مقاله بیس سابقه و هدف مدیر بخش مالی شرکت و هزینه حسابرسی
ACCOUNTING HORIZONS American Accounting Association Vol. 30, No. 3 DOI: 10.2308/acch-51442
September 2016 pp. 325–339
CEO Financial Background and Audit Pricing
Rachana Kalelkar University of Houston–Victoria
Sarfraz Khan University of Louisiana at Lafayette
SYNOPSIS: Accounting scholars theorize that audit price is a function of a client’s audit and business risk. Existing research finds that the functional expertise of Chief Executive Officers (CEOs) in finance improves financial reporting quality (Matsunaga, Wang, and Yeung 2013), increases profitability, and reduces the likelihood of firm failure (Custodio and Metzger 2014). These factors suggest that auditors’ engagement risk decreases when incumbent CEOs possess financial expertise, raising the likelihood that auditors will charge these firms lower fees. In this study, we examine whether CEOs’ work experience in accounting- and finance-related jobs affects audit fees. Using a panel of U.S. firms between 2004 and 2013, we find that firms that have a financial expert CEO pay lower audit fees. Our results are robust to various specifications, including firm-fixed effect model and specifications that control for other CEO- and Chief Financial Officer (CFO)-specific and audit committee characteristics. Our findings thus add to the literature on the advantages and disadvantages of a functional background of top managers and how this background can create value for a firm through savings in audit fees.
Keywords: CEO ﬁnancial expertise; audit fees.
INTRODUCTION Accounting scholars theorize that audit price is a function of a client’s audit and business risk (e.g., Hay, Knechel, and Wong 2006; Simon and Francis 1988; Simunic 1980). Empirically, there is evidence that auditors consider the CEO’s characteristics when determining the client’s business and audit risk. For instance, Johnson, Kuhn, Apostolou, and Hassell (2012) show that auditors charge higher fees when a client’s CEO exhibits narcissistic behavior, and Kim, H. Li, and S. Li (2015) and Wysocki (2010) ﬁnd CEOs’ incentive-based compensation to be positively related to audit fees. We extend the literature in this area by studying a relatively neglected aspect of CEO characteristics—the work experience of CEOs in accounting and ﬁnance-related jobs—and its impact on auditors’ engagement risk and, therefore, audit pricing. The upper echelons theory suggests that executives’ background and experiences shape the choices they make (Hambrick and Mason 1984; Hitt and Ireland 1985). Prior research shows that CEOs’ functional backgrounds make them more effective at addressing problems in related functional areas. For example, Koyuncu, Firﬁray, Claes, and Hamori (2010) document that CEOs with a background in operations are better able to handle problems related to the supply chain, while CEOs with marketing backgrounds are better able than their counterparts to manage marketing policies (Boyd, Chandy, and Cunha 2010). A recent article in the Wall Street Journal highlights an upward trend in the appointment of CEOs with a functional background in ﬁnance (henceforth referred to as ‘‘ﬁnancial expert’’)(Johnson 2015).1 The rise in the number of CEOs with a ﬁnancial background prompts the question of why an increasing number of ﬁrms hire CEOs with such a background. One possible explanation is the increased focus on ﬁnancial reporting and disclosure policies and potentially increased liabilities due to accounting failure in the post-Sarbanes-Oxley of 2002 (SOX) period (Cao and Narayanamoorthy 2014). Another possible
We thank Sharad Asthana, Emeka Nwaeze, and participants at the 2015 AAA Southwest Region Meeting in Houston for comments on earlier drafts. Both authors contributed equally.
Editor’s note: Accepted by Paul A. Grifﬁn.
Submitted: July 2014 Accepted: March 2016 Published Online: March 2016
1 In a similar study, Custodio and Metzger (2014) analyze S&P 1500 ﬁrms from 1993 to 2007 and ﬁnd that 41 percent of their sample ﬁrms had CEOs with a functional background in ﬁnance.
explanation is the increased ﬁnancial constraint inﬂicted by the recession of 2008–2009, which makes CEOs with ﬁnancial acumen more desirable candidates as they are likely to possess a better ability to manage the limited ﬁnancial resources and use them more productively (Custodio and Metzger 2014). The existing literature documents that ﬁrms managed by ﬁnancial expert CEOs beneﬁt in terms of improved ﬁnancial policies (Custodio and Metzger 2014) and better disclosure practices (Matsunaga, Wang, and Yeung 2013). We investigate another potential channel (i.e., audit fees) through which a ﬁnancial expert CEO can create value to the ﬁrm.2 We hypothesize that ﬁrms with ﬁnancial expert CEOs pay lower audit fees, because the ﬁnancial expertise of CEOs (1) improves the quality of earnings (Matsunaga et al. 2013), thus reducing the risk of material misstatement, and (2) increases proﬁtability and reduces ﬁrm failure (Custodio and Metzger 2014), reducing a client’s business risk. In other words, auditors’ engagement risk appears to decline with the ﬁnancial expertise of the CEO, raising the prospect that audit fees will be relatively lower for ﬁrms where CEOs are ﬁnancial experts. We use a sample of data consisting of 77 ﬁrms and 81 changes in ﬁnancial expertise for the period ranging from 2004 to 2013. We hand-collect relevant CEO background information and examine the link between CEOs’ ﬁnancial expertise and audit fees. Following Custodio and Metzger (2014), we deﬁne CEO as a ﬁnancial expert if he or she worked in banking or investment ﬁrms, was an employee of an auditing ﬁrm, or worked as a CFO, treasurer, or vice president (VP) of ﬁnance. One of the concerns with research related to executive characteristics is the endogenous nature of CEO-ﬁrm matching (or the prospect that CEO expertise and ﬁrms’ ﬁnancial and disclosure practices are jointly determined). To the extent that some unobserved quality simultaneously affects both ﬁrms’ accounting outcome and ﬁnancial expert CEOs’ selections, our observed results may be biased. To address such a concern, we follow an approach similar to the managerial-ﬁxed effect model used in Bertrand and Schoar (2003). Speciﬁcally, we restrict our analysis only to those ﬁrms that either changed CEOs from a nonﬁnancial expert to ﬁnancial expert or otherwise. Furthermore, we require ﬁrms to have at least a three-year range of data before and after the change within a panel to be included in the sample. We also consider a ﬁrm-ﬁxed effect model to control for unobservable persistent ﬁrm effects. Our results show that ﬁrms with a ﬁnancial expert CEO pay lower audit fees. In terms of economic magnitude, we ﬁnd that annual audit fees are lower by approximately 8.5 percent, or approximately $310,000 for ﬁrms that have a ﬁnancial expert CEO. We also use an instrumental variables approach to further validate our results. Using local density of ﬁnancial ﬁrms as an instrument, we continue to ﬁnd that ﬁrms with ﬁnancial expert CEOs pay lower audit fees. Additionally, in unreported regressions, we control for managerial ability and audit committee diligence; the results based on these speciﬁcations (not tabulated) remain similar. This study makes an important contribution to the literature that focuses on CEOs’ personal traits and ﬁrm performance. Our study adds to the literature on the role of CEOs’ ﬁnancial expertise in corporate governance. In particular, we argue that, given the fact that CEOs are the ultimate authority over policy decisions, they play a more inﬂuential role in the reporting process when they are ﬁnancial experts. Our ﬁndings increase our understanding that auditors ﬁnd CEOs’ functional background in ﬁnance to be a relevant factor in audit pricing decisions. Our results, however, should be interpreted with caution. In particular, the sample used in the study is relatively small and comprises relatively large ﬁrms. The possible implication is that the ﬁndings and inferences are limited to such ﬁrms, and may not be generalizable to a broader population. Despite such a limitation, our analysis and the robustness of the results suggest that CEOs’ ﬁnancial expertise has a statistically and economically signiﬁcant effect on audit pricing. The remainder of the paper proceeds as follows. The following section provides a literature review and develops the hypothesis. The third section illustrates our sample selection process, and the fourth describes the research design. The ﬁfth section then reports the results. The sixth section reports the additional tests, and the seventh concludes the paper.
THEORY AND HYPOTHESIS DEVELOPMENT
The role of managers in ﬁrm performance is subject to debate among scholars. On the one hand, organizational ecologists question the value of the manager to the ﬁrm. Their argument is that the success of an organization is determined largely by its quality of products, core competency, life cycle, and to some extent even luck, and not by CEO ability (e.g., Cohen, March, and Olsen 1972; Hannan and Freeman 1977). On the contrary, the upper echelon theory proposes that top managers make a signiﬁcant contribution to ﬁrm performance (e.g., Hambrick and Mason 1984; Harris and Helfat 1997; Hayes and Schaefer 2000). The key proposition of the theory is that organizational outcomes reﬂect the values and abilities of top managers. Furthermore, the human capital theory suggests that knowledge and abilities possessed by managers can be an important determinant of organizational performance (e.g., Amit and Shoemaker 1993; Becker 1962; Coff 2002).
2 The ﬁnancial expertise of CEOs can create value through various sources, such as more efﬁcient use of cash, better investment decisions, etc. (see Custodio and Metzger 2014 for more details). In this paper, we focus only on value created through savings in audit fees.
326 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
Drawing from alternative views about the role of executives in organizations, a substantial body of literature examines the role of CEOs in accounting outcomes. These studies have variously explored the link between ﬁnancial reporting outcomes and CEOs’ personal attributes, such as gender (Betz, O’Connell, and Shepard 1989; Butz and Lewis 1996; Mason and Mudrack 1996), age (Peterson, Rhoads, and Vaught 2001; Sundaram and Yermack 2007), and education (e.g., Bhagat, Bolton, and Subramanian 2010; Kimberly and Evansiko 1981). Other studies focus on the effect of the CEO’s motives on the accounting outcomes. For instance, Burns and Kedia (2006) focus on the implications of CEOs’ equity incentive on ﬁnancial misreporting (see also Baik, Brockman, Farber, and Lee 2011; Bergstresser and Philippon 2006; Cheng and Warﬁeld 2005). Turning to auditing, a growing body of knowledge ﬁnds that CEO characteristics affect audit risk. To illustrate, Kim et al. (2015) and Wysocki (2010) ﬁnd that auditors perceive CEO equity incentive to increase their audit risk and, thus, incorporate such risk in their pricing decisions. Hribar, Kim, Wilson, and Yang (2012) ﬁnd that overconﬁdent CEOs report more aggressively and that auditors increase audit fees to compensate for aggressive reporting. Other factors shown to be related to audit fees include CEO narcissism (Johnson et al. 2012; Judd, Olsen, and Stekelberg 2015) and CEO gender (T. Huang, H. Huang, and Lee 2014). Overall, this stream of literature suggests that auditors consider CEO attributes that inﬂuence reporting quality in their pricing decisions. One CEO feature that is still unexplored in the context of auditing is how CEO’s functional experience in ﬁnance affects audit pricing. Few studies have examined the effect of CEO’s work experience in ﬁnance on ﬁrm performance and ﬁnancial reporting. Custodio and Metzger (2014) ﬁnd that ﬁrms that appoint ﬁnancial expert CEOs hold less cash, have better access to capital markets, and are less sensitive to cash-ﬂow shocks. Matsunaga et al. (2013) suggest that ﬁrms run by former CFOs report conservatively. In particular, Matsunaga et al. (2013) emphasize that the exposure of CEOs to ﬁnancial policies over their careers make them better monitors of ﬁnancial reporting quality. The increase in the appointment of CEOs with functional backgrounds in ﬁnance in recent years and the ﬁndings of these two above studies raise the question of whether and the extent to which the CEO’s ﬁnancial expertise is reﬂected in audit pricing. The existing literature shows that auditors charge higher fees when they perceive an increase in audit and business risk of an audit engagement (e.g., see Hay et al. 2006 for a review). Abbott, Parker, and Peters (2006) ﬁnd the audit fee to be positively associated with discretionary accruals (see also Gul, Chen, and Tsui 2003), and Feldmann, Read, and Abdolmohammadi (2009) ﬁnd that audit fees are higher for ﬁrms that restate their earnings. These authors attribute their results to increased perceived audit risk and a loss of organizational legitimacy. Other studies show that higher audit fees are associated with ﬁrms that disclose a material weakness in their internal controls (see Bedard, Ettredge, and Johnstone 2007; Hogan and Wilkins 2008; Raghunandan and Rama 2006). Additionally, DeFond, Lim, and Zang (2016) ﬁnd the audit fee to be negatively related to conservatism, while Bell, Landsman, and Shackelford (2002) and Bedard and Johnstone (2004) ﬁnd that a client’s business risk increases the fees auditors charge. As it pertains to engagement risk, the CEO’s ﬁnancial expertise is likely to be an important mitigating factor for two reasons. First, the ﬁnancial background of CEOs mitigates the risk of poor performance and risk of ﬁrm failure (Custodio and Metzger 2014), reducing a client’s business risk. Second, the ﬁnancial background of CEOs improves the quality of ﬁnancial reporting (Matsunaga et al. 2013), which reduces the probability of material misstatements, thus reducing the risk related to the audit engagement. The reductions of business and audit risks will translate into lower audit fees. We state our hypothesis as follows:
H1: After controlling for ﬁrm-level characteristics, the ﬁnancial expertise of a CEO is negatively associated with audit fees.
Although we predict a negative relationship between CEOs’ ﬁnancial expertise and audit fees, an alternative perspective is that a greater understanding of generally accepted accounting principles (GAAP) provides these CEOs with more avenues to misuse ﬂexibility allowed by GAAP (Demerjian, Lev, Lewis, and McVay 2013). If the use of discretion in ﬁnancial reporting increases engagement risk, then auditors will adjust audit fees to compensate for the increased risk, resulting in higher fees.
We start the analysis of the effect of CEOs’ ﬁnancial expertise on audit fees with a sample of nonﬁnancial and non-utility ﬁrms between 2004 and 2013. We use three different sources to obtain the data for the analysis: Compustat, Audit Analytics, and Execucomp. We use Compustat to calculate ﬁrm-speciﬁc variables, Audit Analytics for auditor-speciﬁc variables, and Execucomp for CEO-speciﬁc variables. We then hand-collect information about CEO background from Businessweek and Forbes. We are able to hand-collect CEO background information for 6,811 observations. To examine the effect of the appointment of ﬁnancial expert CEOs on audit fees, we use methodology similar to the managerial-effect model used in Bertrand and Schoar (2003). Speciﬁcally, we focus on ﬁrms that change their CEOs from
CEO Financial Background and Audit Pricing 327
Accounting Horizons Volume 30, Number 3, 2016
nonﬁnancial experts to ﬁnancial experts or from ﬁnancial experts to nonﬁnancial experts. For this purpose, we delete ﬁrms from our sample that do not use both ﬁnancial expert and nonﬁnancial expert CEOs during our sample period, resulting in a loss of 5,877 observations. This restriction leaves us with 934 ﬁrm observations. Furthermore, we require ﬁrms to have three-year observations before and after a CEO change. This restriction results in the deletion of 357 observations. We follow accounting literature for our choice of a three-year window (see Bamber, Jiang, and Wang 2010; McInnis and Collins 2011). This window is long enough to allow ﬁrms’ audit fees to adjust to the implications of ﬁnancial expert CEOs yet short enough to avoid picking up other economic factors that could affect ﬁrms’ audit fees (McInnis and Collins 2011). Another potential advantage of such analysis is that the events of ﬁnancial expert CEO appointments are scattered over time, thus allowing us to differentiate the impact of ﬁnancial expert CEOs from other economic events. Overall, our design tracks a ﬁrm over time and requires a ﬁrm to have both nonﬁnancial expert and ﬁnancial expert CEOs with at least three years of pre- and post-change data. With all of the restrictions, our sample consists of 577 observations from 77 ﬁrms. Out of these 77 ﬁrms, we ﬁnd that there are 69 ﬁrms that changed from nonﬁnancial expert CEOs to ﬁnancial expert CEOs and 12 ﬁrms that changed their CEOs with ﬁnancial expertise to nonﬁnancial expertise.3 Table 1, Panel A provides more information on the sample selection.4 In Panel B we present the frequency distribution of ﬁrms in our sample. Our analysis shows an average of 7.5 yearobservations per ﬁrm. The observations per ﬁrm range from six to ten years. The results show that almost 32 percent of the total ﬁrms have six years of observations while only 5 percent of ﬁrms have ten years of observations.
TABLE 1 Sample Selection
Panel A: Sample Selection
Firm Observations 2004–2013
Firm observations with valid CEO background information 6811 Less: ﬁrms with all years of nonﬁnancial expert CEO during the sample period (4664) Less: ﬁrms with all years of ﬁnancial expert CEO during the sample period (1213) Less: ﬁrms with fewer than three-year observations in pre- and post-post change (357)
Final Sample 577
Panel B Frequency Distribution of Firms No. of Years No. of Firms Frequency
6 25 32.47% 7 19 24.68% 8 7 9.09% 9 22 28.57% 10 4 5.19%
3 In order to be included in our sample, a ﬁrm must have changed its CFO from a non-ﬁnancial expert to ﬁnancial expert at least once during the sample period. There are four ﬁrms in our sample that appoint their CEOs more than once during the sample period and, therefore, the total number of changes are 81 (69þ12).4 Since our sample is small, which may result in the lack of generalizability of our result, we compare our ﬁrms to a broader sample. The broader dataset consists of ﬁrms after combining all the databases used in our study and after excluding our sample ﬁrms. This results in a broader sample of 13,687 ﬁrm-year observations. We compare the average ﬁrm size and the average audit fees of our sample with this broader dataset. We ﬁnd that the average total assets (proxy for ﬁrm size) of our sample ﬁrms are $7,435.57 million and for the broader sample is $7,061.89 million. The test of difference in means reveals that this difference is not signiﬁcantly different from zero (p-value¼0.66). We also compare the average audit fees (in thousands of dollars) of our sample ($3,673.02) with the broader sample ($3,312.09). We ﬁnd that audit fees in our sample are signiﬁcantly higher than in the broader sample (p-value¼0.06). In another test, we compare the ﬁrm size and audit fees of our sample ﬁrms with the industry average over the same period of time. We ﬁrst calculate industry average over the sample period time for the broader sample and then subtract the industry averages from our observations and test (t-test) whether the resulting value is different than zero. Again, we ﬁnd no difference in total assets (p¼0.26), but we ﬁnd statistically a signiﬁcant difference in audit fee; the difference is $717.26 and is signiﬁcant at less than 0.05.
328 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
Audit Fee Model
To test the hypothesis, we develop our model using the speciﬁcation from Munsif, Raghunandan, Rama, and Singhvi (2011). For brevity, we suppress ﬁrm and time subscripts. The model used is as follows: AFee ¼b0 þb1FinExpþb2Sizeþb3InvRecþb4Segmentsþb5Foreignþb6CurrRatioþb7GCþb8Distress þb9Big4þb10ICweakþb11Exordþb12AudChgþe: ð1Þ In the above model, we measure audit fees, AFee, as the natural logarithm of audit fees. Our variable of interest in the above speciﬁcation is FinExp, which takes a value of 1 for the ﬁrm-year observations when the CEO is a ﬁnancial expert and 0 otherwise. Following Custodio and Metzger (2014), we deﬁne CEO as a ﬁnancial expert if he or she worked in banking or investment ﬁrms, was an employee of an auditing ﬁrm, or worked as a CFO, treasurer, or vice president (VP) of ﬁnance. If auditors perceive a lower audit risk or expend less effort when the CEO of a ﬁrm is a ﬁnancial expert, then we expect a negative coefﬁcient on FinExp. Additionally, we include industry and year effect in the model. We calculate standard errors robust to heteroscedasticity and clustered by ﬁrm. We also consider a ﬁrm-ﬁxed effect model with standard errors clustered by ﬁrm. In this speciﬁcation, the coefﬁcient on FinExp is driven by changes in CEOs’ ﬁnancial background over time. The other variables included in the model control for audit risk, audit complexity, and auditor characteristics (Ashton, Willingham, and Elliott 1987; Ashton, Graul, and Newton 1989; Cushing 1989; Ng and Tai 1994; Simunic 1980). Prior studies document that audit risk increases with ﬁrm operations, internal control weakness, and poor ﬁrm performance (Higgs and Skantz 2006). Following these studies, we include ﬁrm size (Size), measured as a logarithm of total assets, and current ratio (CurrRatio), measured as a ratio of current assets to total assets to measure ﬁrm operations. To measure internal control weakness, we use ICweak measured as 1 if a ﬁrm has a material weakness in its internal control, else 0. For ﬁrm performance, we use two proxies: bankruptcy risk (Distress) and receipt of going concern opinion (GC). We measure Distress as Zmijewski’s (1984) probability of bankruptcy, and GC is 1 if the audit opinion is modiﬁed for going concern, else 0. Audit complexity increases audit fees (Ashton et al. 1987; Ashton et al. 1989; Cushing 1989; Ng and Tai 1994; Simunic 1980). To control for audit complexity, we include the ratio of accounts receivable and inventories to total assets (InvRec), the number of business segments (Segments), measured as the square root of the number of business segments, and an indicator variable for the client’s foreign operations (Foreign), measured as 1 if the ﬁrm has foreign operations and 0 otherwise. Following Munsif et al. (2011), we control for special events (Exord), a dummy variable that equals 1 if the ﬁrm reports extraordinary items or discontinued operations, else 0. Last, we include auditor size and auditor change to control for auditor characteristics. We measure auditor size (Big4), as an indicator variable that equals 1 if the auditor represents one of the Big 4 auditing ﬁrms and auditor change, AudChg, as an indicator variable equal to 1 if there was a change in auditor in the model.
Univariate Results Descriptive Statistics
In Table 2, Panel A we present the descriptive statistics of the variables used in the audit fee analysis. Our univariate results for the sample in Column 1 show that mean and median audit fees are $3,673,022 and $1,983,560, respectively. On average, 7 percent and 2 percent of the ﬁrms report internal control weakness and extraordinary items, respectively. The statistics on other variables used in the model reveal that the average size of the ﬁrms, measured in terms of total assets, is $7,435,570, 40 percent of the ﬁrms are engage in foreign activities, and 1 percent of the ﬁrms receive going concern opinions. The analysis on auditor characteristics shows that 90 percent of ﬁrms use Big 4 auditors and 3 percent of ﬁrms change their auditor. We divide the sample into ﬁrm-observations with ﬁnancial expert CEO and nonﬁnancial expert CEO and present the mean and median statistics in Columns 2 and 3, respectively. The results show that mean (median) audit fees (in thousands of dollars) for the ﬁnancial expert CEO sample is 3,392.34 (1,754.34) and 3,706.46 (2,004.69) for nonﬁnancial expert CEOs. This suggests that the audit fees are lower when ﬁnancial expert CEOs manage ﬁrms. Additionally, the results show that ﬁrmobservations with ﬁnancial expert CEOs have lower internal control weakness, and the difference is signiﬁcant at less than a 5 percent signiﬁcance level. The summary statistics further reveal that observations with ﬁnancial expert CEOs are larger ﬁrms and have better ﬁrm performance than observations with nonﬁnancial expert CEOs. Last, ﬁrm-years with ﬁnancial expert CEOs are less likely to appoint Big 4 auditors and are less likely to change auditors.
CEO Financial Background and Audit Pricing 329
Accounting Horizons Volume 30, Number 3, 2016
In Panel B of Table 2 we present a distribution of different classiﬁcations of ﬁnancial expert CEO. Our results indicate that 12 percent of the CEOs have worked as CPAs, 34 percent have worked in a ﬁnancial industry, 67 percent have worked as CFOs, 25 percent have worked as treasurers, and 19 percent have worked as vice presidents of ﬁnance. The total of the distribution is greater than 100 percent as some of the CEOs have multiple classiﬁcations.
TABLE 2 Summary Statistics
Panel A: Descriptive Analysis
(1) All Observations
(2) Financial Expert
(3) Nonﬁnancial Expert
MeanMedianMeanMedianMeanMedianDifference AFee 3,673.02 1,983.56 3,392.34 1,754.34 3,706.46 2,004.69 314.12 Size 7,435.57 1,916.82 7,719.20 1,963.90 7,184.30 1,845.27 534.90 InvRec 0.22 0.22 0.23 0.23 0.21 0.21 0.02 Segments 2.15 1.00 2.08 1.00 2.20 1.00 0.12 Foreign 0.40 0.00 0.43 0.00 0.37 0.00 0.06 CurrRatio 2.42 1.69 2.33 1.72 2.50 1.63 0.17 GC 0.01 0.00 0.00 0.00 0.02 0.00 0.02 Distress 2.29 1.24 1.95 1.04 2.59 1.41 0.64** Big4 0.90 1.00 0.88 1.00 0.93 1.00 0.05** ICweak 0.07 0.00 0.02 0.00 0.12 0.00 0.1** Exord 0.02 0.00 0.00 0.00 0.04 0.00 0.04*** AudChg 0.03 0.00 0.02 0.00 0.05 0.00 0.03*
Panel B: Classiﬁcations of Financial Experts Banking Financial Industry Former CFO Treasurer VP of Finance
FinExp 0.34 0.12 0.67 0.25 0.19
Panel C: Time-Series of Audit Fees and Financial Expertise
Size-Adjusted Audit Fees Log of Audit Fees
Mean Std. Dev. Median Mean Std. Dev. Median t 3 1,502.84 1,191.11 1,094.45 14.69 1.02 14.7 t 2 1,531.46 1,651.40 1,002.26 14.54 1.19 14.55 t 1 1,769.30 2,067.22 972.52 14.58 1.04 14.5 t 1,617.81 1,746.90 966.04 14.55 1.06 14.42 tþ1 1,565.41 2,000.36 850.33 14.51 1.07 14.56 tþ2 1,397.90 1,553.56 769.53 14.59 1.06 14.51
Panel D: Test of Equality between Pre- and Post-CEO Appointment Audit Fee
Size-Adjusted AF Log AF Pre-turnover year ( 1) versus Post-turnover year (þ1) Difference 203.89 0.07 p-value 0.402 0.912
***, **, * Denotes signiﬁcance at the 1 percent, 5 percent, and 10 percent levels, respectively. This table provides summary statistics for all, ﬁnancial expert, and nonﬁnancial expert ﬁrm-observations in Panel A. The Difference column represents differences between variables in ﬁnancial expert and nonﬁnancial expert groups. Panel B provides different classiﬁcations of ﬁnancial expert CEO. Panel C presents time-series analyses of audit fees in the pre- and post-ﬁnancial expert CEO’s appointment period, and Panel D presents p-values from parametric t-tests of change in audit fees from nonﬁnancial expert CEO period (year t 1) to ﬁnancial expert CEO period (year tþ1). Audit fees reported in the table are in thousands of dollars. Difference reﬂects the reduction in fee from year 1 to yearþ1.
330 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
Panel C of Table 2 shows a trend in audit fees where t is the ﬁrst year of a ﬁnancial expert CEO being in charge. Since ﬁrm size inﬂuences audit fees, we adjust the audit fee for ﬁrm size by scaling the audit fees by total assets or by taking the logarithm of audit fees. We restrict our analysis only to those ﬁrms that change their CEO from a nonﬁnancial expert to a ﬁnancial expert. For the size-adjusted audit fees, we ﬁnd that audit fees generally increase leading up to the appointment of ﬁnancial expert CEOs but decline subsequently. However, we fail to ﬁnd a similar trend for the log of audit fees. Panel D of Table 2 provides the p-value from t-tests of change in audit fees from nonﬁnancial expert CEO period (year t 1) to ﬁnancial expert CEO period (year tþ1). Although, the change in scaled audit fees from year t 1 to year tþ1 is about 11.5 percent ([1,769.30 1,565.41]/1,769.30), this change is not statistically signiﬁcant (two-tailed p . 0.05). A similar result persists when the log of audit fees is used.5 However, the univariate t-tests of change do not account for other confounding variables, and thus may be biased to the extent other confounding variables affect audit fee. To overcome these shortcomings, we use multivariate regression analyses.
In Table 3, we present the correlation among the variables used in our analysis. The results show that the correlation between the audit fee and a ﬁnancial expert CEO is negative (coefﬁcient¼ 0.04). This indicates that auditors charge lower fees when ﬁrms are managed by ﬁnancial expert CEOs. Consistent with prior studies, we ﬁnd AFee to be positively associated with Size (coefﬁcient¼0.81), InvRec (coefﬁcient¼0.04), Segments (coefﬁcient¼0.10), Foreign (coefﬁcient¼0.47), and Big4 (coefﬁcient¼0.23) and negatively associated with CurrRatio (coefﬁcient¼ 0.34), Distress (coefﬁcient¼ 0.35), and AudChg (coefﬁcient¼ 0.09). We also ﬁnd FinExp to be negatively and signiﬁcantly associated with Distress (coefﬁcient¼ 0.10), Big4 (coefﬁcient¼ 0.08), ICweak (coefﬁcient¼ 0.17), Exord (coefﬁcient¼ 0.11), and AudChg (coefﬁcient¼ 0.09), thus indicating that ﬁrms with ﬁnancial expert CEOs have lower audit risk and are less likely to switch their auditors and appoint Big 4 auditors.
Regression Results Audit Fees and Financial Expert CEOs
Table 4 provides the results from the regressions with the log of audit fees as the dependent variable. Column 1 (base model) includes our CEOs’ ﬁnancial expertise variable (FinExp) and standard ﬁrm-level controls. Our main variable of interest,
TABLE 3 Correlations AFee FinExp Size InvRec Segments Foreign CurrRatio GC Distress Big4 ICweak Exord AudChg
AFee 1 FinExp 0.04 1 Size 0.81* 0.05 1 InvRec 0.04 0.06 0.05 1 Segments 0.10* 0.02 0.01 0.12* 1 Foreign 0.47* 0.08 0.25* 0.06 0.02 1 CurrRatio 0.34* 0.04 0.33* 0.11* 0.01 0.02 1 GC 0.02 0.06 0.00 0.09* 0.05 0.02 0.07 1 Distress 0.35* 0.10* 0.36* 0.08 0.06 0.04 0.62* 0.07 1 Big4 0.23* 0.08* 0.26* 0.20* 0.00 0.15* 0.05 0.03 0.06 1 ICweak 0.02 0.17* 0.11* 0.06 0.04 0.12* 0.05 0.04 0.04 0.04 1 Exord 0.07 0.11* 0.03 0.00 0.01 0.03 0.06 0.10* 0.06 0.00 0.01 1 AudChg 0.09* 0.09* 0.10* 0.07 0.03 0.05 0.01 0.02 0.05 0.05 0.10* 0.04 1 * Variable is statistically signiﬁcant at 5 percent. This table provides pairwise correlations between variables used in the base model. See Appendix A for variable deﬁnitions.
5 We also compare audit fees for the three years before and three years after the ﬁnancial expert CEO appointment. We ﬁnd that the average change in the size-adjusted audit fees from year t 3 to year tþ3 is statistically signiﬁcant (p-value¼0.06). We do not ﬁnd any difference between pre- and postappointment period when we use the log of audit fee (p-value¼0.87).
CEO Financial Background and Audit Pricing 331
Accounting Horizons Volume 30, Number 3, 2016
FinExp, is signiﬁcant and in the predicted direction (b¼ 0.176; p , 0.01). This result suggests that, on average, ﬁrms with ﬁnancial expert CEOs pay approximately 17 percent lower audit fees. The sign on the coefﬁcients of other control variables are consistent with existing literature. Speciﬁcally, we ﬁnd that audit fees increase with client size and other audit risk factors. We also ﬁnd coefﬁcients on Segments and Foreign to be positive, suggesting that complexity of business is positively associated with audit fees. Additionally, we ﬁnd that audit fees are signiﬁcantly and negatively associated with CurrRatio and positively associated with Size and ICweak, thus indicating that audit fees increase with audit risk. The coefﬁcients on auditor-speciﬁc variables—such as auditor size and auditor change—are insigniﬁcant however. Overall, the results from Column 1 indicate that ﬁrms whose CEOs’ background is in ﬁnance pay signiﬁcantly lower audit fees. In Column 2, we add a dummy for a potential exogenous turnover (Departure) and an interaction term (FinExp  Departure) to our base model. We identify possible exogenous turnovers if a CEO turnover is caused by death or health issues (Fee, Hadlock, and Pierce 2013), move elsewhere in a same position (Brochet, Faurel, and McVay 2011), or retirement (Custodio and Metzger 2014).6 Thus, we create a dummy variable that equals 1 if CEO turnover is caused by any of the above reasons and 0 otherwise. This variable, to some degree, helps us separate two groups of turnovers: situations in which nonﬁnancial (ﬁnancial) expert CEOs are replaced by the board of directors speciﬁcally to appoint ﬁnancial (nonﬁnancial) expert CEOs and situations where a CEO either voluntarily retires or takes a position at another ﬁrm and the appointment of a CEO is potentially unrelated to a ﬁrm’s internal demand. We assume the latter situation to be exogenous for the purpose of our analyses; however, we acknowledge the imperfectness of our measures as, in many cases, these turnovers may not necessarily be strictly exogenous.
TABLE 4 Financial Expert and Audit Fees Log of Audit Fees
(1) Coefﬁcient t-stat
(2) Coefﬁcient t-stat
(3) Coefﬁcient t-stat
(4) Coefﬁcient t-stat
Intercept 10.550*** (28.57) 9.930*** (28.16) 9.869*** (26.48) 11.276*** (25.87) FinExp 0.176*** ( 4.99) 0.165*** ( 3.64) 0.221*** ( 3.48) 0.083** ( 2.61) Size 0.449*** (12.07) 0.453*** (11.88) 0.447*** (11.85) 0.392*** (8.90) InvRec 0.594 (1.27) 0.629 (1.26) 0.640 (1.30) 0.869 (1.54) Segments 0.208** (2.24) 0.213** (2.32) 0.213** (2.28) 0.080 (1.24) Foreign 0.497*** (5.25) 0.492*** (4.92) 0.485*** (5.08) 0.070 (1.09) CurrRatio 0.051*** ( 3.04) 0.051*** ( 3.03) 0.053*** ( 3.17) 0.023*** ( 3.22) GC 0.232 ( 1.17) 0.235 ( 1.20) 0.232 ( 1.12) 0.018 (0.15) Distress 0.005 ( 0.55) 0.006 ( 0.56) 0.004 ( 0.41) 0.006 ( 1.10) Big4 0.093 (1.01) 0.092 (1.02) 0.115 (1.08) 0.053 (0.27) ICweak 0.233** (2.56) 0.238** (2.55) 0.178** (2.04) 0.126 (1.63) Exord 0.064 (0.64) 0.064 (0.66) 0.070 (0.68) 0.013 (0.16) AudChg 0.036 (0.28) 0.040 (0.31) 0.045 (0.34) 0.033 (0.28) Departure 0.015 ( 0.13) FinExp  Departure 0.032 ( 0.47) CFO Change 0.082 (0.76) FinExp  CFO Change 0.048 (0.60) Fixed Effects Industry and Year Industry and Year Industry and Year Firm No of Obs. 577 577 577 577 Adj. R2 85.50% 85.40% 85.40% 74.19%
***, **, * Denotes signiﬁcance at the 1 percent, 5 percent, and 10 percent levels, respectively. This table provides results from regressing log of audit fees on ﬁnancial expert CEO, a dummy variable that equals 1 if the ﬁrm is headed by a ﬁnancial expert CEO, and various control variables. The data cover the period from 2004–2013. All models include industry and year dummies. Standard errors are robust to heteroscedasticity and clustering by ﬁrm. All tests are two-sided and t-statistics are provided in parentheses. Column 1 presents the result of the base model. Column 2 and Column 3 additionally control for CEO turnover type and a simultaneous CFO change, respectively. Column 4 presents a regression with ﬁrm-ﬁxed effect model. See Appendix A for variable deﬁnitions.
6 None of the turnovers in our sample occurs due to the death of a CEO.
332 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
We repeat our analysis and ﬁnd that FinExp continues to remain negatively signiﬁcant (b ¼ 0.165; p , 0.01). Additionally, we do not ﬁnd any signiﬁcant association between audit fees and the interaction of FinExp  Departure. This result suggests that the effect of ﬁnancial expert CEOs on audit fees are not conditional on the type of CEO turnovers. That is, the change in audit fees due to the presence of a ﬁnancial expert CEO does not depend on whether an outgoing CEO voluntarily leaves his position or is replaced by a ﬁnancial (nonﬁnancial) expert CEO due to some unobserved demand for such an individual within the ﬁrm. Thus, our result of a negative association between audit fees and ﬁnancial expert CEOs is less likely to be driven by the different types of CEO departure. In Column 3, we control for CFO turnover. Some studies in the recent past have argued that CEOs are responsible for reporting quality while other studies have documented that CFOs have greater inﬂuence over ﬁnancial reporting. Kannan, Skantz, and Higgs (2014) ﬁnd that auditors charge higher fees following an increase in CFO equity incentives, suggesting that auditors perceive higher audit risk associated with CFO equity incentives, while Geiger and North (2006) ﬁnd that a negative association between discretionary accruals and the appointment of a new CFO. Additionally, ﬁnancial expert CEOs are more likely to replace the current CFO in order to gain more control with respect to implementing ﬁnancial policies (Custodio and Metzger 2014). Thus, a possibility exists that a reduction in audit fees following the appointment of a ﬁnancial expert CEO is driven by the appointment of a new CFO. If CEO and CFO changes are simultaneous, then it is important to separate the effect of these two changes on audit fees to provide a cleaner inference related to our main variable (FinExp). Although we do not intend to examine who is more responsible for reporting quality, we do ensure that simultaneous CFO changes do not determine our reported association between CEOs’ ﬁnancial expertise and audit fees. To separate the effect of new CFOs on ﬁnancial reporting quality, we rerun our base model after including a dummy for ﬁrms that changed CFOs during the sample period (CFO Change). Speciﬁcally, the dummy takes a value of 1 for the periods t, tþ1, and tþ2 if a ﬁrm changes its CFO. Additionally, we also include an interaction between CEO ﬁnancial expertise and the CFO change variable (FinExp  CFO Change). Our result in Column 3 shows that FinExp is still negatively and signiﬁcantly associated with audit fees. Additionally, we ﬁnd the coefﬁcient on the interaction term to be positive but insigniﬁcant. In Column 4, we present results using a ﬁrm-ﬁxed effect model to control for unobserved ﬁrm effects. We use this speciﬁcation because it is possible that some unobserved ﬁrm-speciﬁc quality is correlated with our explanatory variable (i.e., audit fee), which may bias our inferences. Our results using the ﬁxed-effect model continues to remain same (b¼ 0.083, p , 0.05); however, the magnitude of the coefﬁcient on FinExp reduces from the ﬁrst three columns. This highlights that some of the variation in audit fees reported in the ﬁrst three columns is due to unobserved ﬁrm characteristics. In terms of economic signiﬁcance, we observe a reduction of about 8.3 percent or about $310,000 in audit fees for ﬁrms with ﬁnancial expert CEOs. Overall, the results are consistent across all four models, suggesting that work experience makes managers more effective in handing challenges in related areas. In particular, ﬁnancial background of CEOs affects auditors’ pricing decision.7
Instrumental Variable Regression
In Table 5, we report results from an instrumental variable (IV) approach. We use endogenous treatment effect speciﬁcation to implement instrumental variable regressions (see Maddala 1983).8 We use the local density of ﬁnancial ﬁrms (Density) as our instrument in the ﬁrst stage. We deﬁne Density as the log of the number of ﬁnancial ﬁrms available within a 100-mile radius from the focal ﬁrm. Alam, Chen, Ciccotello, and Ryan (2014) and A. Knyazeva, D. Knyazeva, and Masulis (2013) show that the geographical location of ﬁrms inﬂuences board composition. Mobbs (2014) suggests that ﬁrms located near the larger pool of ﬁnancial ﬁrms have access to more ﬁnancial experts and are thus more likely to have outside expert
7 Since our tests use CEO turnover with the change in their ﬁnancial expertise, there is a possibility that our results are driven by CEO turnover and not by change in CEO ﬁnancial expertise. To rule out this possibility, in an unreported regression, we conduct a falsiﬁcation test around a CEO turnover that does not involve a change in the ﬁnancial expertise of a CEO. We rerun Model (1) with one difference. Model 1 includes CEO turnover with change in ﬁnancial expertise (FinExp), while our falsiﬁcation test includes a CEO turnover variable with no changes in expertise. Our analysis, that includes 73 ﬁrms with 577 ﬁrm-year observations and 73 CEO turnovers, reveals that no difference exists in audit fee between pre- and post-CEO turnover. Speciﬁcally, we ﬁnd that although the coefﬁcient on CEO turnover is negative, it is statistically insigniﬁcant (b¼ 0.01, p¼0.92). This suggests that our results are not driven by CEO turnovers. 8 We use a full sample for which we hand-collected ﬁnancial expert data and for which all relevant variables are available to conduct instrumental variable regression, resulting in 6,600 observations. We do this (as opposed to restricting our analyses to managerial ﬁxed-effect [or change] sample), because we need enough variations both in our dependent variable (dummy for ﬁnancial expert) and our instrument (the density of ﬁnancial ﬁrm). Since our instrument is sticky, as ﬁrms do not change their headquarters frequently, using the restricted sample does not allow us to properly estimate the ﬁrst stage model, because all ﬁrms in the restricted sample have both a ﬁnancial expert and a nonﬁnancial expert within a panel. Thus, our ability to separate the likelihood of a ﬁrm, given its location, appointing a ﬁnancial expert CEO is severely curtailed in the restricted sample. Therefore, to properly estimate the likelihood of a ﬁrm appointing a ﬁnancial expert CEO, we require a full sample that includes both ﬁrms that never appointed a ﬁnancial expert CEO during our sample period and ﬁrms that have.
CEO Financial Background and Audit Pricing 333
Accounting Horizons Volume 30, Number 3, 2016
directors on board, thereby reducing the demand for a ﬁnancial expert from inside of the ﬁrm to serve on the board. Adapting Mobbs’s argument, we suggest that if a ﬁrm is located near a large pool of ﬁnancial ﬁrms, then it can fulﬁll its ﬁnancial expert needs by appointing these experts as directors to its board. Thus, the likelihood of a ﬁnancial expert CEO should be negatively associated with the local density of ﬁnancial ﬁrms. The density of ﬁnancial ﬁrms is a valid instrument, because it is negatively related to the selection of a ﬁnancial expert CEO but not systematically associated with ﬁrm-speciﬁc policies such as accruals and audit fees. In Table 5, Column 1 we present results from the ﬁrst stage regression. The results show that after netting out other control variables, the local density of ﬁnancial ﬁrms is negatively associated with the selection of ﬁnancial expert CEOs (p , 0.05). The results further reveal an F-statistic of 1,716.05 (p , 0.01), which is signiﬁcantly greater than the critical value of 10. In Column 2, we present results from the second stage, which show that, after controlling for all other factors, the ﬁnancial expertise is negatively associated with audit fees (p , 0.01), although the magnitude is larger than expected.9 The signs on the coefﬁcients of other control variables are consistent with existing literature.
Improvement in Reporting Information
The results from Tables 4 and 5 indicate that ﬁrms with ﬁnancial expert CEOs pay lower audit fees. To shed more light on a possible cause, we examine, in an unreported regression, whether ﬁrms with ﬁnancial expert CEOs report better quality of earnings. Caramanis and Lennox (2008) ﬁnd an inverse relationship between audit effort and earnings quality (see also Kinney and McDaniel 1993; Mitra, Deis, and Hossain 2009; Shibano 1990). It follows that improvement in the perceived earnings quality is likely to reduce audit effort. We use absolute value of discretionary accruals, calculated using Kothari, Leone, and
TABLE 5 Instrumental Variable Regression FinExp Log of Audit Fees
(1) Coefﬁcient Z-stat
(2) Coefﬁcient Z-stat
Intercept 1.714* (1.77) 9.795*** (30.66) Density 0.269** ( 2.29) FinExp 0.587*** ( 5.92) Size 0.0275 (0.96) 0.522*** (38.04) InvRec 0.477 ( 1.61) 0.571*** (3.93) Segments 0.043 ( 0.62) 0.108*** (3.97) Foreign 0.129* (1.66) 0.222*** (7.02) CurrRatio 0.049** ( 2.07) 0.031*** ( 3.50) GC 0.343 ( 1.19) 0.050 ( 0.47) Distress 0.002 ( 0.20) 0.007* ( 1.65) Big4 0.127 ( 0.92) 0.122** (2.29) ICweak 0.148 ( 1.29) 0.371*** (6.35) Exord 0.055 (0.41) 0.136** (2.52) AudChg 0.173 ( 1.56) 0.267*** ( 4.59) Fixed Effects Industry and Year Industry and Year No of Obs. 6,660 6,660 Log likelihood 3,633.33 8,209.96 ***, **, * Denotes signiﬁcance at the 1 percent, 5 percent, and 10 percent levels, respectively. This table presents results from instrumental variable approach. Column 1 presents the results from the selection model and Column 2 reports results of the outcome model. The data cover the period from 2004–2013. All models include industry and year dummy. Standard errors are robust to heteroscedasticity and clustering by ﬁrm. All tests are two-sided and z-statistics are provided in parentheses. See Appendix A for variable deﬁnitions.
9 This larger than expected coefﬁcient on FinExp in the second stage is probably due to problems with a weak instrument in a ﬁnite sample (see Hahn and Hausman 2003).
334 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
Wasley’s (2005) performance-matched modiﬁed Jones model (Dac), to proxy for reporting quality.10 Our unreported result conﬁrms the ﬁndings of Matsunaga et al. (2013) and indicates that ﬁrms with ﬁnancial expert CEOs report better quality of earnings (b¼ 0.011; p , 0.05). This result provides some explanation that auditors perceive earnings quality to be better when ﬁrms are headed by ﬁnancial expert CEOs and possibly reduce audit effort as reﬂected in lower audit fees.
Inside versus Outside Appointment
We also consider whether the effect of ﬁnancial expertise of CEOs on audit fees differs between internal and external appointments of ﬁnancial expert CEOs. We include an interaction variable between ﬁnancial expert CEOs (FinExp) and a dummy variable that equals 1 if the ﬁnancial expert CEO is recruited from outside (Outsider) the ﬁrm in our primary audit fee speciﬁcation (Model 1). The coefﬁcient on the interaction term is the incremental effect of a ﬁnancial expert CEO if he or she is recruited from outside the ﬁrm. In unreported results, we ﬁnd a positive but insigniﬁcant coefﬁcient on the interaction term (FinExpOutsider). The result suggests that no signiﬁcant difference exists in audit fee reduction between internally promoted and externally recruited ﬁnancial expert CEOs.
We consider a couple of additional analyses to conﬁrm the robustness of our reported results. In the ﬁrst test we control for managerial ability. Demerjian, Lev, and McVay (2012) show that ﬁrms with better managers provide higher quality of earnings. If CEOs’ ﬁnancial background is correlated with ﬁrm-speciﬁc managerial ability, then our results may be biased due to an omitted-variable problem. To address this concern, we include an industry-year decile rank of managerial ability as calculated by Demerjian et al. (2012).11 Inclusion of this managerial ability variable results in the loss of 90 observations, resulting in 487 observations. The inclusion of managerial ability does not change our results related to ﬁnancial expert CEOs. For example, the coefﬁcient on FinExp in our primary audit fees model remains negative and signiﬁcant (b¼ 0.177, p , 0.01). Additionally, we ﬁnd that managerial ability is signiﬁcantly and negatively associated with audit fees, results that are consistent with the ﬁndings of Krishnan and Wang (2015). Therefore, our results of ﬁnancial expert CEOs are in addition to the ﬁrm-speciﬁc managerial ability. In our last test, we control for audit committee characteristics. The role of audit committees in monitoring reporting quality is well documented. Existing literature shows that auditors charge lower fees when the audit committee monitoring is diligent (Abbott et al. 2003; Chang, Chen, and Zhou 2013; Goodwin-Stewart and Kent 2006; Lee and Mande 2005; Vafeas and Waegelein 2007), when the audit committee is more ﬁnancially literate (e.g., Krishnan and Visvanathan 2008), and when it is chaired by a female (Ittonen, Miettinen, and Va¨ha¨maa 2010). Following these studies, we control for audit committee characteristics in our model. We include the percentage of female members, average director tenure, size of the audit committee (number of audit committee members), and average age of the audit committee members. Risk Metrics database provides these variables. Inclusion of these audit committee variables in the model results in the loss of 386 ﬁrm-year observations. However, our results remain consistent even for the smaller sample. In particular, the coefﬁcient on FinExp remains negative and signiﬁcant (b ¼ 0.165, p , 0.01). Additionally, we ﬁnd that director tenure is negative and signiﬁcantly associated with audit fees, and average female board membership is positively and signiﬁcantly associated with audit fees. Our results are consistent with Gul, Srinidhi, and Tsui (2008), suggesting that female directors require higher audit effort and therefore result in higher audit fees.
We consider the importance of ﬁnancial expert CEOs on audit fees. We examine whether the ﬁnancial background of CEOs results in an improved trustworthiness of the ﬁnancial reporting process, which can potentially reduce audit effort and audit risk as perceived by auditors. Our expectations are based on the upper echelon theory, which predicts that training and experience in a particular domain improves a manager’s performance in related areas (Hambrick and Mason 1984). Consistent with this upper echelon theory, Matsunaga et al. (2013) document an improvement in the quality of earnings and Custodio and Metzger (2014) document an increase in ﬁrm proﬁtability and a reduction in ﬁrm failure, following the appointment of a ﬁnancial expert CEO. The ﬁndings of these studies thus suggest that auditors’ engagement risk decreases with the appointment of ﬁnancial expert CEOs, raising a prospect that audit fees would be lower for these ﬁrms.
10 We follow a speciﬁcation similar to the one used in Jiang, Petroni, and Wang (2010) for the choice of our control variables to be included in the regression model. Speciﬁcally, we include ﬁrm size, leverage, cash-ﬂow volatility, sales volatility, sales growth volatility, a dummy that equals 1 if the ﬁrm is 20 years old, the log of the number of analysts following, and an entrenchment index developed by Bebchuk, Cohen, and Ferrell (2008). 11 Please see Demerjian et al. (2012) for the calculation of this variable.
CEO Financial Background and Audit Pricing 335
Accounting Horizons Volume 30, Number 3, 2016
Our analyses indicate that ﬁrms where CEOs possess a background in ﬁnance pay lower audit fees, suggesting that auditors consider such ﬁrms to have a lower engagement risk. The results highlight that auditors consider ﬁnancial background of CEOs to be an important determinant of audit fees. Our results are robust to a number of alternative speciﬁcations. The results also hold after controlling for managerial ability, CFO changes, and audit committee characteristics. The instrumental variable and ﬁrm-ﬁxed effect regressions further conﬁrm that ﬁrms with ﬁnancial expert CEOs pay lower audit fees. Overall, our result adds to the literature on the advantages and disadvantages of a functional background of top managers and how this background can create value for a ﬁrm through savings in audit fees.
Abbott, L. J., S. Parker, and G. F. Peters. 2006. Earnings management, litigation risk, and asymmetric audit fee responses. Auditing: A Journal of Practice & Theory 25 (1): 85–98. Abbott, L. J., S. Parker, G. F. Peters, and K. Raghunandan. 2003. The association between audit committee characteristics and audit fees. Auditing: A Journal of Practice & Theory 22 (2): 17–32. Alam, Z., M. Chen, C. Ciccotello, and H. Ryan. 2014. Does the location of directors matter? Information acquisition and board decisions. Journal of Financial and Quantitative Analysis 49 (1): 131–164. Amit, R., and P. Shoemaker. 1993. Strategic assets and organizational rent. Strategic Management Journal 14: 33–46. Ashton, R. H., J. J. Willingham, and R. K. Elliott. 1987. An empirical analysis of audit delay. Journal of Accounting Research 2: 275– 292. Ashton, R. H., P. R. Graul, and J. D. Newton. 1989. Audit delay and the timeliness of corporate reporting. Contemporary Accounting Research 2: 657–673. Baik, B., P. Brockman, D. B. Farber, and S. Lee. 2011. CEO Reputation and Corporate Opacity. Working paper, Seoul National University. Bamber, L. S., J. Jiang, and I. Y. Wang. 2010. What’s my style? The inﬂuence of top managers on voluntary corporate ﬁnancial disclosure. The Accounting Review 85 (4): 1131–1162. Bebchuk, L., A. Cohen, and A. Ferrell. 2008. What matters in corporate governance? Review of Financial Studies 22 (2): 783–827. Becker, G. S. 1962. Investment in human capital: A theoretical analysis. Journal of Political Economy 70 (5): 9–49. Bedard, J. C., and K. M. Johnstone. 2004. Earnings manipulation risk, corporate governance risk, and auditors’ planning and pricing decisions. The Accounting Review 79 (2): 277–304. Bedard, J. C., M. L. Ettredge, and K. M. Johnstone. 2007. Using electronic audit workpaper systems in audit practice: Task analysis, learning, and resistance. Advances in Accounting Behavioral Research 10: 29–53. Bell, T., W. Landsman, and D. Shackelford. 2002. Auditors’ perceived business risk and audit fees: Analysis and evidence. Journal of Accounting Research 39: 35–43. Bergstresser, D., and T. Philippon. 2006. CEO incentives and earnings management. Journal of Financial Economics 80: 511–529. Bertrand, M., and A. Schoar. 2003. Managing with style: The effect of managers on ﬁrm policies. The Quarterly Journal of Economics 118 (4): 1169–1208. Betz, M., L. O’Connell, and J. M. Shepard. 1989. Gender differences in proclivity for unethical behavior. Journal of Business Ethics 8: 321–324. Bhagat, S., B. J. Bolton, and A. Subramanian. 2010. CEO Education, CEO Turnover, and Firm Performance. Working paper, University of Colorado, Portland State University, and Georgia State University. Boyd, D. E., R. Chandy, and M. Cunha, Jr. 2010. When do chief marketing ofﬁcers impact value? A customer power explanation. Journal of Marketing Research 47: 1162–1176. Brochet, F., L. Faurel, and S. McVay. 2011. Manager-speciﬁc effects on earnings guidance: An analysis of top executive turnovers. Journal of Accounting Research 49: 1123–1162. Burns, N., and S. Kedia. 2006. The impact of performance-based compensation on misreporting. Journal of Financial Economics 79: 35– 67. Butz, C. E., and P. V. Lewis. 1996. Correlation of gender-related values of independence and relationship and leadership orientation. Journal of Business Ethics 15: 1141–1149. Cao, Z., and G. Narayanamoorthy. 2014. Accounting and litigation risk: Evidence from directors’ and ofﬁcers’ insurance pricing. Review of Accounting Studies 19 (1): 1–42. Caramanis, C., and C. Lennox. 2008. Audit effort and earnings management. Journal of Accounting and Economics 45 (1): 116–138. Chang, H., X. Chen, and N. Zhou. 2013. Determinants and Consequences of Audit Committee Effectiveness: Evidence from China. Working paper, Drexel University, Shanghai Jiao Tong University, and Binghamton University–SUNY. Cheng, Q., and T. Warﬁeld. 2005. Equity incentives and earnings management. The Accounting Review 80: 441–476. Coff, R. W. 2002. Human capital, shared expertise, and the likelihood of impasse in corporate acquisitions. Journal of Management 28: 107–128. Cohen, M. D., J. G. March, and J. P. Olsen. 1972. A garbage can model of organizational choice. Administrative Science Quarterly: 1–25.
336 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
Cushing, B. 1989. Discussion of the association between audit technology and audit delay. Auditing: A Journal of Practice & Theory 8: 38–47. Custodio, C., and D. Metzger. 2014. Financial expert CEOs: CEO’s work experience and ﬁrm’s ﬁnancial policies. Journal of Financial Economics 114 (1): 125–154. DeFond, M. L., C. Y. Lim, and Y. Zang. 2016. Client conservatism and auditor-client contracting. The Accounting Review 91 (1): 69–88. Demerjian, P., B. Lev, and S. McVay. 2012. Quantifying managerial ability: A new measure and validity tests. Management Science 58 (7): 1229–1248. Demerjian, P., B. Lev, M. Lewis, and S. McVay. 2013. Managerial ability and earnings quality. The Accounting Review 88 (2): 463–498. Fee, C., C. Hadlock, and J. Pierce. 2013. Managers with and without style: Evidence using exogenous variation. Review of Financial Studies 23 (3): 567–601. Feldmann, D. A., W. J. Read, and M. J. Abdolmohammadi. 2009. Financial restatements, audit fees, and the moderating effect of CFO turnover. Auditing: A Journal of Practice & Theory 28: 205–223. Geiger, M. A., and D. S. North. 2006. Does hiring a new CFO change things? An investigation of changes in discretionary accruals. The Accounting Review 81: 781–809. Goodwin-Stewart, J., and P. Kent. 2006. Relation between external audit fees, audit committee characteristics and internal audit. Accounting and Finance 46 (3): 387–404. Gul, F. A., C. J. Chen, and J. S. Tsui. 2003. Discretionary accounting accruals, managers’ incentives, and audit fees. Contemporary Accounting Research 20 (3): 441–464. Gul, F. A., B. Srinidhi, and J. S. Tsui. 2008. Board Diversity and the Demand for Higher Audit Effort. Working paper, Monash University, The University of Texas at Arlington, and The Hong Kong Polytechnic University. Hahn, J., and J. Hausman. 2003. Weak instruments: Diagnosis and cures in empirical econometrics. American Economic Review 93: 118– 125. Hambrick, D. C., and P. A. Mason. 1984. Upper echelons: The organization as a reﬂection of its top managers. Academy of Management Review 9: 193–206. Hannan, M. T., and J. Freeman. 1977. The population ecology of organizations. American Journal of Sociology 82 (5): 929–964. Harris, D., and C. Helfat. 1997. Speciﬁcity of CEO human capital and compensation. Strategic Management Journal 18 (11): 895–920. Hay, D. C., W. R. Knechel, and N. Wong. 2006. Audit fees: A meta-analysis of the effect of supply and demand attributes. Contemporary Accounting Research 23: 141–191. Hayes, R. M., and S. Schaefer. 2000. Implicit contracts and the explanatory power of top executive compensation for future performance. The RAND Journal of Economics 31 (2): 273–293. Higgs, J. L., and T. R. Skantz. 2006. Audit and nonaudit fees and the market’s reaction to earnings announcements. Auditing: A Journal of Practice & Theory 25: 1–26. Hitt, M. A., and R. D. Ireland. 1985. Corporate distinctive competence, strategy, industry and performance. Strategic Management Journal 6: 273–293. Hogan, C. E., and M. S. Wilkins. 2008. Evidence on the audit risk model: Do auditors increase audit fees in the presence of internal control deﬁciencies. Contemporary Accounting Research 25: 219–242. Hribar, P., J. Kim, R. Wilson, and H. Yang. 2012. Counterparty Responses to Managerial Overconﬁdence. Working paper, The University of Iowa and University of Pennsylvania. Huang, T. C., H. W. Huang, and C. C. Lee. 2014. Corporate executive’s gender and audit fees. Managerial Auditing Journal 29 (6): 527– 547. Ittonen, K., J. Miettinen, and S. Va¨ha¨maa. 2010. Does female representation on audit committees affect audit fees? Quarterly Journal of Finance and Accounting 49 (3–4): 113–139. Jiang, J., K. R. Petroni, and I. Y. Wang. 2010. CFOs and CEOs: Who has the most inﬂuence on earnings management? Journal of Financial Economics 96: 513–526. Johnson, E. N., J. R. Kuhn, Jr., B. A. Apostolou, and J. M. Hassell. 2012. Auditor perceptions of client narcissism as a fraud attitude risk factor. Auditing: A Journal of Practice & Theory 32 (1): 203–219. Johnson, K. 2015. Finance chiefs getting shot at CEO role. Wall Street Journal. Available at: http://blogs.wsj.com/cfo/2015/08/07/ ﬁnance-chiefs-getting-shot-at-ceo-role/ Judd, J. S., K. J. Olsen, and J. Stekelberg. 2015. CEO Narcissism, Accounting Quality, and External Audit Fees. Working paper, University of Illinois at Chicago, Utah State University, and The University of Arizona. Kannan, Y. H., T. R. Skantz, and J. L. Higgs. 2014. The impact of CEO and CFO equity incentives on audit scope and perceived risks as revealed through audit fees. Auditing: A Journal of Practice & Theory 33 (2): 111–139. Kim, Y., H. Li, and S. Li. 2015. CEO equity incentives and audit fees. Contemporary Accounting Research 32: 608–638. Kimberly, J. R., and M. J. Evansiko. 1981. Organizational innovation: The inﬂuence of individual, organizational, and contextual factors on hospital adoption of technical and administrative innovations. Academy of Management Journal 24: 689–713. Kinney, W., and L. McDaniel. 1993. Audit delay for ﬁrms correcting quarterly earnings. Auditing: A Journal of Practice & Theory (Fall): 135–142.
CEO Financial Background and Audit Pricing 337
Accounting Horizons Volume 30, Number 3, 2016
Knyazeva, A., D. Knyazeva, and R. Masulis. 2013. The supply of corporate directors and board independence. Review of Financial Studies 26 (6): 1561–1605. Kothari, S. P., A. J. Leone, and C. Wasley. 2005. Performance matched discretionary accrual measures. Journal of Accounting and Economics 39 (1): 163–197. Koyuncu, B., S. Firﬁray, B. Claes, and M. Hamori. 2010. CEOs with a functional background in operations: Reviewing their performance and prevalence in the top post. Human Resource Management 49: 869–882. Krishnan, G., and G. Visvanathan. 2008. Does the SOX deﬁnition of an accounting expert matter? The association between audit committee directors’ accounting expertise and accounting conservatism. Contemporary Accounting Research 25 (3): 827–857. Krishnan, G., and C. Wang. 2015. The relation between managerial ability and audit fees and going concern opinions. Auditing: A Journal of Practice & Theory 34 (3): 139–160. Lee, H. Y., and V. Mande. 2005. The relationship of audit committee characteristics with endogenously determined audit and non-audit fees. Quarterly Journal of Business and Economics 44 (3/4): 93–112. Maddala, G. 1983. Limited-Dependent and Qualitative Variables in Econometrics. Cambridge, MA: Cambridge University Press. Mason, E. S., and P. E. Mudrack. 1996. Gender and ethical orientation: A test of gender and occupational socialization theories. Journal of Business Ethics 15: 599–605. Matsunaga, S. R., S. Wang, and P. E. Yeung. 2013. Does Appointing a Former CFO as CEO Inﬂuence a Firm’s Accounting Policies? Working paper, University of Oregon. McInnis, J., and D. W. Collins. 2011. The effect of cash ﬂow forecasts on accrual quality and benchmark beating. Journal of Accounting and Economics 51 (3): 219–239. Mitra, S., D. Deis, and M. Hossain. 2009. The association between audit fees and reported earnings quality in pre- and post-SarbanesOxley regimes. Review of Accounting and Finance 8 (3): 232–252. Mobbs, S. 2014. Chief Financial Ofﬁcer Board Membership: Causes and Consequences. Working paper, The University of Alabama. Munsif, A., K. Raghunandan, D. V. Rama, and M. Singhvi. 2011. Audit fees after remediation of internal control weaknesses. Accounting Horizons 25 (1): 87–105. Ng, P. P. H., and B. Y. K. Tai. 1994. An empirical investigation of the determinants of audit delay in Hong Kong. British Accounting Review 26: 43–59. Peterson, D., A. Rhoads, and B. C. Vaught. 2001. Ethical beliefs of business professionals: A study of gender, age and external factors. Journal of Business Ethics 31: 225–233. Raghunandan, K., and D. V. Rama. 2006. SOX section 404 material weakness disclosures and audit fees. Auditing: A Journal of Practice & Theory 25: 99–104. Shibano, T. 1990. Assessing audit risk from errors and irregularities. Journal of Accounting Research 28 (3): 110–140. Simon, D., and J. Francis. 1988. The effects of auditor change on audit fees: Tests of price cutting and price recovery. The Accounting Review 63: 255–269. Simunic, D. A. 1980. The pricing of audit services: Theory and evidence. Journal of Accounting Research 18: 161–190. Sundaram, R. K., and D. Yermack. 2007. Pay me later: Inside debt and its role in managerial compensation. The Journal of Finance 62: 1551–1587. Vafeas, N., and J. F. Waegelein. 2007. The association between audit committees, compensation incentives, and corporate audit fees. Review of Quantitative Finance and Accounting 28: 241–255. Wysocki, P. 2010. Corporate compensation policies and audit fees. Journal of Accounting and Economics 49: 155–160. Zmijewski, M. 1984. Methodological issues related to the estimation of ﬁnancial distress. Journal of Accounting Research 22: 59–82.
APPENDIX A Variable Descriptions
Main Variables AFee¼the natural logarithm of audit fees; and FinExp¼1 for the ﬁrm-year observations when the CEO is a ﬁnancial expert, else 0.
Control Variables Size¼logarithm of total assets; CurrRatio¼ratio of current assets to total assets; ICweak¼1 if there is a material weakness in internal control, else 0; Distress¼Zmijewski’s (1984) probability of bankruptcy; GC¼1 if the audit opinion is modiﬁed for going concern, else 0; InvRec¼ratio of accounts receivables and inventories to total assets;
338 Kalelkar and Khan
Accounting Horizons Volume 30, Number 3, 2016
Segments¼square root of the number of business segments; Foreign¼1 if the ﬁrm has foreign operations, else 0; Exord¼1 if the ﬁrm reports extraordinary items or discontinued operations, else 0; Big4¼1 if the auditor represents one of the Big 4 auditing ﬁrms, else 0; and AudChg¼1 if there was a change in auditor, else 0. Other Variables Departure¼1 if CEO turnover is caused by death, retirement, or move elsewhere in a same position, else 0; CFO Change¼1 if there is a CFO change, else 0; Dac¼absolute value of residuals from Kothari et al. (2005) performance matched modiﬁed Jones model; Outsider¼1 if the ﬁnancial expert CEO is recruited from outside the ﬁrm, else 0; and Density¼log of number of ﬁnancial ﬁrms within 100 miles of the focal ﬁrm.
CEO Financial Background and Audit Pricing 339
Accounting Horizons Volume 30, Number 3, 2016
Copyright of Accounting Horizons is the property of American Accounting Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder’s express written permission. However, users may print, download, or email articles for individual use.