Audit and nonaudit fees from different reporting regimes and perceived audit quality.

Parkash, Mohinder ; Singhal, Rajeev ; Zhu, Yun "Ellen" 等

Abstract

Auditor independence in fact as well as in appearance has been one of the concerns of the Securities and Exchange Commission. The debate on perceived audit quality is concentrated on short reporting periods. In this paper we study the issue of auditor independence by analyzing how investors and financial analysts perceive audit quality in the presence of auditor provided nonaudit services under different reporting regimes. We analyze the issue by focusing on three distinct disclosure periods. The first period of our analysis covers the relatively noncontroversial accounting disclosures made between 1978 and 1980. The second period of our analysis includes accounting scandals and perceived audit failures over February 2001 and July 2002. Our third period of analysis includes restrictions on provisions of certain nonaudit services and other audit related radical reforms with the enactment of the Sarbanes-Oxley Act in July 2002.

I. INTRODUCTION

The Securities and Exchange Commission (SEC) has been concerned about auditor provided nonaudit services and its impact on auditor independence for last four. decades. SEC's concern pertains to auditor's independence in fact as well as in appearance (Levitt 2000). The debate on perceived audit quality has focused on how auditor provided nonaudit services are viewed by financial information users affecting their perception of auditor independence and objectivity. Concentrating on short reporting periods, several recent studies have provided insights into this debate. Data used from short disclosure periods do not address an important issue of whether perception of audit quality with auditor provided nonaudit services remains unchanged across different disclosure periods. In this study, we focus on how investors and financial analysts perceive audit quality in the presence of auditor provided nonaudit services under different reporting regimes.

Under ASR No. 250, the SEC required registrants to disclose information about nonaudit fees in relation to audit fees. The disclosure requirement was rescinded with ASR No. 304. This resulted in disclosure of the ratio of nonaudit to audit fees for 1978-1980, which was a period without any controversial accounting issues. Subsequently, in November 2000, the SEC issued Final Rule S7-13-00, which mandated disclosure of nonaudit and audit fees paid to auditors in proxy statements filed after February 5, 2001. This period involved accounting scandals and perceived audit failure resulting in criticism of accounting profession and independence of auditors. Following these controversies, the Sarbanes-Oxley Act of 2002 (SOX) was enacted to impose restrictions on provisions of certain nonaudit services along with other audit related radical reforms. On February 6, 2003 the SEC issued Final Rule S7-49-02, requiring companies listed on U.S. exchanges to disclose audit fees, audit-related fees, tax fees and all other fees for each of the two most recent fiscal years. Given the history of required nonaudit fees disclosures, we identify three distinct periods for our analysis: (i) non-controversial period of 1978-80 under ASR No. 250, (ii) controversial period which extends from pre-SOX time period of February 2001 through July 2002, and (iii) SOX period after enactment of SOX in July 2002.

In requiring these disclosures in different time periods, the SEC's focus has been to provide information to users of financial statements and disclosures to assess impartiality and objectivity of auditors. Under varying time periods and circumstances, the SEC emphasized independence in appearance to justify the mandated disclosure of nonaudit and audit fees to enable the users of financial information to conclude if auditor is exercising objectivity and independence. In forming their assessment about auditor independence and audit quality, users of financial information judge the benefits and sacrifices of joint provision of audit and nonaudit services. Users consider the benefits of knowledge spillovers and economies of scope as well as the sacrifices of possible compromised independence resulting from economic bonding with joint provision of audit and nonaudit services. Users' perception of audit quality with provision of audit and nonaudit services is combined assessment of these benefits and sacrifices. If assessed benefits are higher than sacrifices, the audit quality would increase with higher level of nonaudit services. If benefits are assessed to be lower than sacrifices, the audit quality would have an inverse relation with the magnitude of nonaudit fees.

In the current study, we focus on fee ratio, measured as the ratio of nonaudit to audit fees, to assess perceived audit quality. On yearly basis, the fee ratio decreased insignificantly in non-controversial ASR No. 250 period but decreased in each year of controversial pre-SOX period and in post-SOX era. This indicates that the required disclosure did not have a significant impact on the fee ratio in the non-controversial regime, but helped reduce the ratio significantly during the controversial pre-SOX and post-SOX time periods. The disclosure requirements can have an impact on the procuring patterns of nonaudit fees in relation to audit fees when auditing profession is under scrutiny. Our empirical results also imply that the fee ratio had increased dramatically in scandalous era of pre-SOX since ASR No. 250 time period and that the passage of the SOX resulted in significantly lower fee ratio for (i) each client, (ii) for each audit firm and (iii) for each industry. The increases in fees ratio were consistent with the SEC's concerns.

We use dispersion of analysts' earnings forecast to assess the impact of auditor provided nonaudit services on audit quality. Our results show an inverse relationship between the dispersion of analysts' forecast and fee ratio to show higher audit quality with higher fee ratio during ASR No. 250 period. In contrast, during the pre-SOX controversial period, we observe a positive relationship between analysts' forecast dispersion and fee ratio to signify lower audit quality with higher fee ratio. In post-SOX period, we find an insignificantly negative relationship between analysts' forecast dispersion and fee ratio establishing no relation between audit quality and fee ratio.

Finally, we explore investors' assessment of audit quality in the presence of auditor provided nonaudit fees by examining stock market reaction around filing of proxy statements containing information about the fee ratio, Our empirical analysis shows that there is significantly negative reaction to announcements of fee ratios during controversial pre-SOX period but insignificantly positive reaction during ASR No. 250 and the post-SOX periods. Collectively, our empirical results provide evidence that auditor provided nonaudit fees relative to audit fees are related to (i) increased audit quality in ASR No. 250 reporting period, (ii) decreased audit quality in controversial period of pre-SOX and (iii) no effect on audit quality in post-SOX period. Our results provide some support in favor of enacting SOX.

Our study contributes to the extant literature by making a temporal analysis of the patterns of fee ratio during three different mandatory disclosure periods. Our study provides evidence that procurement of nonaudit fees can have varying impact under different reporting environments. The empirical evidence in this study also contributes to financial analysts' forecast literature by showing that the dispersion of earnings forecast is influenced by perceived auditor independence. This study provides some support for the SEC's concern about financial markets confidence in the US disclosure system when auditors are perceived to lack independence. Further, this study contributes to the debate about auditor independence by showing that disclosure can have significant influence on the behavior of companies and auditors.

The paper is organized as follows. Section II provides a review of the relevant literature and presents our hypotheses. Section III contains our main results and Section IV concludes.

II. COMBINATION OF AUDIT AND NONAUDIT SERVICES

In the agency context, although stockholders and creditors rely on management to safeguard their assets, they cannot directly observe if managers are acting in accordance with these expectations. Hence, accounting variables and other financial disclosures become proxies for direct observations with an external third party, the auditor, attesting to the accuracy of the disclosures (Fama and Jensen 1983a, 1983b). The value of attestation varies with audit quality of the external auditor. In the minds of outside users of financial information, actions that imply a conflict of interest impair the independence of the auditor and any indication of knowledge transfer between audit and nonaudit tasks enhances audit quality. One such action is the provision of nonaudit services (POB 1986). Increases in economic bonding associated with the provision of such services give auditors added incentive to maintain ongoing relationships (DeAngelo 1981; Simunic 1984; Beck, Frecka and Solomon 1988). To the extent that nonaudit fees are generated from audit clients, auditors face increased economic incentives to bear higher risk levels in audit engagements. Based on the magnitude of nonaudit fees when compared to audit fees, users of financial information form their perception about how much the loss of independence reduces the audit quality. Therefore, there is a sacrifice in perceived audit quality because of auditor provided nonaudit services. When a disclosure of nonaudit services is made, the assessment of independence made by external users indicates a sacrifice of perceived audit quality.

Economies of scope exist when Provision of one kind of service has a favorable impact on provision of other services. These economies of scope may result in increased revenues or reduced costs. While audit services are compulsory for registrant firms, nonaudit services may be purchased from auditors. Auditors have idiosyncratic knowledge of client's operations, which places them in a position of comparative advantage in providing audit and nonaudit services. It has been suggested in the accounting literature that knowledge gained by auditors in their work would facilitate their provision of nonaudit services and vice versa. This facet has been termed "knowledge spillover" in the literature (see Simunic 1984). Users of financial information observe the extent of nonaudit fees with respect to audit fees and make their assessment of benefits from economies of scope and knowledge spillover.

The Users' perception of audit quality when auditors provide both audit and nonaudit services is a combined assessment of benefits from economies of scope as well as knowledge spilloverand the sacrifices because of compromise of auditor independence. With higher assessed benefits, larger nonaudit fees as compared to audit fees would result in enhanced audit quality. If the users' assessment indicates higher sacrifices then an increased level of nonaudit fees in comparison to audit fees would mean lower audit quality.

DeAngelo (1981) models the audit pricing structure and finds that the future quasi-rents from audit engagements create an economic bond that provides incentives for the auditor and client to maintain their relationship which may result in compromises. Simunic (1984) analyzes the joint demand for nonaudit and audit services and shows economic bonding of an auditor to the client when attendant costs savings are retained. Beck et al. (1988) show that increase in bonding depend on the magnitude of nonaudit start-up and switching costs.

(A) Audit and Nonaudit Fees Disclosure Regimes and Related Evidence

The SEC had required registrant firms to disclose information about nonaudit fees during 1978-80 under ASR No. 250, during 2001-02 under Final Rule S7-13-00, and post 2002 under Final Rule S7-49-02.

1. Non-controversial Disclosure Period

Under ASR No. 250, SEC registered firms were required in 1978 to disclose the amount of auditor provided nonaudit services as a percentage of audit fees. This requirement was rescinded under ASR No. 304 in 1982. The commission implemented the disclosure requirement because it felt that reporting by external accountants substantially increased the reliability of financial statements. In order to assess the independence of the accountants, the SEC wanted to consider all relevant evidence bearing on the audit relationship. In its rescission of ASR No. 250, SEC cited the comments to the proposal to rescind ASR No 250 and concluded that there was no evidence of investor interest in disclosure of auditor provided nonaudit services. There were no observable accounting controversies during this period which made it easy for the SEC to repeal the nonaudit fees disclosure requirement under ASR No. 304.

In 1979, the Public Oversight Board (POB) issued a report which concluded that available evidence did not reveal any actual instances where the furnishing of nonaudit services impaired independence. McKinley, Pany, and Reckers (1985) examine whether provision of nonaudit services to audit clients affects the perception of the users of financial statements about independence of auditors. They observe that financial statements users believe that Big-8 auditors were more independent and that the provision of nonaudit services did not significantly affect loan decisions. Pany and Reckers (1988) study the perception of auditor independence by presenting identical loan and investment packages to loan officers and financial analysts with nonaudit variable systematically manipulated across individuals. The results of this study indicate that provision of nonaudit services by auditors has little effect on the perception of auditor independence. Nonexistence of accounting controversies and the empirical evidence from the twentieth century gives a general impression that the assessed benefits from economies of scope and knowledge spillovers were higher than or equal to the sacrifices due to compromise of auditor independence.

2. Controversial Disclosure Period

A dramatic increase in the proportion of revenues to auditors from nonaudit services in the 1990's renewed concerns about auditor independence and audit quality. Former SEC Chairman Levitt (2000) pointed out that, "auditing ... accounts for just 30 percent of total revenues--down from 70 percent in 1977." He further added to his concern that, "consulting and other [than audit] services may shorten the distance between the auditor and management. Independence--if not in fact, then certainly in appearance--becomes a more elusive proposition." After considerable debate, the SEC issued new rules on November 15, 2000. The rule required registered companies to disclose, in proxy statements filed after February 5, 2001, information about audit and nonaudit fees paid to auditors. Specifically, the companies were required to disclose audit fees, financial information systems design and implementation fees, and all other nonaudit fees:

(a) Audit Fees The fees paid for the annual audit and
 review of financial statements for the
 most recent fiscal year.

(b) Financial Information The aggregate fees billed for specified
 and System Design and information technology services rendered
 Implementation by the outside auditor during the most
 recent fiscal year. Specified information
 technologies services include the
 operation or management of a company's
 information system or local area network,
 and the design and implementation of
 hardware or software system that
 aggregates data underlying the financial
 statements or otherwise generates
 information significant to a company's
 financial statements.

(c) All Other Fees The aggregate fees for services rendered
 by the principal accountant other than
 those in items (a) and (b) above for the
 most recent fiscal year.

Frankel, Johnson and Nelson (2002) assessed investors' perception of auditor independence when auditors provide nonaudit services. Using nonaudit fees data collected from proxy statements filed between February 5, 2001 and June 15, 2001, Frankel et al. (2002) report an association, indicating a lack of auditor independence because of provision of nonaudit services by an auditor, between nonaudit fees and share values on the date the fees are disclosed. Krishnan, Sami and Zhang (2005) examine whether investors perceive auditor independence to be impaired when auditors provide nonaudit services to their clients. Based on the data from proxies statements filed in 2001, they find that various measures of nonaudit fees are negatively associated with earnings response coefficients, consistent with the notion that investors did perceive nonaudit services to impair auditor independence. Brandon, Crabtree and Maher (2004) analyze the effect of nonaudit fees on perceived auditor independence in the bond market. They note that bond-rating analysts are clearly identifiable users of financial disclosures. The authors analyze the ratings of new bond issues between February 2001 and December 2002 and report that auditor provided nonaudit services are negatively related to a client's bond rating.

Based on accounting controversies and the empirical evidence from early 2000s, it appears that the users assessed benefits from economies of scope and knowledge spillover are lower than the sacrifices because of compromise of auditor independence.

3. SOX Disclosure Period

In an effort to improve audit quality, the SOX prohibits auditors from providing any nonaudit services unless the company's audit committee pre-approves those services. In June 2002, the SEC also banned auditors from providing nonaudit services in nine specific areas that might impair auditor independence. The nine categories of prohibited non-audit services included in the Act are: (1) Bookkeeping or other services related to the accounting records or financial statements of the audit client; (2) Financial information systems design and implementation; (3) Appraisal or valuation services, fairness opinions, or contribution-in-kind reports; (4) Actuarial services; (5) Internal audit outsourcing services; (6) Management functions or human resources; (7) Broker or dealer, investment advisory, or investment banking services; (8) Legal and expert services unrelated to the audit; and (9) Any other service determined by the Board to be impermissible because of regulations. On February 6, 2003 the SEC issued "Final Rule: Strengthening the Commission's Requirements Regarding Auditor Independence" requiring companies listed on U.S. exchanges to disclose audit fees, audit-related fees, tax fees and all other fees for each of the two most recent fiscal years:

(a) Audit Fees The aggregate fees billed for each of the
 last two fiscal years for professional
 services rendered by the principal
 accountant for audit registrant's annual
 financial statements and review of
 financial statements included in the
 registrant's Form 10Q or 10-QSB or
 services that are normally provided by the
 accountant in connection with statutory
 and regulatory filings and engagements for
 those fiscal years.

(b) Audit-Related Fees The aggregate fees billed for each of the
 last two years for assurance and related
 services by the principal accountant that
 reasonably related to the performance of
 the audit or review of the registrant's
 financial statements.

(c) Tax Fees The aggregate fees billed for each of the
 last two years for professional services
 rendered by the principal accountant for
 tax compliance, tax advice, and tax
 planning.

(d) All Other Fees The aggregate fees billed for each of the
 last two years for other products and
 services provided by the principal
 accountant.

The last three categories of fees disclosed relate to nonaudit services. If new restrictions are perceived to be working, the users will assess benefits from economies of scope and knowledge spillover to be higher than or equal to the sacrifices because of compromise of auditor independence. Asthana, Balsam, and Kim (2009) observe that average audit fees charged by the Big-4 audit firms significantly increased after SOX. Cohen, Dey, and Lys (2008) report that management of accounting earnings by firms decreased significantly, without any change in the informativeness of earnings, after the passage of SOX. Heflin and Hsu (2008) report a significant decline in the use of non-GAAP earnings measures and the probability that disclosed earnings meet or exceed financial analysts' forecasts after the passage of SOX. Jain, Kim, and Rezaee (2008) find evidence consistent with an increase in investor confidence after the passage of SOX. Bushee and Leuz (2005) report that the SEC mandated disclosure are associated with positive stock returns and improved liquidity for companies already complying with those regulatory requirements. Jain and Rezaee (2006) document positive abnormal returns around the passage of SOX. Their results suggest that the market perceived the passage and implementation of SOX as a positive signal. Lobo and Zhou (2006) document fewer income increasing earnings management in the first year the CEOs/ CFOs certified the financial statements as mandated by the SOX than in the immediately preceding year when no certifications were required. Markelevich, Hoitash, and Barragato (2005) document a significant positive association between nonaudit fees and discretionary accruals in the pre-SOX time period but no association in the two years after the passage of SOX.

In summary, the audit fee ratio (i) increased from non-controversial period to the controversial pre-SOX period, and (ii) decreased from the controversial pre-SOX period to the post-SOX period. We state our first hypothesis in the null form:

H1: There is no difference in the fee ratios in the three reporting regimes.

(B) User's Assessment Metrics

We use dispersion of analysts' earnings forecast and stock market reaction to nonaudit fees disclosure in proxy statements to measure users' perception. Analysts use financial information to predict earnings. Lower disclosure quality should result in lower financial information leading to a greater disagreement among analysts about forecast of future earnings. The dispersion of earnings forecasts by analysts can be used to measure audit quality as perceived by analysts(see, for instance, Imhoff and Lobo 1992; and Barron and Stuerke 1998). Furthermore, the theoretical predictions of Abarbanell, Lanen, and Verrecchia (1995) and empirical evidence of Ajinkya and Gift (1985) point to a positive association between stock price variance and earnings forecast dispersion.

We also use the reaction of stock market to the disclosure of nonaudit fees as an investors perception of audit quality in the presence of nonaudit services. Frankel et al. (2002) suggest an association between share prices and the disclosure of nonaudit fees. In a slightly different context, Firth (1990) reports negative market reaction for firms when the Department of Trade of U.K. criticized the firms' auditors for the work peribrmed. Fcroz, Park, and Pastena (1991) and Dechow, Sloan, and Sweeney (1996) provide similar evidence for the announcements of SEC enforcement actions for companies accused of manipulating earnings. It is reasonable to assume that stock market reaction at the time of nonaudit fees announcements in proxy statements provides an appropriate measure of investors' perception of audit quality in the presence of auditor provided nonaudit services.

In light of the above discussion, we claim that the dispersion of analysts forecast is (i) negatively related to nonaudit fees in the non-controversial period, (ii) positively related to nonaudit fees in the controversial pre-SOX period, and (rid negatively related to nonaudit fees in the SOX period. Further, stock market reaction to disclosure of nonaudit fees is (i) positively related to nonaudit fees in the noncontroversial period, (ii) negatively related to nonaudit fees in the controversial pre-SOX period, and (rid positively related to nonaudit fees in the SOX period. The SEC takes the position that fee ratio, the ratio of nonaudit to audit fees, is useful for users to assess auditor independence (SEC 2000). Prior research has also analyzed the impact of the fee ratio to assess auditor independence (Parkash and Venable 1993; Firth 1997). We state our hypotheses about investors' perception in the null form:

H2: Fee ratio is not associated with analysts forecast dispersion immediately after nonaudit fees announcement.

H3: Fee ratio is not associated with stock market reaction to the disclosure of nonaudit fees.

III. EMPIRICAL RESULTS

(A) Sample Selection

Under ASR No. 250, SEC registrants were required to disclose the amount of" nonaudit services as a percentage of audit fees in proxy statements filed between September 30, 1978 and January 28, 1982. Under the rule, in addition to total nonaudit services (NONADT) firms were also required to disclose the nature of' specific audit services that were at least three percent of audit fees. Specific audit services disclosed were for tax services; pension and personnel services; system services; merger and acquisition services; and other services. This study uses nonaudit information for Fortune 500 companies in the three fiscal years (1978-80) during which ASR No.250 was in effect. Only the Fortune 500 companies are chosen because they are more likely to have data available for all the three reporting regimes.

The nonaudit data were collected manually from proxy statements filed during 1978-80. The SEC issued Final Rule S7-13-00 mandating disclosure of nonaudit and audit fees paid to auditors in proxy statements filed after February 5, 2001. The new rule required disclosure of audit fees, financial information systems related nonaudit fees, and other nonaudit fees in proxies filed on or after February 5, 2001. Further, the SOX required registrants to provide information about audit fees, audit related nonaudit fees, audit fees related to tax services, and other nonaudit fees. We obtain audit and nonaudit fees information after February 5, 2001 from proxy statements filed with the SEC. Firm specific financial data are collected from COMPUSTAT and returns data are obtained from CRSP. Financial analysts' forecasts were collected from the Institutional Brokers Estimate System (IBES). Our sample includes 353 firm year observation in the non-controversial period, 315 firm year observations in the controversial pre-SOX period and 205 firm year observations in the post-SOX period (see Table 1). As reported in Table 1, the number of observations ranged from 96 (in 2000) to 119 (in 1980).

(B) Analysis of Fee Ratios

We analyze the yearly distribution of the fee ratio to assess the impact of mandated fee disclosures under various reporting regimes. The decrease in fee ratio indicates that companies concerned about fee ratio may reduce procurement of nonaudit services in relation to audit fees resulting in a lower fee ratio. The mean fee ratio decreased insignificantly from 0.299 to 0.278 in the non-controversial ASR No. 250 period during 1978-80. The mean fee ratio decreased significantly from 2.983 in 2000 to 2.43 in 2001 and further to 1.264 in 2002 (0.01 level two-tailed) in each year of the controversial pre-SOX period. The mean fee ratio decreased significantly from 0.800 in 2003 to 0.383 in 2004 (0.01 level, two-tailed) in the post-SOX era also. Our interpretation of the fee ratio changes within a reporting regime is that although the required disclosure did not have an impact in the non-controversial regime, there was a significant impact during the controversial pre-SOX and post-SOX time periods.

Panel B, Table 1 reports the fee ratio distribution and fee ratio changes across the three reporting regimes. The mean fee ratio in the non-controversial period was 0.29 which increased to 2.136 during the controversial pre-SOX period. The increase in mean fee ratio from non-controversial to controversial period was significant at the 0.01 level (two-tailed). The mean fee ratio during post-SOX is 0.566, which is a significant (0.01 level, two-tailed) decrease from the controversial pre-SOX period. In general, these results indicate that fee ratio had increased dramatically in scandalous era after the ASR No. 250 time period and that the passage of SOX resulted in significantly lower fee ratio. The evidence indicates that the hypothesis H1 of no difference in fee ratios in three reporting regimes is rejected.

As reported in Table 2, similar observation can be made about the fee ratios for each auditor during the three reporting regimes. Some of the audit firms merged between reporting regimes. For a meaningful comparison, we combined observations the Big-8 auditors present during ASR No. 250 to obtain the auditors present during the 2000s. We find that the controversial pre-SOX period had significantly higher fee ratio for each auditor as compared to the non-controversial time period of ASR No. 250. The passage of SOX was accompanied by significantly reduced fee ratios for each auditor in the post-SOX period as compared to the pre-SOX period.

Table 3 presents fee ratio distribution for clients in eight industries. The trend of fee ratios for clients in each industry during the three reporting regimes is similar to the trends reported for auditors. The mean fee ratios in each industry in the pre-SOX period are larger than those reported in the ASR No. 250 reporting period. Mean fee ratios in each industry during the post-SOX period are lower than those reported in the controversial pre-SOX period. The evidence from Tables 2 and 3 indicates that fee ratios during the controversial pre-SOX period increased significantly as compared to the ASR No. 250 period. With the passage of SOX, fee ratios during post-SOX era decreased in comparison to the pre-SOX period.

(C) Regression Analysis

To test the impact of fee ratio on the dispersion of analysts' forecasts, we used the following equation:

DISP = [[alpha].sub.0] + [[alpha].sub.1]BETA + [[alpha].sub.2]MV + [[alpha].sub.3]FEERATIO + [[alpha].sub.4]AUDITOR + [[alpha].sub.5]NUMEST + [[alpha].sub.6]PHARMA + [[alpha].sub.7]CHEMICAL + [[alpha].sub.8]COMPUTERS + [[alpha].sub.9]TEXTILE + [[alpha].sub.10]FOOD + [[alpha].sub.11]MINNING & CONST + [[alpha].sub.12]ELECT MANUF + [epsilon](1)

To test the relationship between dispersion of analysts' forecasts and the fee ratio (H2), we assess the sign and significance of [[alpha].sub.3]. Other variables are included in equation (1) to control for uncertainty due to production, investment and financing decisions, and level of information availability shown in prior literature to determine dispersion of forecast (Parkash, Dhaliwal, and Salatka 1995). Equation (1) is estimated separately for each disclosure regime and results are presented in Table 4. DISP is defined as the standard deviation of analysts' forecasts divided by the absolute value of the mean analysts' forecast. BETA is estimated by regressing the daily stock return for each firm on the CRSP value weighted index around the proxy filing date. We use 6 months of daily stock returns before and 6 months of daily stock returns after the proxy filing date, excluding a month on each side of the proxy filing date. MV is the natural log of market value of equity for the sample firms. FEERATIO is defined as nonaudit fees scaled by audit fees paid to the auditor. AUDITOR is an indicator variable that takes value one if the auditor belongs to Big-5, otherwise it is zero. NUMEST is the number of estimates by analysts for each of the sample firm as reported in the IBES database. In our regressions, we control for industry specific effects by including seven indicator variables for those industries. To minimize the effect of extreme observations on our analysis, we rank DISP, BETA, and FEERATIO by percentiles and use the ranked numbers in our analysis.

For the ASR No. 250 non-controversial period, the coefficient of FEERATIO is significantly (0.01 level, two-tailed) negative indicating inverse relationship between the dispersion of" analysts' forecast and fee ratio indicating higher audit quality with higher fee ratio. In contrast, during the pre-SOX controversial period, the coefficient of FEERATIO is significantly (0.01 level, two-tailed) positive showing a direct relationship between the dispersion of analysts' forecast and fee ratio to signify lower audit quality with higher fee ratio. In the post-SOX period, the coefficient of FEERATIO is negative but insignificant indicating no relationship between dispersion of' forecast and fee ratio to mean no relationship between audit quality and the ratio.

To explore investors' assessment of audit quality in the presence of auditor provided nonaudit fees, we examine the five-day CARs around filing of proxy statements disclosing either the the ratio or information to calculate the fee ratio. We follow the standard event study methodology proposed by Brown and Warner (1985) to calculate CARs. Daily returns for 360 days (covering the period t-390 through t-30) are used to estimate market model parameters for each firm, where day t represents the filing date of the proxy statements. Cumulative abnormal returns (CAR) summarize the price impact over the five-day period (-2,+2) surrounding the proxy date. Average standardized abnormal returns and average standardized cumulative abnormal returns are utilized to compute the test statistics. Table 5 reports the five-day CARs around the proxy filing dates for the three reporting regimes. CARs for ASR No. 250 and post-SOX periods are positive but insignificant. CARs of -0.0074 associated with fee ratio announcements during the pre-SOX controversial period is significant (0.01 level). This analysis shows that the investors reacted negatively to announcements of fee ratios during controversial period but positively during ASR No. 250 and post-SOX periods.

The evidence presented in Tables 5 supports the rejection of H3 for controversial pre-SOX time period. For the ASR No. 250 non-controversial period and the post-SOX, we are unable to reject H3.

IV. CONCLUDING REMARKS

We examine the trend of the fee ratio measured as the ratio of nonaudit to audit fees for three reporting periods--the non-controversial period of 1978-80 under ASR No. 250, the controversial pre-SOX time period between February 2001 and July 2002, and the post-SOX period after enactment of SOX in July 2002--when it was mandatory to disclose information related to the fee ratio. We observe that the fee ratio decreased insignificantly in the non-controversial ASR No. 250 period, but decreased in each year of the controversial pre-SOX period as well as the post-SOX era indicating that although required disclosure did not have an impact in the non-controversial regime, they had a significant impact during the controversial pre-SOX and the post-SOX time periods. Our empirical results also imply that the fee ratio had increased dramatically in the scandalous pre-SOX era after the ASR No. 250 time period and that the passage of SOX resulted in significantly lower fee ratio for each client, firm, and industry.

We use dispersion of analysts' earnings forecast to assess the impact of auditor provided nonaudit services. Our results indicate an inverse relationship between forecast dispersion and fee ratio to show higher audit quality with higher fee ratio during the ASR No. 250 period. In contrast, during the pre-SOX controversial period, the direct relationship between forecast dispersion and fee ratio signifies lower audit quality with higher fee ratio. In the post-SOX period, the insignificant relationship between forecast dispersion and fee ratio establishes no relation between audit quality and fee ratio. We also explore investors' assessment of audit quality in the presence of auditor provided nonaudit fees by examining stock market reaction around filing of proxy statements disclosing information about the fee ratio. Our empirical analysis shows that investors reacted negatively to announcements of fee ratios during the controversial pre-SOX period, but insignificantly to fee ratio disclosures during the ASR No. 250 and the post-SOX periods. Collectively, our empirical results provide evidence that auditor provided nonaudit fees as compared to audit fees are related to (i) increased audit quality in the ASR No. 250 reporting period, (ii) decreased audit quality in the controversial pre-SOX period, and (iii) no effect on audit quality in the post-SOX period. Our results provide some support in favor of enacting SOX.

References

Abarbanell, J. S., W.N. Lanen and R. E. Verrechia (1995), "Analysts' Forecasts as Proxies for Investor Beliefs in Empirical Research," Journal of Accounting and Economics, Vol. 20, No. 1, 31-60.

Ajinkya, B. P. and M. J. Gift (1985), "Dispersion of Financial Analysts' Earnings Forecasts and the (Option Model) Implied Standard Deviation of Stock Returns," Journal of Finance, Vol. 40, 1353-1365.

Asthana, S., S. Balsamand S. Kim (2009), "The Effect of Enron, Andersen, and Sarbanes-Oxley on the US Market for Audit Services," Accounting Research Journal, Vol. 22, No. 1, 4-26.

Barron, O. E. and P. S. Stuerke (1998), "Dispersion in Analysts' Earnings Forecasts as a Measure of Uncertainty," Journal of Accounting, Auditing, and Finance, Vol. 13, 245-270.

Beck, P. J., T J. Frecka and I. Solomon (1988), "A Model of the Market for MAS and Audit Services: Knowledge Spillovers and Auditor-Auditee Bonding," Journal of Accounting Literature, Vol. 7, 50-64.

Brandon, D., A. Crabtree and J. Maher (2004), "Non-Audit Fees, Auditor Independence and Bond Ratings," Auditing: A Journal of Theory and Practice, Vol. 23, No. 2, 89-103.

Brown, S. J. and J. B. Warner (1985), "Using Daily Stock Returns: The Case of Event Studies," Journal of Financial Economics, Vol. 14, No. 1, 3-31.

Bushee, B. J. and C. Leuz (2005), "Economic Consequences of SEC Disclosure Regulation: Evidence from the OTC Bulletin Board," Journal of Accounting and Economics, Vol. 39, No. 2, 233-264.

Cohen, D. A., A. Dey and T. Lys (2008), "Real and Accrual-Based Earnings Management in the Pre- and Post-Sarbanes Oxley Periods," The Accounting Review, Vol. 82, No. 3,757-787.

DeAngelo, L. (1981), "Auditor Independence, Low-Bailing, and Disclosure Regulation," Journal of Accounting and Economics, Vol. 3, No. 2, 113-127.

Dechow, P. M., R. G. Sloan and A. P. Sweeney (1996), "Causes and Consequences of Earnings Manipulation: An Analysis of Firms Subject to Enforcement Actions by the SEC," Contemporary Accounting Research, Vol. 13, No. 1, 1-36.

Fares, E. F. and M. C. Jensen (1983a), "Separation of Ownership and Control," Journal of Law and Economics, Vol. 26, 301-325.

Fama, E. F. and M. C. Jensen (1983b), "Agency Problems and Residual Claims," Journal of Law and Economic, Vol. 26, 327-349.

Feroz, E. H., K. Park and V.S. Pastena (1991), "The Financial and Market Effects of the SEC's Accounting and Auditing Enforcement Releases," Journal of Accounting Research, Vol. 29, 107-142.

Firth, M. (1990), "Auditor Reputation: The Impact of Critical Reports Issued by Government Inspectors," The Rand Journal of Economics, Vol. 21, No. 3, 374-387.

Firth, M. (1997), "The Provision of Nonaudit Services by Accounting Firms to Their Audit Clients," Contemporary Accounting Research, Vol. 14, 1-21.

Frankel, R., M. Johnson and K. Nelson (2002), "The Relation between Auditors' Fees for Nonaudit Services and Earnings Management," The Accounting Review, Vol. 77, 71-105.

Heflin, F. and C. Hsu (2008), "The Impact of the SEC's Regulation of Non-GAAP Disclosures," Journal of Accounting and Economics, Vol. 46, No. 2-3, 349-365.

Imhoff, E. and G. Lobo (1992), "The Impact of Ex Ante Earnings Uncertainty on Earnings Response Coefficient," The Accounting Review, Vol. 67, 427-39.

Jain, P. K. and Z. Rezaee (2006), "The Sarbanes-Oxley Act of 2002 and Capital-Market Behavior: Early Evidence," Contemporary Accounting Research, Vol. 23, No. 3, 629-654.

Jain, P. K, J. C. Kim and Z. Rezaee (2008), "The Sarbanes-Oxley Act of 2002 and Market Liquidity," Financial Review, Vol. 43, No. 3, 361-382.

Krishnan, J., H. Sami and Y. Zhang (2005), "Does the Provision of Nonaudit Services Affect Investor Perceptions of Auditor Independence?" Auditing: A Journal of Practice & Theory, Vol. 24, No. 2, 111-135.

Levitt, A. (2000), "Renewing the Convent with Investors," Remarks delivered at the NYU Center for Law and Business, New York, NY.

Lobo, G. J. and J. Zhou (2006), "Did Conservatism in Financial Reporting Increase after the Sarbanes-Oxley Act? Initial Evidence," Accounting Horizons, Vol. 20, No. 1, 57-73.

Markelevich, A. J., R. Hoitash and C.A. Barragato (2005), "Auditor Fees, Abnormal Fees and Audit Quality Before and After the Sarbanes-Oxley Act," Available at SSRN: http://ssrn.com/abstract=646681 or doi:10.2139/ssrn.646681

McKinley, S., K. Pany, and P.M.J. Reekers (1985), "An Examination of the Influence of CPA Firm Type, Size, and MAS Provision on Loan Officer Decisions and Perceptions," Journal of Accounting Research, Vol. 23, 887-896.

Pany, K., and P. M. J. Reekers (1988), "Auditor Performance of MAS: A Study of Its Effects on Decisions and Perceptions," Accounting Horizons, Vol. 2, 31-38.

Parkash, M., D. S. Dhaliwal and W. K. Salatka (1995), "How Certain Firm-Specific Characteristics Affect the Accuracy and Dispersion of Analysts' Forecasts: A Latent Variables Approach," Journal of Business Research, Vol. 34, No. 3, 161-169.

Parkash, M. and C. Venable (1993), "Auditee Incentives for Auditor Independence: The Case of Nonaudit Services," The Accounting Review, Vol. 68, No. 1, 113-133.

Public Oversight Board, Securities and Exchange Commission Practice Section, American Institute of Certified Public Accountants (1986), Public Perceptions of Management Advisory Services Performed by CPA Firms for Audit Clients. New York: POB.

Securities and Exchange Commission (SEC) (1978), ASR No.250: Disclosure of Relationships with Independent Public Accountants. Washington, D.C.: U.S. Government Printing Office (June).

Securities and Exchange Commission (SEC) (1982), ASR No.304: Relationships between Registrants and Independent Accountants. Washington, D.C.: U.S. Government Printing Office (January).

Securities and Exchange Commission (SEC) (2000), Final Rule: Revision of the Commission's Auditor Independence Requirements. File No. S7-13-00. Washington, D.C.: U.S. Government Printing Office.

Simunic, D. A. (1984), "Auditing, Consulting, and Auditor Independence," Journal of Accounting Research, 679-702.

MOHINDER PARKASH *, RAJEEV SINGHAL * AND YUN (ELLEN) ZHU *

* Oakland University, School of Business Administration, Department of Accounting and Finance, Rochester, M148309, E-mails: [email protected]; [email protected]; [email protected]

Table 1
Distribution Fee Ratios by Years and Regimes
Panel A: Distribution of Fee Ratio by Years

 Mean Median
Year #Firms (t-stat for change) (z-stat for change)

1978 116 0.299 0.210

1979 118 0.292 0.165
 (-0.17) (-0.47)

1980 119 0.278 0.210
 (-0.34) (0.18)

2000 96 2.983 2.199
 (8.81 *) (11.65 *)

2001 108 2.438 1.770
 (-1.41) (-1.52)

2002 101 1.264 0.902
 (-4.29) * (-5.49) *

2003 110 0.800 0.617
 (-3.14) * (-3.91) *

2004 111 0.383 0.276
 (-5.06) * (-6.14) *

Panel B: By Regime

1978-1980 353 0.290 0.200

2000-Pre Sox 325 2.136 1.419
 (13.69 *) (19.21 *)

Post Sox 201 0.566 0.400
 (-11.09 *) (-13.23 *)

Year Min Max

1978 0.000 1.940

1979 0.000 1.980

1980 0.020 1.430

2000 0.084 16.045

2001 0.003 20.000

2002 0.078 8.108

2003 0.049 6.151

2004 0.039 2.551

Panel B: By Regime

1978-1980 0.000 1.980

2000-Pre Sox 0.003 20.000

Post Sox 0.039 6.151

* Significant at 1% level using two-tailed tests

Table 2
Distribution of Fee Ratio by Auditor

 AA DT

 N Mean Median N Mean Median

1978-1980
(1) 66 0.52 0.37 44 0.20 0.17

2000-Pre Sox
(2) 41 1.68 * 1.53 * 72 1.96 1.63

Post Sox (3) 31 0.46 0.39

t-stat change
1 to 2 5.73 * 6.00 * 10.61 * 8.48 *

t-stat change
2 to 3 -8.60 * -6.40 *

t-stat change
1 to 3 4.10 * 3.98 *

 EY KPMG

 N Mean Median N Mean Median

1978-1980
(1) 77 0.23 0.18 49 0.21 0.21

2000-Pre Sox
(2) 46 1.57 1.03 51 1.40 1.00

Post Sox (3) 50 0.61 0.51 35 0.34 0.29

t-stat change
1 to 2 6.08 * 7.92 * 3.76 * 6.76 *

t-stat change
2 to 3 -3.80 * -4.44 * -3.33 * -5.20 *

t-stat change
1 to 3 3.16 * 5.36 * 2.90 * 2.30 *

 PWC OTH

 N Mean Median N Mean Median

1978-1980
(1) 103 0.28 0.18 14 0.15 0.08

2000-Pre Sox
(2) 115 2.96 1.82

Post Sox (3) 85 0.67 0.45

t-stat change
1 to 2 8.77 * 11.71 *

t-stat change
2 to 3 -7.31 * -8.85 *

t-stat change
1 to 3 4.83 * 5.57 *

* Significant at 1% level using two-tailed tests.

Table 3
Distribution of Fee Ratio by Industry

 Chemical Computer

 Mean Median Mean Median

1978-1980 (1) 0.24 0.25 0.35 0.21
(#Firms) (42) (75)

2000-Pre Sox (2) 1.70 1.37 3.01 1.85
(#Firms) (38) (68)

Post Sox (3) 0.47 0.39 0.59 0.42
(#Firms) (25) (44)

t-stat change 6.13 * 7.09 * 5.72 * 8.96 *
1 to 2

t-stat change -4.91 * -4.86 * -5.09 * -6.75 *
2 to 3

t-stat change 3.03 * 3.26 * 2.25 * 2.25 *
1 to 3

 Elec. Man. Food

 Mean Median Mean Median

1978-1980 (1) 0.27 0.16 0.29 0.28
(#Firms) (92) (32)

2000-Pre Sox (2) 2.03 1.63 2.57 1.80
(#Firms) (84) (31)

Post Sox (3) 0.62 0.46 0.72 0.68
(#Firms) (48) (20)

t-stat change 9.26 * 10.14 * 5.24 * 6.64 *
1 to 2

t-stat change -6.16 * -6.88 * -4.09 * -4.58 *
2 to 3

t-stat change 2.52 * 4.96 * 3.33 * 3.77 *
1 to 3

 Machinery Pharma

 Mean Median Mean Median

1978-1980 (1) 0.30 0.28 0.21 0.14
(#Firms) (27) (18)

2000-Pre Sox (2) 1.48 1.36 1.80 1.09
(#Firms) (24) (18)

Post Sox (3) 0.40 0.39 0.53 0.56
(#Firms) (13) (14)

t-stat change 5.84 * 5.28 * 3.02 * 4.64 *
1 to 2

t-stat change -5.28 * -3.90 * -2.40 ** -2.87 *
2 to 3

t-stat change 1.28 1.53 3.43 * 3.59 *
1 to 3

 Textile

 Mean Median

1978-1980 (1) 0.28 0.12
(#Firms) (59)

2000-Pre Sox (2) 1.71 0.98
(#Firms) (56)

Post Sox (3) 0.51 0.29
(#Firms) (36)

t-stat change 5.29 * 6.73 *
1 to 2

t-stat change - 4.24 * -4.30 *
2 to 3

t-stat change 2.19 ** 3.18 *
1 to 3

Significant at 1% level using two-tailed tests.

Table 4
Regression of Dispersion of Analyst Forecasts

 1978-1980 2000-Pre Sox Post Sox
 Coefficient Coefficient Coefficient
Variable (p-value) (p-value) (p-value)

INTERCEPT 1.036 * 0.454 * 0.762 *
 (0.00) (0.00) 0.00)

BETA 0.189 * 0.479 * 0.525 *
 (0.00) (0.00) (0.00)

MV -0.102 * -0.078 * -0.071 *
 (0.00) (0.00) (0.00)

FEERATIO -0.129 * 0.092 *** -0.038
 (0.01) (0.06) (0.48)

AUDITOR 0.049 *** 0.055 *** 0.166 *
 (0.10) (0.06) (0.00)

NUMEST 0.011 * 0.012 * 0.009 *
 (0.00) (0.00) (0.01)

PHARMA -0.412 * 0.190 -0.014
 (0.00) (0.12) (0.95)

CHEMICAL 0.057 0.317 * -0.002
 (0.54) (0.00) (0.99)

COMPUTERS -0.014 0.208" -0.089
 (0.88) (0.04) (0.67)

TEXTILE -0.005 0.341 * -0.010
 (0.96) (0.00) (0.96)

FOOD -0.133 0.232 ** -0.002
 (0.17) (0.03) (0.99)

MINING & CONSTRUCTION 0.197 ** 0.649 * 0.463 **
 (0.04) (0.00) (0.03)

ELECT. MANUFACTURER 0.039 0.243 * -0.010
 (0.66) (0.01) (0.96)

Obs 353 325 193

[R.sup.2] 0.35 0.43 0.54

F-statistic 15.17 19.78 17.73

(p-value) 0.00 * 0.00 * 0.00 *

*,**,*** Significant at 1%, 5%, and 10% levels respectively using
two-tailed tests

Table 5
Five Day Cumulative Average Abnormal Returns around the Proxy Dates

Period N CHAR z-stat

1978-1980 324 0.0039 1.41
2000-Pre Sox 343 -0.0074 -2.25 *
Post Sox 111 0.0020 0.61

* Significant at 1% level using two-tailed tests.