the audit fees. Audit complexity is explained as the tricky to audit. Level of audit work will increase with the level of auditee complexity. The auditee complexity leads to higher level of audit fees due to more time being needed and to a higher fee per hour (Cameran, 2005).

Few researchers (for example, Besacier and Schatt, 2006) constitutes that the risky balance sheets' evaluation is complex and requires more in depth inspection as well as relatively stronger involvement which requires most experienced and expensive auditors. Some researchers (for example, Simunic, 1980) observed that the service or financial sectors are less complex for the auditing process than the manufacturing sector. Whereas Besacier and Schatt (2006) observed that the high growth firms are complex and risky auditing firms. Besacier and Schatt (2006) and Maher et al. (1992) considered receivables and inventories, the risk categories of balance sheet. Evaluation of receivables and inventories is a bit complex.

Different measures were used to calculate audit complexity. Subsidiaries (Simunic, 1980; Simon and Francis, 1988; Beattie et al., 2000; Carson et al., 2003; Ji-hong, 2007), foreign operations (Simunic, 1980; Simon and Francis, 1988; Joshi and Al-Bastaki, 2000; Beattie et al., 2000; Carson et al., 2003), locations (Palmrose, 1986; Davis et al., 1993; Beattie et al., 2000), receivables and inventories (Cameran, 2005; Ji-hong, 2007), Current assets to total assets (Ji-hong, 2007) were used as the proxies of audit complexity. Simon and Taylor (2002) related subsidiaries, receivables and inventories to audit complexity because he (Simon and Taylor, 2002) considered these items difficult to audit.

After considering all the measures of auditee complexity by different researchers, researcher in this study considered inventories, receivables and current assets, the complex items of balance sheet. The formula for the calculation of receivables is total receivables divided by total assets; an inventory is total inventories divided by total assets; a current asset is current assets divided by total assets. After calculating receivables, inventories and current assets, average is taking for these three for the calculation of audit complexity (AUDCOMP). Therefore, a positive association is expected between audit fee and auditee complexity (AUDCOMP). A third hypothesis is developed which is as follows:

H3: Audit fee is positively associated with auditee complexity

Profitability:

Profitability (ROA) is also an important variable in determining audit fees. Efficient use of assets and other resources results in higher return on assets i-e profitability. Profitable firms pay more audit fees to their auditors in view of the fact that higher profits may require accurate audit testing of the authority for the identification of revenue and expenses which require more audit time.

Only few researchers (Simon and Francis, 1988; Joshi and Al-Bastaki, 2000; Whisenant et al., 2003; Besacier and Schatt, 2006) have used profitability in their studies. All of them except Besacier and Schatt (2006) measured profitability as return on assets (ROA). Besacier and Schatt (2006) measured profitability as return on equity (ROE). Studies carried out by Simunic, 1980; Joshi and Al-Bastaki, 2000; Besacier and Schatt, 2006 found profitability a significant variable, subsequently studies carried out by Simon and Francis, 1988; Whisenant et al. (2003) found it insignificant.

Following the majority of the work done on profitability (PROFIT), researcher in this research used return on assets (ROA) as the measure of profitability (PROFIT). The formula for the calculation of return on assets would be income divided by total assets. The last and fourth hypothesis is as follows:

H4: Audit fees are positively associated with the profitability of the firm.

CHAPTERV

Results

The result of this study presents the basic regression equation that is similar to those used in the previous studies (Simunic, 1980; Palmrose, 1986; Simon and Francis, 1988; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Cameran, 2005; Besacier and Schatt, 2006) of audit services in different countries. The overall results suggest a good linear fit in which a large proportion of the audit fees is explained. Dependent variable, audit fee is taken as the natural log of audit fee.

Correlation is the source of estimating all regression relationships. Correlation measures the interdependence between two variables. According to Hair, Black, Babin, Anderson, Tatham (2006), in regression analysis, the correlation between the independent variable provides the basis for forming the regression variate by estimating regression coefficients for each independent variable that maximize the prediction of the dependent variable.

Hair et al. (2006) further said, as independent variables are added to the variate, however, the calculations must also consider the intercorrelations among independent variables. If the independent variables are correlated, then they share some of their predictive powers. Because we use only the prediction of the overall variate, the shared variance must not be double counted by using just the bivariate correlations.

Hair et al. (2006) explained partial correlation as the correlation of an independent and dependent variable when the effects of the other independent variables have been removed. The partial correlation represents the incremental predictive effect of independent variable from the collective effect of all others and is used to identify independent variables that have the greatest incremental predictive power when a set of independent variables is already in the regression variate.

Table 1 shows the correlations among all four independent variables i-e., client size (CSIZE), audit risk (AUDRISK), audit complexity (AUDCOMP) and profitability (PROFIT) and their correlations with the dependent variable i-e., audit fee (logfee).

Table 1 Correlation

Control Variables

CSIZE

AUDRISK

AUDCOMP

PROFIT

Logfee

CSIZE

Correlation

1.000

0.026

-0.153

-0.256

Significance (2-tailed)

.

0.848

0.264

0.059

AUDRISK

Correlation

0.026

1.000

0.674

0.304

Significance (2-tailed)

0.848

.

0.000

0.024

AUDCOMP

Correlation

-0.153

0.674

1.000

0.761

Significance (2-tailed)

0.264

0.000

.

0.000

PROFIT

Correlation

-0.256

0.304

0.761

1.000

Significance (2-tailed)

0.059

0.024

0.000

.

This correlation matrix reveals that profitability (PROFIT) has the highest correlation with the dependent variable (-.256). Profitability (PROFIT) is sent to the controlling variables to check the partial correlation among other 3 independent variables.

Table 2 Partial Correlation

Control Variables

CSIZE

AUDRISK

AUDCOMP

PROFIT

CSIZE

Correlation

1.000

-0.146

-0.217

Significance (2-tailed)

.

0.288

0.111

AUDRISK

Correlation

-0.146

1.000

0.751

Significance (2-tailed)

0.288

.

0.000

AUDCOMP

Correlation

-0.217

0.751

1.000

Significance (2-tailed)

0.111

0.000

.

In table 2 show the result which indicates that after controlling profitability (PROFIT), audit complexity has the highest correlation with the profitability (-.217). Now Auditee complexity will be sent to controlling variables.

Now partial correlation is performed in table 3 by controlling both profitability (PROFIT) and audit complexity (AUDCOMP) to calculate correlation among remaining variables.

Table 3 Partial Correlation

Control Variables

CSIZE

AUDRISK

PROFIT

CSIZE

Correlation

1.000

0.027

Significance (2-tailed)

.

0.849

AUDRISK

Correlation

0.027

1.000

Significance (2-tailed)

0.849

.

Results for table 3 show that size is a significant variable whereas; audit risk is an insignificant variable. Audit risk will not be sent to controlling variables because this variable is not significant and very lowest correlation is found. If audit risk is added in the equation, it affects the value of adjusted R square. Reduction in the value of adjusted R square is observed after adding audit risk in the equation that is why this variable is excluded as it is not significant.

Regression analysis is a statistical practice that can be used to evaluate the association between a dependent variable and many independent variables. Before regression analysis it is essential to ensure that essential assumptions of regression analysis are not violated. These assumptions are linearity of individual variables, constant variance and normality (Hair et al., 2006).

The backward estimation method of regression is used to analyze the data. This method eliminates the variables that do not contribute significantly i-e audit risk. The summarized results of regression are presented below:

Table 4 Descriptive Statisticsa

Minimum

Maximum

Mean

Std. Deviation

N

Predicted Value

11.61

13.4077

12.224

0.411

56

Residual

-0.9123

0.72354

0.00000

0.362

56

Std. Predicted Value

-2.587

2.88

0.000

1.000

56

Std. Residual

-2.45

1.943

0.000

0.972

56

a. Dependent Variable: logfee

Table 4 shows the descriptive statistics for the audit fee (logfee) used in this study. Audit fee averages to 1, and has a standard deviation greater than the mean i-e. 9.38. Minimum predicted value for audit fee is 11.61 and maximum is 13.4077. On the 95 percent confidence interval the mean value lies between 11.61 and 13.

Table 5 Model Summarya,b

Model

R

R Square

Adjusted R Square

1

0.754

0.568

0.530

a. Predictors: (constant), CSIZE, PROFIT, AUDRISK, AUDCOMP

b. Dependent Variable: logfee

Table 5 show the summary of dependent and independent variables used in this study. Results indicate that the value of R2 is 0.568 when all the independent variables are added. It indicates that 56.8 percent variation in the audit fee is explained by the client size, audit risk, auditee complexity and profitability. Adjusted R2 summarizes all the variables in this model.

Table 6 ANOVAa,b

Model

Sum of Squares

F

Sig.

1

Regression

9.378

16.788

0.000

Residual

7.122

Total

16.500

a. Predictors: (constant), CSIZE, PROFIT, AUDRISK, AUDCOMP

b. Dependent Variable: logfee

Table 6 shows the ANOVA analysis which provides the statistical test for the overall model fit in terms of the F ratio. The total sum of squares (9.378 + 7.122 = 16.500) is the squared error that would occur if we used only the mean to predict the audit fee. After using the values of all variables, this error reduces by 56.84%. At this stage the value of F ratio is 16.788 which is significant.

Table 7 Coefficientsa

Model

Unstandardized Coefficients

t

Sig.

Collinearity Statistics

B

Std. Error

Tolerance

VIF

1

(Constant)

11.997

0.08

150.409

0.000

AUDRISK

-0.002

0.002

-0.806

0.424

0.417

2.399

AUDCOMP

-0.002

0.002

-1.090

0.281

0.209

4.783

PROFIT

0.024

0.008

3.096

0.003

0.359

2.784

CSIZE

6.32E-11

0.000

6.361

0.000

0.952

1.050

a. Dependent Variable: logfee

The regression coefficient (B) in table 7 shows the change in the audit fee measure for each unit change in the independent variables. The values -.002, -.002, .024, and 6.32*10-11 are the REG COEFFICIENT for the independent variables: audit risk, auditee complexity, profitability and client size respectively. The predicted value for each observation is the intercept (11.997) plus the regression coefficients (-.002, -.002, .024, 6.32*10-11) times its value of independent variables (audit fee = 11.997 + [-.002] audit risk + [-.002] auditee complexity + [.024] profitability + [6.32*10-11] client size). Here audit risk (AUDRISK) is an insignificant variable. We will exclude this variable because of its insignificance and analyze results later.

Both Collinearity measures (tolerance and VIF) as shown in table 7, provide an impact of collinearity on the independent variables. The tolerance value explains the amount of variability of an individual independent variable which is unexplained by other independent variables. Small tolerance value denotes high collinearity. In table 7 client size has the highest tolerance value i-e., .952 which means that size is .048 affected by other three independent variables (audit risk, auditee complexity and profitability). Smaller value is of auditee complexity (AUDCOMP) i-e., .209 which means that auditee complexity is .791 affected by other three explanatory variables.

Variance inflation factor (VIF) is the inverse of tolerance value. When VIF is high there is high multicollinearity. VIF for client size is 1.050 which means the standard error is 1.025 when compared to other three explanatory variables. Whereas VIF value for auditee complexity (AUDCOMP) is 4.783 which means standard error is 2.187 when compared to other explanatory variables.

Tolerance value for client size is .953 it means that client is 0.047 affected by other independent variables. Variance inflation factor (VIF) is the inverse of tolerance value. VIF for client size is 1.049 which measures that standard deviation is increased to 1.024.

As discussed earlier audit risk (AUDRISK) is an insignificant variable. We have excluded that variable. Now lets analyze results that what are results after the elimination of audit risk.

Table 8 Model Summarya,b

Model

R

R Square

Adjusted R Square

1

0.750

0.563

0.538

a. Predictors: (Constant), CSIZE, PROFIT, AUDCOMP

b. Dependent Variable: logfee

After the elimination of one variable i-e audit risk (AUDRISK), this is not significant at 0.424. The value of R2 decreased to 0.563 which indicates that almost 56.3 percent of the variation in the audit fee is explained by the client size, auditee complexity and profitability. But the value of adjusted R2 improves after eliminating audit risk. Adjusted R2 summarizes all the variables in this model. In this study increase of adjusted R2 shows that the elimination of audit risk improves the model.

Table 9 ANOVAa,b

Model

Sum of Squares

F

Sig.

1

Regression

9.287

22.318

0.000

Residual

7.213

Total

16.500

a. Predictors: (Constant), CSIZE, PROFIT, AUDCOMP

b. Dependent Variable: logfee

After the elimination of audit risk (AUDRISK) which is not a significant variable in the model, results improved. The total sum of squares (9.287 + 7.213 = 16.500) is the squared error that would occur if we used only the mean to predict the audit fee. After using three significant variables (client size, auditee complexity and profitability) the error reduces by 56.28 percent. This reduction is deemed statistically significant with a F ratio of 22.318 and a significant level of 0.000.

Table 10 Coefficientsa

Model

Unstandardized Coefficients

t

Sig.

Collinearity Statistics

B

Std. Error

Tolerance

VIF

1

(Constant)

11.973

0.074

162.389

0.000

AUDCOMP

-0.004

0.001

-2.541

0.014

0.469

2.130

PROFIT

0.028

0.007

4.027

0.000

0.480

2.084

CSIZE

6.30E-11

0.000

6.363

0.000

0.953

1.049

a. Dependent Variable: logfee

After the elimination of audit risk results were found to be improved. The improved predicted value for each observation is in the intercept (11.973) plus the regression coefficients (-.004, .028, 6.3*10-11) times its value of three significant independent variables (audit fee = 11.973 + [-.004]audit complexity + [.028]profitability + [6.3*10-11]client size). The value of tolerance and VIF also improved after the elimination of audit risk from the model.

The results of Client size are strong. Audit fees are positively associated with the total assets and statistically significant. Therefore, the client size is the major determinant of audit fees. Size is importantly concerned with auditing transactions and the value of the assets. Thus, the hypothesis (H1) is accepted. Prior researches (Simunic, 1980; Palmrose, 1986; Francis and Simon, 1987; Joshi and Al-Bastaki, 2000; Simon and Taylor, 2002; Carson and Fargher, 2007; Cameran, 2005; Ji-Hong, 2007) also provide evidence that the client size is the leading variable in explaining the audit fees and found size a strong variable.

Audit fees are also positively associated with the auditee complexity (receivable, inventories and current assets). This variable is also found significant. Therefore, the hypothesis (H3) is accepted. These findings support previous findings such as Simon and Taylor, 2002; Cameran, 2005; Ji-hong, 2007.

Profitability is another variable which is positively associated with audit fees in Pakistan. Profitability (ROA) found highly significant in Pakistani audit market. Therefore, the hypothesis (H4) is accepted. These findings support the finding of Simunic, 1980; Joshi and Al-Bastaki, 2000; Besacier and Schatt, 2006 and does not support Simon and Francis, 1988; Whisenant et al. 2003.

Audit fees are not positively associated with the Audit risk (debt ratio). Therefore, the hypothesis (H2) is not accepted. Findings for audit risk in this study are not consistent with most of the previous researchers (Joshi and Al-Bastaki, 2000; Rubin, 1988). Thus, the finding proved that the companies in Pakistan have no impact of long term debt on audit fees.

Table 11 Summary

Variables

Measured As

Hypothesis

Results

Client Size

Log of Total assets

H1: Audit fee is positively associated with total assets at year end.

Size is a strong determinant of audit fees. Larger size companies pay higher audit fees.

Accept

Audit Risk

Debt Ratio (long-term debt to TA)

H2: Audit fee is positively associated with the auditee's debt ratio

Higher risk doesn't affect audit fees.

Reject

Audit Complexity

Average of

receivable turnover (receivables/TA),

inventory turnover (inventories/TA) and

current assets turnover (CA/TA)

H3: Audit fee is positively associated with auditee complexity

Companies with receivables, inventories and current assets pay higher fees because of complex operations.

Accept

Profitability

ROA

H4: Audit fees are positively associated with the profitability of the firm.

Highly profitable firms also pay higher audit fees.

Accept

CHAPTER VI

Conclusions and Limitations of the study

This study highlighted the determination of audit fee in Pakistan for 56 listed companies on the Karachi Stock Exchange (KSE). Data was collected from the annual reports of the companies. Four explanatory variables namely: Client size, Audit Risk, Auditee Complexity and Profitability were examined in this study to determine the variation in the audit fees in Pakistan. Regression model was used in this study as used by previous researchers. The findings for this study strong support the previous findings on audit fees that the audit fees are strongly associated with client size, auditee complexity and profitability.

Based on the present study, it is found that in Pakistan, audit fees are significantly associated with Client size (total assets), auditee complexity (receivables, inventories and current assets) and profitability (ROA). These finding fully supports the previous studies. However, the findings also indicate that audit risk (debt ratio) is not a significant variable in Pakistani audit market which is different from other researches done on different countries. The important of non-audit services is recommended for future research.

This research is considered in the light of their limitation. Limitation is that the data is considered for one year only. It is not sure that if the data was analyzed for two or more years, findings world be similar.

In order to generalize the outcome of the study, the same study needs to be conducted over a long period of time. Other variables such as the market share of the audit firm and the economic conditions of the country can be included in the regression model in future research. Some other important issues can be considered in further research such as non-audit services, the extent of market concentration and the audit services in the unregulated market.

Source: Essay UK - http://turkiyegoz.com/free-essays/finance/the-auditing-profession.php


Not what you're looking for?

Search:


About this resource

This Finance essay was submitted to us by a student in order to help you with your studies.


Rating:

Rating  
No ratings yet!

  • Order a custom essay
  • Download this page
  • Print this page
  • Search again

Word count:

This page has approximately words.


Share:


Cite:

If you use part of this page in your own work, you need to provide a citation, as follows:

Essay UK, The auditing profession. Available from: <http://turkiyegoz.com/free-essays/finance/the-auditing-profession.php> [19-12-18].


More information:

If you are the original author of this content and no longer wish to have it published on our website then please click on the link below to request removal:


Essay and dissertation help

badges

 
Blooming Sequence | Venom (1080p)[Subtitulado] | openload.co