Do firms that have a common signing auditor exhibit higher earnings comparability

THE ACCOUNTING REVIEW Vol 95, No May 2020 pp 115–143 American Accounting Association DOI: 10.2308/accr-52522 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? Jeff Zeyun Chen Texas Christian University Mei-Hui Chen National Defense University Chen-Lung Chin National Chengchi University Gerald J Lobo University of Houston ABSTRACT: We hypothesize that if individual auditors possess unique audit styles that they consistently apply to different audit engagements, then client firms with a common signing auditor will exhibit higher earnings comparability Using a large sample of Chinese firms, we find that client firms report more comparable earnings when they are audited by the same individual auditor than when they are audited by (1) different audit firms, (2) the same audit firm, but different audit offices, and (3) the same audit office, but different individual auditors The individual auditor style effect is stronger for larger audit firms, senior signing auditors, and signing auditors with more stable teamwork experience We also document that having a common signing auditor is associated with lower analyst earnings forecast error and dispersion for client firms This study contributes to the literature by showing that individual auditors have a significant impact on client firms’ earnings comparability Keywords: individual auditor style; earnings comparability; auditor demographic characteristics; teamwork experience I INTRODUCTION R egulators, practitioners, investors, and academics often stress the importance of earnings comparability (De Franco, Kothari, and Verdi 2011) As indicated by the Financial Accounting Standards Board (FASB 2010), comparability is a desirable property of financial statements However, despite its importance, we know relatively little about the extent to which economic agents and institutional incentives influence comparability beyond the role of accounting standards In this study, we examine whether individual auditors exhibit unique audit styles that they consistently apply to different engagements, and the extent to which individual auditor styles affect client firms’ earnings comparability Our study is motivated by recent findings in the audit literature that highlight the importance of understanding auditor style and audit outcome at the individual level (Wallman 1996; Gul, Wu, and Yang 2013; Knechel, Vanstraelen, and Zerni 2015; Li, Qi, Tian, and Zhang 2017) An individual auditor often has considerable autonomy and flexibility to interpret and implement the in-house working rules established by the audit firm or audit office Consequently, given their personal attributes, individual auditors develop their own unique set of judgments and working rules that standardize the application of accounting and We thank Rodrigo S Verdi (editor), two anonymous referees, and workshop participants at Bocconi University, ESSEC Business School, Free University of Bozen-Bolzano, Fudan University, Peking University (Shenzhen), Renmin University of China, and University of Padova for helpful comments and suggestions Chen-Lung Chin is grateful for financial support provided by E Sun Bank in Taiwan Editor’s note: Accepted by Rodrigo S Verdi, under the Senior Editorship of Mark L DeFond Submitted: December 2016 Accepted: June 2019 Published Online: August 2019 115 116 Chen, Chen, Chin, and Lobo auditing standards among their clients The individual-level working rules and standardization process give rise to individual auditor styles Francis, Pinnuck, and Watanabe (2014; hereafter, FPW) find that clients of the same Big audit firm, subject to the common audit firm style, exhibit higher earnings comparability than clients of different Big audit firms.1 FPW use the term ‘‘audit (firm) style’’ to characterize the unique set of internal working rules of each Big audit firm for implementation of auditing standards and enforcement of generally accepted accounting principles (GAAP) within their clienteles Kawada (2014) further argues that it is the local audit office that implements the in-house working rules, and a localized audit style may exist because staff training is largely decentralized in large audit firms He documents a significant effect of audit office style on client firms’ earnings comparability that is not explained by the audit firm effect Given FPW’s and Kawada’s (2014) findings, it remains an empirical question whether the implementation of comparability is still influenced by individual auditors Although personal characteristics of individual auditors are associated with commonly studied earnings attributes (Gul et al 2013), these are firm-specific measures that are calculated independently of the attributes of other firms Comparability is related to similarities in earnings attributes across different firms (De Franco et al 2011) Each audit firm or audit office has its own quality control system to ensure consistent application of auditing standards and enforcement of GAAP across different engagements Therefore, it is possible that the individual auditor effect on earnings comparability is muted in the presence of audit firm style and/or audit office style An important objective of our analysis is to shed light on the question of whether individual auditor style has a unique role in shaping earnings comparability and, if so, whether it is more or less important than audit firm style and audit office style We examine the relation between individual auditor style and client firms’ earnings comparability in China, where data on the identities of signing auditors are publicly available Chinese regulators require that each audit report be signed by two individual auditors, one junior (engagement) auditor and one senior (review) auditor (Ministry of Finance of the People’s Republic of China [MOF] 1995a, 1995b) The role of the signing auditors in China is similar to that of engagement partners in other markets (Gul et al 2013) Signing auditors are responsible for decision-making during the audit process and hence can influence the audit outcomes and client financial statements Our analyses are based on pairs of nonfinancial client firm-year observations in the same industry-year from the China Stock Market and Accounting Research (CSMAR) database between 2003 and 2013 Following FPW, we measure comparability by the closeness of total and abnormal accruals between pairs of client firms and the degree to which the earnings of a pair of client firms covary over time To ensure that these measures are not contaminated by client firm characteristics related to the choice of audit firms, we further restrict the pairings to be between client firms that hire similar-sized auditors Our main hypothesis is that firm-pairs that share the same signing auditor have a unique individual auditor style effect on earnings comparability, in addition to the effects of audit firm style and audit office style To isolate individual auditor style from audit firm style and audit office style, we classify the sample into four groups: (1) firm-pairs with different audit firms, (2) firm-pairs with the same audit firm, but different audit offices, (3) firm-pairs with the same audit office, but different signing auditors, and (4) firm-pairs with the same signing auditor We find a positive effect on client firms’ earnings comparability when firm-pairs have a common audit office or a common signing auditor However, after controlling for audit office style and individual auditor style, we not find such an effect when firm-pairs share a common audit firm only Importantly, we show that client firms’ earnings comparability is more sensitive to individual auditor style than it is to audit office style, suggesting that individual auditors play a more pivotal role in ensuring consistent enforcement of GAAP during the audit process This result is not only statistically, but also economically, significant Having the same signing auditor leads to a decrease of 11.9 percent from the mean value of differences in total accruals for firm-pairs in our sample The corresponding decrease for firmpairs audited by the same audit office (but different signing auditors) is just 5.5 percent We perform several cross-sectional analyses to examine whether the individual auditor style effect on client firms’ earnings comparability varies with audit firm and signing auditor characteristics First, individual auditors at large audit firms are more competent and better trained than their counterparts at small audit firms, making them more confident of their own styles As a result, they may rely more on their professional judgment than on standardized audit procedures We find that the effect of having the same signing auditor on earnings comparability is stronger for large audit firm clienteles than for small audit firm clienteles.2 Second, we focus on signing auditors’ professional experience Professional experience is a key factor in forming individual auditor style because it helps auditors develop mental models for interpreting and integrating evidence during the FPW define comparability as the closeness of two firms’ reported earnings due to the consistency with which rules are applied across firms We adopt the same definition and follow FPW to operationalize the concept of comparability In our study, we define large audit firms as the Top Ten audit firms in China We obtain audit firm ranking data from the website of the Chinese Institute of Certified Public Accountants (CICPA) The most important ranking factor is the audit firm’s total revenues Over our sample period (2003–2013), the Big international firms invariably occupy the top four ranks measured by total revenues in China The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 117 audit process in a systematic way (Bonner and Lewis 1990) We find a stronger relation between having a common senior signing auditor and client firms’ earnings comparability than having a common junior signing auditor This result suggests that more experienced auditors are more confident of their audit styles and apply GAAP on a more consistent and correct basis Third, we consider the senior-junior paired signing auditors as a team and investigate the role of teamwork experience in shaping individual auditor style We measure teamwork experience by the number of years the two signing auditors have worked together on any engagement in the past The longer they have worked together, the more stable their teamwork experience and the more likely they are to have formed a persistent audit style We find that client firm-pairs subject to the same individual auditor style exhibit higher earnings comparability when their signing auditors have more stable teamwork experience In additional analyses, we use several approaches to draw causal inferences First, we limit our sample to firm-pairs subject to mandatory auditor rotation requirements We find a significant drop in earnings comparability for firm-pairs that switch from having the same signing auditor to having different signing auditors, and vice versa for those that switch from having different signing auditors to having the same signing auditor Second, we examine the change in the individual auditor effect on earnings comparability under the mandatory international financial reporting standards (IFRS) reporting regime IFRS leaves more room for individual auditors to exercise professional judgment, so their personal work styles are likely to be more important in shaping clients’ financial reporting We find that although the individual auditor style effect manifests in both periods, it is markedly stronger in the post-IFRS period Last, our results are robust to using De Franco et al.’s (2011) measure of accounting system comparability, and continue to hold when we use the propensity score matching technique to address concerns about misspecification of the regression model (Shipman, Swanquist, and Whited 2017) These robustness checks enhance confidence that the observed individual auditor effect does not merely reflect a spurious correlation between client firms’ economic similarity and their common auditor choices An alternative explanation for our results is that individual auditor style results in earnings uniformity between client firmpairs Unlike comparability, uniformity makes different things look alike and, thus, reduces earnings informativeness De Franco et al (2011) argue that earnings comparability should reduce information acquisition costs for financial analysts, and find that forecast accuracy increases and forecast dispersion decreases in earnings comparability Similarly, we find that having a common signing auditor leads to higher analyst forecast accuracy and lower forecast dispersion for client firm-pairs This result is more consistent with individual auditor style enhancing earnings comparability than with forcing earnings uniformity on client firms Our study makes several contributions First, FPW find a positive effect of having the same Big audit firm on client firms’ earnings comparability Using more granular data, we show that the effect of audit style on earnings comparability exists at the audit office level and, more prominently, at the signing auditor level.3 The importance of individual auditors in shaping audit outcomes has recently attracted attention from regulators Effective in 2017, the Public Company Accounting Oversight Board (PCAOB) adopted a rule that, for the first time, requires all U.S public firms to disclose the name of the engagement partner This rule is intended to improve audit quality and benefit investors.4 We offer the first evidence that individual auditors have a significant influence on client firms’ earnings comparability, which, in turn, affects investors’ information acquisition costs Second, prior research has primarily focused on accounting standards as a key determinant of earnings comparability (Yip and Young 2012; Barth, Landsman, Lang, and Williams 2012; Cascino and Gassen 2015) We extend the comparability literature by showing that individual auditors are important to enforcing consistent implementation of accounting standards across different audit engagements We find that our results are stronger under the IFRS regime, highlighting a more significant individual auditor effect on reporting outcome when client firms move toward a less-specific reporting environment Third, we show that when a senior auditor is paired with a junior auditor on an engagement team, both can contribute to client firms’ earnings comparability, but to different degrees In addition, the individual auditor effect increases with stable teamwork experience, suggesting that social learning is a key factor in forming personal work style (Su and Wu 2016) Overall, our results indicate that individual auditor style can be developed and strengthened over time and in stable teams, thus highlighting the importance of training and staffing designs at audit firms to improve consistency in audit outcomes The Chinese and the U.S audit markets differ along two important dimensions: market competition and litigation risk To the extent that these countrylevel institutional factors affect how individual auditors exert influence on audit outcomes, our findings may not be readily generalizable to the U.S Therefore, we suggest that caution be exercised when directly comparing our results to FPW’s For example, former PCAOB board member Steven B Harris (2016) argues that audit quality will improve from the public identification of the engagement partner as it emphasizes the individual auditor’s sense of accountability, and investors will be able to take advantage of this information by evaluating and comparing the performance of individual engagement partners over time The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 118 The rest of this study is organized as follows We develop the hypotheses in Section II and present the research design in Section III We describe the sample and data in Section IV, discuss the main results in Section V, and the results of additional analyses in Section VI We provide our conclusions in Section VII II HYPOTHESIS DEVELOPMENT The Individual Auditor Style Effect on Earnings Comparability Comparability enables financial statement users to identify and understand similarities in and differences among financial reporting items Many studies have demonstrated the value of comparability to financial statement users (De Franco et al 2011; Kim, Kraft, and Ryan 2013; Kim, Li, Lu, and Yu 2016; Chen, Collins, Kravet, and Mergenthaler 2018) However, we know relatively little about factors that affect comparability at the firm level Similar to other qualitative characteristics of financial reporting, comparability is sensitive to the incentives of the managers and auditors responsible for financial statement preparation FPW argue that each Big audit firm has a unique audit style, in the sense that each has its unique audit testing approach for implementing generally accepted auditing standards (GAAS) and in-house working rules for interpreting and implementing GAAP Kawada (2014) contends that it is the local office that implements the rules and applies them to client engagements Auditing is inherently a professional judgment process of the engagement team, and the audit outcome ultimately depends on the competency and incentives of individual auditors (Nelson and Tan 2005; Nelson 2009) In recent years, with partner-level data made available to the public in many countries, an emerging line of research has examined the characteristics of individual auditors that influence audit quality (Carey and Simnett 2006; Chen, C Lin, and Y Lin 2008; Chi, Huang, Liao, and Xie 2009; Lennox, Wu, and Zhang 2014; Zerni 2012; Gul et al 2013) This literature finds strong evidence that individual auditors can influence audit outcomes An interesting question that arises is, to what extent individual auditors possess unique audit styles that persist over time and extend to different audit engagements? Individual auditors exhibit different audit styles due to idiosyncratic personal attributes (Kennedy and Peecher 1997; Messier, Owhoso, and Rakovski 2008) In an experimental study, Pincus (1990) finds that auditors who are more fieldindependent and ambiguity-intolerant are more likely to detect inventory manipulation by the management Amer, Hackenbrack, and Nelson (1994) provide experimental evidence that interpretation of a variety of the probability phrases used in audit practice exhibits significant between-auditor differences, and auditors’ awareness of the extent of such interpretation variance is generally low These results suggest that individual auditors interpret and implement the working rules in systematically different ways, leading to persistent variation in the application of professional standards and audit engagement tools at the individual auditor level Using archival data, Knechel et al (2015) confirm that aggressive and conservative audit reporting persists for individual auditors over time Li et al (2017) find that individual auditors who have performed failed audits also deliver lower-quality audits on other audit engagements Such a contagion effect is only confined to the failed auditor and does not spread to nonfailed auditors in the same local office as the failed auditor The latter result reinforces the notion that each auditor has a unique audit style that is likely to be independent of office-level factors, such as working rules, training programs, and office culture, that influence audit outcome Although firm-level and office-level risk management and quality control mechanisms are important for choosing the level of acceptable risk and the materiality threshold, individual auditors exercise a substantial amount of discretion in the field when implementing internal working rules and testing procedures This discretion, if exercised in a systematic way, will imprint an individual auditor’s mark on her clients’ financial reporting To the extent that individual auditor style is not subsumed by audit firm style and audit office style, we expect greater comparability in financial statements of client firm-pairs that share the same signing auditor than of client firm-pairs that have different signing auditors Accordingly, we state our first hypothesis (in the alternative form) as follows: H1: Firm-pairs that share the same signing auditor have a unique individual auditor style effect on earnings comparability, in addition to the effects of audit firm style and audit office style Cross-Sectional Predictions In this section, we develop three hypotheses based on cross-sectional differences in the extent to which audit firm and signing auditor characteristics influence individual auditor style FPW argue that the effect of audit firm style on client firms’ earnings comparability is stronger for Big 4s than for non-Big 4s Large audit firms have more resources and greater capacity to develop unique in-house rules for interpreting and implementing GAAP than small audit firms They invest more in risk control, technical support, and personnel training because they have a larger and more dispersed staff If the in-house rules are The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 119 more faithfully and consistently implemented at large audit firms, then clients of Big 4s should report more comparable financial statements than clients of non-Big 4s The above reasoning suggests that individual auditor style may be muted at large audit firms In addition, more rigorous and detail-oriented quality control procedures at large audit firms constrain individual auditors from exercising professional judgment and making decisions in the field, thus leaving less room for them to exert influence on the audit outcome There are also reasons that individual auditors may play a more prominent role in shaping clients’ financial reporting at large audit firms Individual auditors at large audit firms are generally more competent, more diligent, harder working, and better trained to interpret and implement GAAP.5 They are more confident to uphold their personal work styles and to exercise professional judgment in a consistent manner for similar economic events across different engagements McKnight and Wright (2011) find that higher-performing auditors are inclined to extend standard audit procedures, leading to a higher level of professional skepticism Audit firms provide general rules and guidance, but highly competent auditors have knowledge and skills that they can quickly adapt to new and complex situations As FPW point out, high-quality standards and in-house rules by themselves not necessarily result in comparability—they must be faithfully implemented Highly competent auditors are more likely to apply accounting standards correctly and deliver higher audit quality Given the significant discretion in financial reporting afforded by GAAP, comparability will be higher among firms whose auditors are more competent because the accounting standards are applied on a more consistent and correct basis Given the competing views of the role of individual auditor style in financial statement comparability for client firms of large versus small audit firms, we not offer a directional hypothesis Instead, we leave it to our empirical testing to reject the following null hypothesis in favor of either of the two alternatives: H2 (null): The effect of individual auditor style on earnings comparability for client firm-pairs does not differ between large and small audit firms An important factor that determines whether individual auditors correctly implement the accounting standards in a consistent manner is their professional experience, which helps them develop effective mental models for interpreting and integrating evidence during the audit process (Bonner and Lewis 1990) Kaplan, O’Donnell, and Arel (2008) find that as auditors gain experience, they are less likely to be persuaded by management’s favorable assessment Shelton (1999) finds that irrelevant information has a diluting effect on the judgments of inexperienced auditors in making going-concern assessments, but experienced auditors are not affected by the presence of irrelevant information Abdolmohammadi and Wright (1987) show that the experience effect on audit judgments is more prominent when the audit task becomes more complex These studies suggest that professional experience fosters the development and use of personal work style Chinese law requires that two individual auditors sign each audit report The senior (review) signing auditor conducts quality control and final review, and the junior (engagement) signing auditor performs the detailed fieldwork Our next hypothesis examines whether the individual style effect of senior signing auditors differs from that of junior signing auditors If the more experienced auditor exhibits a more confident and consistent audit style, then we should observe a stronger relation between having a common senior signing auditor and clients’ earnings comparability than having a common junior signing auditor.6 Based on the above discussion, we present our hypothesis (in the alternative form) as follows: H3: The effect of individual auditor style on earnings comparability for client firm-pairs is stronger for senior signing auditors than for junior signing auditors Next, we consider the paired signing auditors as a team Specifically, we examine whether the two signing auditors’ teamwork experience enhances the individual style effect on client firms’ earnings comparability Teamwork experience is an important channel through which audit behavior is transmitted between individual auditors Su and Wu (2016) find that individual auditors who have worked on the same team with a sanctioned auditor are more likely to issue lenient audit opinions and deliver lower audit quality than those who have not overlapped with the sanctioned auditor This contagion effect exists only among team members and does not spread to non-overlapping colleagues in the same audit firm Su and Wu’s (2016) results suggest that social learning via teamwork experience affects individual auditor behavior and audit outcome, in addition to audit firm style Large audit firms have national training programs and resources to invest in information technology that facilitates firm-wide knowledge-sharing practices, resulting in better trained auditors (Francis and Yu 2009; Eshleman and Guo 2014) Survey evidence suggests that auditors at large audit firms have a higher educational level and more work experience, and large audit firms spend more money on continuing professional development (Cheng, Liu, and Chien 2009) Prior literature also finds evidence that auditors with greater competency exert greater audit effort (Bae, Choi, and Rho 2016) We note that all signing auditors are CPAs in China, who must pass the CPA exam and have at least two years of qualifying work experience However, Lennox et al (2014) find that the positive effect of mandatory partner rotation on audit quality is limited to engagement auditors This result suggests that it might be the junior signing auditor who plays a more important role in shaping the audit outcome The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 120 In a stable team, signing auditors develop mutual trust, which is key to social learning Social learning, in turn, facilitates collaborative work and sharing of knowledge, ideas, and experiences between signing auditors As teamwork experience accumulates, individual auditor style emerges Signing auditors who have worked together longer are more likely to develop routines for solving problems that can be applied to different engagements in a consistent manner As teamwork experience increases, each signing auditor gets better at executing existing routines and developing new ones Consequently, a stable signing auditor team ensures development and use of individual auditor style By contrast, individual auditors may diversify away their work styles if the signing auditor team composition constantly changes, and each team member must frequently adapt to new audit testing approaches for implementing GAAP and in-house working rules on different engagements The above discussion leads to the following hypothesis (in the alternative form): H4: The effect of individual auditor style on earnings comparability for client firm-pairs is stronger if the common signing auditor has more stable teamwork experience than if the common signing auditor has less stable teamwork experience III RESEARCH DESIGN Measuring Earnings Comparability Following De Franco et al (2011) and FPW, we use firm-pairs as the unit of observation To mitigate industry- and timespecific effects on earnings characteristics, we pair each sample firm with a peer firm in the same industry and year We also restrict the pairing to firms with similar-sized auditors (either Top Ten or non-Top Ten audit firms) to remove some inherent differences between large and small audit firm clienteles for the paired firms Similar to FPW, we measure earnings comparability based on cross-sectional similarities in accruals for firm-pairs and the correlation of earnings for firm-pairs over time Closeness of Accruals To the extent that a firm-pair audited by the same signing auditor is subject to the same type of accrual adjustments due to her unique audit style, the accruals structure of the two firms should be similar We measure the closeness of accruals for firmpairs as follows:   Tacc Diff ijt ¼ Taccit À Taccjt  ð1Þ where Tacc_Diffijt is the absolute value of the difference between signed total accruals for each firm i-firm j pair in the same industry and with auditors of similar size in year t We measure total accruals, Tacc, as income before extraordinary items less cash flows from operations adjusted for cash flows from extraordinary items, divided by total assets at the beginning of the year Because earnings comparability decreases with increasing values of Tacc_Diffijt, we multiply Tacc_Diffijt by À1 for ease of interpretation and comparison Closeness of Abnormal Accruals To separate the effect of managerial discretion from firm fundamentals on accruals, we also measure earnings comparability by the similarity of abnormal accruals for firm-pairs We calculate abnormal accruals (AbnAcc) using the performance-adjusted Jones model (Jones 1991; Kothari, Leone, and Wasley 2005) and measure the closeness of abnormal accruals for firm-pairs, AbnAcc_Diffijt, as the absolute value of the difference between signed abnormal accruals for each firm i-firm j pair.7 Because earnings comparability decreases with increasing values of AbnAcc_Diffijt, we multiply AbnAcc_Diffijt by À1 Earnings Covariation The third measure of earnings comparability captures the extent to which earnings for firm-pairs covary over time Specifically, we measure the level of covariation as the R2 from the following regression: Earningsiq ẳ a0ij ỵ a1ij Earningsjq ỵ eijq 2ị To estimate performance-adjusted Jones model, we require at least ten observations in each industry-year The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 121 where Earnings is income before extraordinary items for firm i or firm j in quarter q divided by average total assets of each firm We estimate Model (2) over eight consecutive quarters for all unique pairs of firms in the same industry with auditors of similar size The R2 from Model (2), labeled EComp_Cov, increases with earnings covariation between firm-pairs.8 Models for Testing the Hypotheses We estimate the following ordinary least squares (OLS) regression model to test H1:   À1 Tacc Diff ijt or AbnAcc Diff ijt ẳ b0 ỵ b1 SameFirm Diff Off ice Diff Partnerijt ỵ b2 SameOff ice Diff Partnerijt ỵ b3 SamePartnerijt ỵ Controls ỵ eijt ð3Þ where SameFirm_DiffOffice_DiffPartner is an indicator variable that equals if both firm i and firm j in the pair are audited by the same audit firm, but different audit offices and different individual auditors, and otherwise SameOffice_DiffPartner is an indicator variable that equals if the two firms comprising a pair are audited by the same audit firm and same audit office, but different individual auditors, and otherwise SamePartner is an indicator variable that equals if both firms in the pairwise combination are audited by the same audit firm and same audit office and same individual auditor, and otherwise In a typical audit report in China, the name of the senior (review) signing auditor is disclosed first, followed by the name of the junior (engagement) signing auditor.9 In our tests of H1 and H2, we not differentiate between the senior and junior signing auditors in an auditor-pair signing the audit report If there is at least one common signing auditor for the client firm-pair, we code SamePartner as In Model (3), the benchmark group contains firm-pairs audited by different audit firms (and audit offices, and individual auditors) b1, b2, and b3 reflect the unique effects of audit firm style, audit office style, and individual auditor style on client firms’ earnings comparability, respectively H1 predicts that b3 is positive An advantage of our model specification is that it facilitates comparisons among the three coefficients, which can reveal the contribution of each audit style to client firms’ earnings comparability The control variables are similar to those used in FPW Specifically, we control for firm size, leverage, market-to-book, cash flow from operations, losses, standard deviation of sales, standard deviation of cash flows, and sales growth, all of which prior research shows are related to accruals Because the dependent variable is the difference in total (abnormal) accruals, we include both the difference in and the level of these firm characteristics for firm-pairs as control variables.10 We also include the level of total (abnormal) accruals as a control variable because Becker, DeFond, Jiambalvo, and Subramanyam (1998) show that it is influenced by auditors We add two controls for individual auditor characteristics: average audit tenure for the two signing auditors, and whether either signing auditor is an industry specialist (Carey and Simnett 2006; Lennox et al 2014; Balsam, Krishnan, and Yang 2003; Francis, Reichelt, and Wang 2005; Chi and Chin 2011) We control for industry fixed effects and year fixed effects to reduce potential bias due to correlated omitted variables To reduce the effect of serial dependence in the error term from multiple observations of the same firm-pair over the sample period, we use robust standard errors clustered at the firm-pair (i, j) level in our tests.11 We provide detailed variable descriptions in Appendix A When we measure earnings comparability by EComp_Cov, we add two more controls for economic comparability to Model (3) following FPW: contemporaneous cash flow covariation (CFOComp_Cov) and monthly stock return covariation (RetComp_Cov) for firm-pairs We so because earnings covariation is affected not only by audit style, but also by common economic fundamentals and shocks between the firms in the pair To test H2, we estimate Model (3) separately for firm-pairs that hire the Top Ten audit firms and firm-pairs that hire the non-Top Ten audit firms If the individual auditor style effect is stronger (weaker) in larger audit firms, then b3 should be larger (smaller) for the client firm-pairs of Top Ten audit firms H3 predicts that the individual auditor style effect on client firms’ earnings comparability is stronger for the senior signing auditor To test H3, we estimate Model (3) after replacing SamePartner with three indicator variables: SamePartner_Senior, SamePartner_Junior, and SamePartner_Others.12 10 11 12 FPW estimate Model (2) over 16 consecutive quarters for all unique pairs of firms in the same industry We use only eight consecutive quarters because of our relatively shorter sample period Therefore, we caution against direct comparisons between our EComp_Cov measure and FPW’s Lennox et al (2014) confirm that audit partners whose signatures are disclosed first on audit reports are more experienced compared with audit partners whose signatures are disclosed second We control for levels using the minimum value of the paired control variable for firm i and firm j in year t Our inferences remain the same when we cluster the standard errors at the firm i level For this test, we drop firm-pairs that have the same review auditor and the same engagement auditor The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 122 SamePartner_Senior equals if both client firms in the pair have the same senior signing auditor (but different junior signing auditors), and otherwise SamePartner_Junior equals if the firm-pair has the same junior signing auditor (but different senior signing auditors), and otherwise SamePartner_Others equals for all other firm-pairs with SamePartner ¼ (e.g., the same signing auditor is a review auditor for firm i and an engagement auditor for firm j) A larger coefficient on SamePartner_ Senior than on SamePartner_Junior will be consistent with H3 To measure teamwork experience, we identify the two signing auditors for each client firm in year t and count the number of years in the past (from 2003 to year t) that these two auditors have worked together on any audit engagements (i.e., they both signed the same audit report).13 Because the unit of observation in our analyses is a firm-pair (i, j), we use the minimum value of auditors’ teamwork experience between firm i and firm j as our measure for the firm-pair If this measure is larger (smaller) than the sample median, then we classify the firm-pair in the subsample with auditors having more (less) stable teamwork experience.14 To test H4, we first partition the subsample with SamePartner ¼ into two groups: firm-pairs with two common signing auditors (SameTeam ¼ 1) and firm-pairs with only one common signing auditor (SamePartner_DiffTeam ¼ 1) We then modify Model (3) by replacing SamePartner with SameTeam and SamePartner_DiffTeam and separately estimate the model for the more and less stable signing auditor teamwork experience subsamples H4 predicts that the coefficient on SameTeam will be larger if the common signing auditor team for the paired client firms has more stable teamwork experience IV SAMPLE SELECTION AND DESCRIPTIVE STATISTICS Sample Selection We obtain financial statement data and stock returns data from the CSMAR database We collect identities of signing auditors from annual reports, and other personal information, such as office affiliation and demographic characteristics, from the website of CICPA Table summarizes the sample selection procedures The initial sample comprises 23,707 firm-year observations that issue A-shares traded on the stock exchanges in Shanghai and Shenzhen from 2003 to 2013.15 We drop 346 observations for client firms in the financial industry, and 2,236 observations with missing data for signing auditors (e.g., name, office location, and tenure) In addition, we lose 5,410 observations because of insufficient data to estimate abnormal accruals and calculate various control variables used in Model (3), leaving us with 15,715 firm-year observations for the accrualsdifference sample All firms in this sample are exhaustively paired in the same industry-year and we retain only unique firmpairs with similar-sized auditors.16 Thus, the final sample for testing H1 based on Tacc_Diff and AbnAcc_Diff has 267,312 firmpairs, and includes 107,179 client firm-pairs of the Top Ten audit firms and 160,133 client firm-pairs of the non-Top Ten audit firms The earnings covariation sample is a subsample of the accruals-difference sample It includes only those firms with data on earnings and cash flow from operations for eight consecutive quarters and without any auditor changes over the eight quarters The first time a firm-pair appears in the sample is when its eight consecutive quarters of data first become available Similar to FPW, we calculate EComp_Cov using firm-pairs with non-overlapping two-year periods to mitigate concerns about correlated error terms in the regression analysis The sample based on EComp_Cov contains 98,945 firm-pairs (45,994 client firm-pairs of the Top Ten audit firms and 52,951 client firm-pairs of the non-Top Ten audit firms) To reduce the influence of extreme values, we winsorize all continuous variables at the percent and 99 percent levels Descriptive Statistics Panel A of Table presents descriptive statistics for all variables used in the main tests The mean and median values of À1 Tacc_Diff (À1 AbnAcc_Diff ) are À0.109 (À0.089) and À0.073 (À0.064), respectively The mean and median values of EComp_Cov are 0.261 and 0.169, respectively, which suggests that Chinese firms have a large degree of earnings comovement within the same industry-year.17 Given the highly competitive nature of the audit market in China, it is not 13 14 15 16 17 Alternatively, we count the number of different engagements that the two signing auditors have previously worked on together Our results (untabulated) remain qualitatively the same when we use this measure as a proxy for teamwork experience We find that the distribution of the minimum value of teamwork experience for client firm-pairs is skewed to the right, in that a significant portion of client firm-pairs have the minimum value of teamwork experience of only one year (70.48 percent in the accruals-difference sample and 71.47 percent in the earnings co-movement sample) This positive skewness results in a significantly larger sample size for the less stable teamwork experience subsample (188,407 firm-pairs in the accruals-difference sample and 70,717 firm-pairs in the earnings co-movement sample) than the more stable teamwork experience subsample (78,905 firm-pairs in the accruals-difference sample and 28,228 firm-pairs in the earnings co-movement sample) A-shares can only be owned and traded by Chinese citizens For example, if there are four firms, i, j, k, and l, then we obtain six unique firm-pairs: i-j, i-k, i-l, j-k, j-l, and k-l FPW report that the mean and median of EComp_Cov are 0.114 and 0.052, respectively, for their sample The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 123 TABLE Sample Selection Total client firm-year observations available on CSMAR from 2003 to 2013 Less: Observations in the financial industry Observations with missing identity data for signing auditors (name and office affiliation) Observations without sufficient data to calculate abnormal accruals and control variables Total client firm-year observations in the accruals-difference sample Total unique client firm-pairs in the same industry-year and with auditors of similar size for the accrualsdifference test 23,707 (346) (2,236) (5,410) 15,715 267,312 surprising that only a small proportion of firm-pairs share common auditors Specifically, 2.1 percent of firm-pairs use the same audit firm, but are audited by different audit offices and different signing auditors (SameFirm_DiffOffice_DiffPartner ¼ 1), 5.6 percent of firm-pairs are audited by the same audit office, but have different signing auditors (SameOffice_DiffPartner ¼ 1), and 0.5 percent of firm-pairs have the same signing auditor (SamePartner ¼ 1) Of the firm-pairs with the same signing auditor, 80 percent are audited by different audit teams (mean of SamePartner_DiffTeam ¼ 0.4 percent), and 20 percent are audited by the same audit team (mean of SameTeam ¼ 0.1 percent) After we drop firm-pairs with the same audit team, we find that 0.3 percent of the remaining firm-pairs have a common senior signing auditor, but different junior signing auditors (SamePartner_Senior ¼ 1), and 0.1 percent have a common junior signing auditor, but different senior signing auditors (SamePartner_Junior ¼ 1).18 Panel B of Table reports the mean values of all variables separately for firm-pairs audited by (1) different audit firms, (2) the same audit firm, but different audit offices and individual auditors, (3) the same audit office, but different individual auditors, and (4) the same individual auditor We find that relative to firm-pairs with different audit firms, those with the same signing auditor exhibit a smaller difference in total and abnormal accruals and a higher degree of earnings co-movement, suggesting an individual auditor style effect on earnings comparability We also find that earnings co-movement is higher for firm-pairs audited by the same individual auditor than for both firm-pairs audited by the same audit firm, but different audit offices, and firm-pairs audited by the same audit office, but different individual auditors Although the corresponding results based on the two accrual-difference metrics are in the same direction, they are not statistically significant Finally, several client firm and auditor characteristics differ systematically among the four groups, highlighting the importance of controlling for these variables in our regression analyses Panel C of Table reports the univariate test results of H2 For the subsample of firm-pairs with the same signing auditor, we report the mean values of À1 Tacc_Diff, À1 AbnAcc_Diff, and EComp_Cov separately for clients of the Top Ten audit firms and clients of the non-Top Ten audit firms The Top Ten client firm-pairs that have a common signing auditor exhibit higher earnings comparability, regardless of how we measure it, than the non-Top Ten client firm-pairs audited by a common signing auditor Panel D reports the mean values of the three comparability measures separately for firm-pairs with the same senior signing auditor and firm-pairs with the same junior signing auditor We find a higher degree of earnings co-movement for firm-pairs with the same senior signing auditor, consistent with H3 However, we fail to detect a reliable difference in either total or abnormal accruals closeness between the two groups In Panel E, we focus on firmpairs with the same audit team and further partition this sample into groups whose auditors have more and less stable teamwork experience H4 predicts that comparability is higher for firm-pairs whose common signing auditor has more stable teamwork experience Although the differences in all three comparability measures have the expected sign, they are not statistically significant Panel F of Table reports Spearman correlations between À1 Tacc_Diff, À1 AbnAcc_Diff, and EComp_Cov The correlation between the two accruals-based measures (0.72) is significantly positive, as expected The correlation between À1 Tacc_Diff (À1 AbnAcc_Diff ) and EComp_Cov also exhibits the predicted positive sign and is statistically significant However, its magnitude is small 18 The small sample sizes for firm-pairs with SameTeam ¼ 1, SamePartner_Senior ¼ 1, and SamePartner_Junior ¼ are not surprising because each senior ( junior) signing auditor, on average, audits only 1.09 (1.03) public firms in an industry-year over our sample period Despite the statistical significance, our results should be interpreted with care because of the small sample sizes for the cross-sectional tests The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 124 TABLE Descriptive Statistics Panel A: Summary Statistics for Full Sample Variable Mean STD 25% Median 75% À1 Tacc_Diff À1 AbnAcc_Diff EComp_Cov SameFirm_DiffOffice_DiffPartner SameOffice_DiffPartner SamePartner SamePartner_Junior SamePartner_Senior SamePartner_Others SamePartner_DiffTeam SameTeam Tenure_Diff Tenure_Min IndustrySpec Tacc_Min AbnAcc_Min Size_Diff Size_Min Lev_Diff Lev_Min MB_Diff MB_Min CFO_Diff CFO_Min LossProb_Diff LossProb_Min Std_Sales_Diff Std_Sales_Min Std_CFO_Diff Std_CFO_Min Std_SalesGr_Diff Std_SalesGr_Min CFOComp_Cov RetComp_Cov À0.109 À0.089 0.261 0.021 0.056 0.005 0.001 0.003 0.001 0.004 0.001 0.429 1.048 0.058 À0.057 À0.046 1.176 21.019 0.247 0.370 3.131 1.924 0.104 À0.007 0.233 0.062 0.035 0.027 0.034 0.027 6.225 0.373 0.266 0.303 0.124 0.085 0.262 0.144 0.231 0.070 0.024 0.050 0.036 0.065 0.024 0.326 0.297 0.234 0.091 0.068 0.966 0.919 0.342 0.187 8.774 2.455 0.109 0.098 0.244 0.126 0.047 0.017 0.063 0.016 51.865 0.517 0.264 0.191 À0.140 À0.119 0.040 0 0 0 0 0.182 0.693 À0.095 À0.077 0.445 20.440 0.085 0.218 0.503 1.265 0.033 À0.039 0.000 0.000 0.009 0.015 0.008 0.016 0.086 0.153 0.041 0.150 À0.073 À0.064 0.169 0 0 0 0 0.406 1.099 À0.044 À0.036 0.946 20.943 0.185 0.369 1.203 1.786 0.073 0.010 0.125 0.000 0.020 0.023 0.019 0.023 0.220 0.225 0.174 0.289 À0.033 À0.029 0.422 0 0 0 0 0.619 1.253 À0.003 À0.003 1.651 21.569 0.324 0.510 2.702 2.641 0.135 0.048 0.375 0.125 0.042 0.034 0.039 0.034 0.636 0.361 0.434 0.440 (continued on next page) V MAIN RESULTS The Individual Auditor Style Effect on Earnings Comparability (H1) Table 3, Panel A reports the regression results for testing H1 In the first column, where we measure comparability using À1 Tacc_Diff, we find significantly positive coefficients on SameOffice_DiffPartner (b2 ¼ 0.006, t ¼ 6.13) and SamePartner (b3 ¼ 0.013, t ¼ 6.44), but not on SameFirm_DiffOffice_DiffPartner (b1 ¼ 0.002, t ¼ 1.09).19 These results indicate that audit office style and individual auditor style each have a significant impact on client firms’ earnings comparability and dominate the 19 Using a U.S Big sample, Kawada (2014) finds that firm-pairs audited by the same Big office report more comparable earnings, but does not find a significant effect of having the same audit firm, but different audit offices on client firms’ earnings comparability Therefore, our finding that the audit office style and individual auditor style dominate the audit firm style in shaping client firms’ earnings comparability may not be a unique phenomenon in the Chinese context In addition, if we not isolate the audit office style and individual auditor style from the audit firm style, then we find results consistent with FPW, that firm-pairs audited by the same audit firm (including firm-pairs audited by the same audit office and the same individual auditor) have higher earnings comparability The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 130 TABLE Relation between Individual Auditor Style and Earnings Comparability Common Senior Signing Auditor versus Common Junior Signing Auditor (H3) Panel A: Regression Results À1 Tacc_Diff Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner_Junior (b3) SamePartner_Senior (b4 ) SamePartner_Others À1 AbnAcc_Diff EComp_Cov Coeff t-stat Coeff t-stat Coeff t-stat 0.103*** 0.002 0.007*** 0.009** 0.017*** 0.006 3.97 1.10 6.13 2.33 8.06 1.50 0.035** 0.001 0.002*** 0.007** 0.013*** 0.001 2.29 1.39 3.91 2.40 7.93 0.52 0.170** 0.004 0.022*** 0.069* 0.138*** À0.010 2.21 0.72 4.82 1.90 6.12 À0.39 Controls Industry and Year FE R2 Number of Firm-Pairs Yes Yes 0.680 267,153 Yes Yes 0.724 267,153 Yes Yes 0.051 98,903 Panel B: Test of Coefficient Differences in Panel A À1 Tacc_Diff b2 À b1 ¼ b3 À b2 ¼ b4 À b3 ¼ À1 AbnAcc_Diff EComp_Cov Value t-stat Value t-stat Value t-stat 0.005*** 0.002 0.008* 2.68 0.62 1.80 0.001 0.005* 0.006* 1.41 1.64 1.66 0.018** 0.047 0.069* 2.35 1.28 1.65 ***, **, * Indicate significance levels at percent, percent, and 10 percent, respectively (two-tailed) This table examines the effect of having a common senior signing auditor versus the effect of having a common junior signing auditor on client firms’ earnings comparability For this test, we drop firm-pairs having a common senior signing auditor and a common junior signing auditor (159 firm-pairs for the accruals-difference tests, and 42 firm-pairs for the earnings co-movement test) We decompose the rest of the firm-pairs with a common signing auditor (SamePartner ¼ 1) into three subsamples: firm-pairs with a common senior signing auditor, but different junior signing auditors (SamePartner_Senior ¼ 1), firm-pairs with a common junior signing auditor, but different senior signing auditors (SamePartner_Junior ¼ 1), and all other firm-pairs with a common signing auditor (SamePartner_Others ¼ 1) Earnings comparability is measured by À1 Tacc_Diff in the first column, À1 AbnAcc_Diff in the second column, and EComp_Cov in the third column For brevity, control variables are not tabulated t-values are calculated based on robust standard errors clustered at the client firm-pair level See Appendix A for variable definitions Taken together, the results of H2 and H3 are consistent with individual auditors who are more competent (e.g., Top Ten employees) and more experienced (e.g., senior signing auditors) exhibiting a stronger audit style Our reasoning for H2 and H3 implies that the individual auditor style is at least partially influenced by the auditor’s demographic characteristics, such as innate ability and learned skills.21 In untabulated results, we find corroborating evidence that a signing auditor who has an advanced education degree, stays longer in the current job, or is a former Top Ten audit firm employee exhibits a stronger audit style, resulting in higher earnings comparability for her clients.22 To the extent that large audit firm employees and senior signing auditors are more competent, confident, and experienced, the results of H2 and H3 lend support to the idea that both innate ability and learned skills play important roles in shaping individual auditor style Our last hypothesis, H4, examines whether auditors’ stable teamwork experience solidifies their work styles and leads to higher earnings comparability for client firm-pairs We report the results in Table In Panels A and B, we focus on À1 Tacc_ Diff and estimate the revised Model (3) separately for firm-pairs whose audit teams have more and less stable teamwork experience H4 predicts that the coefficient on SameTeam will be larger if an individual signing auditor is more likely to form a 21 22 Innate ability cannot be altered by the audit firm By contrast, learned skills are a personal trait that can be developed through rigorous on-the-job training Specifically, we regress client firms’ earnings comparability on the following seven variables that capture the common signing auditor’s innate ability and learned skills: 985University, Graduate_Degree, Female, Education_Cohort, CurrentJob_Exp, PriorTop10_Exp, and Rank, after controlling for other variables that affect earnings comparability from Model (3) Variable definitions are detailed in Appendix A Results are available upon request The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 131 TABLE Relation between Individual Auditor Style and Earnings Comparability More Stable versus Less Stable Audit Teamwork Experience (H4) Panel A: Regression Results (Measure of Earnings Comparability ¼ À1 Tacc_Diff ) More Stable Teamwork Experience Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner_DiffTeam (b3) SameTeam (b4 ) Less Stable Teamwork Experience Coeff t-stat Coeff t-stat 0.108*** 0.002 0.007*** 0.016*** 0.031*** 3.07 0.61 4.72 6.05 6.69 0.101*** 0.001 0.006*** 0.011*** 0.016*** 3.86 0.82 5.48 4.70 3.99 Controls Industry and Year FE R2 Number of Firm-Pairs Yes Yes 0.67 78,905 Yes Yes 0.69 188,407 Panel B: Test of Coefficient Differences in Panel A More Stable Teamwork Experience b2 b3 b4 b1 b2 b3 b4 À b1 ¼ À b2 ¼ À b3 ¼ (More Stable) (More Stable) (More Stable) (More Stable) À À À À b1 b2 b3 b4 (Less (Less (Less (Less Stable) Stable) Stable) Stable) ¼ ¼ ¼ ¼ Less Stable Teamwork Experience Value t-stat Value t-stat 0.005 0.009*** 0.015*** 1.62 3.34 3.01 0.005 0.005* 0.005 0.001 0.001 0.005* 0.015** 1.60 1.91 1.22 0.10 0.52 1.70 2.40 0 0 Panel C: Regression Results (Measure of Earnings Comparability ¼ À1 AbnAcc_Diff ) More Stable Teamwork Experience Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner_DiffTeam (b3) SameTeam (b4 ) Controls Industry and Year FE R2 Number of Firm-Pairs Less Stable Teamwork Experience Coeff t-stat Coeff t-stat 0.057*** 0.001 0.003*** 0.011*** 0.018*** 3.30 1.14 3.73 5.97 5.20 0.026 0.001 0.002*** 0.009*** 0.007** 1.53 1.44 3.22 4.89 2.02 Yes Yes 0.72 78,905 Yes Yes 0.73 188,407 (continued on next page) consistent work style after collaborating with the same colleague for relatively longer Consistent with this prediction, we find that the coefficient on SameTeam is significantly larger when the common audit team has more stable teamwork experience than when it has less stable teamwork experience In addition, we find that the coefficient on SameTeam is significantly larger than the coefficient on SamePartner_DiffTeam for firm-pairs with more stable audit teams, suggesting that having the same stable audit team has a larger impact on client firms’ earnings comparability than having only one common signing auditor The results based on À1 AbnAcc_Diff and EComp_Cov, reported in Panels C and D, and Panels E and F, respectively, are qualitatively the same The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 132 TABLE (continued) Panel D: Test of Coefficient Differences in Panel C More Stable Teamwork Experience b2 b3 b4 b1 b2 b3 b4 À b1 ¼ À b2 ¼ À b3 ¼ (More Stable) (More Stable) (More Stable) (More Stable) À À À À b1 b2 b3 b4 (Less (Less (Less (Less Stable) Stable) Stable) Stable) ¼ ¼ ¼ ¼ Less Stable Teamwork Experience Value t-stat Value t-stat 0.002 0.008*** 0.007* 1.36 3.99 1.82 0.001 0.007*** À0.002 0.000 0.001 0.002 0.011** 0.94 3.77 À0.46 0.01 0.90 0.70 2.25 0 0 Panel E: Regression Results (Measure of Earnings Comparability ¼ EComp_Cov) More Stable Teamwork Experience Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner_DiffTeam (b3) SameTeam (b4 ) Controls Industry and Year FE R2 Number of Firm-Pairs Less Stable Teamwork Experience Coeff t-stat Coeff t-stat 0.286** 0.009 0.011 0.057*** 0.285*** 2.51 0.79 1.42 2.69 3.34 0.138* 0.002 0.006 0.034* 0.062* 1.84 0.30 1.17 1.78 1.70 Yes Yes 0.05 28,228 Yes Yes 0.05 70,717 Panel F: Test of Coefficient Differences in Panel E More Stable Teamwork Experience b2 b3 b4 b1 b2 b3 b4 À b1 ¼ À b2 ¼ À b3 ¼ (More Stable) (More Stable) (More Stable) (More Stable) À À À À b1 b2 b3 b4 (Less (Less (Less (Less Stable) Stable) Stable) Stable) ¼ ¼ ¼ ¼ 0 0 Less Stable Teamwork Experience Value tÀstat Value tÀstat 0.002 0.046** 0.228** 0.16 2.10 2.53 0.004 0.028 0.028 0.007 0.005 0.023 0.223** 0.49 1.41 0.68 0.57 0.58 0.81 2.05 ***, **, * Indicate significance levels at percent, percent, and 10 percent, respectively (two-tailed) This table separately examines the effect of having a common signing auditor on earnings comparability for client firm-pairs whose signing auditors have more stable teamwork experience and client firm-pairs whose signing auditors have less stable teamwork experience For each client firm in the pair, we count the number of years that its two signing auditors worked together as a team on any audit engagement in the past For the client firm-pair, we use the minimum value for the two client firms as their measure of signing auditors’ stable teamwork experience Client firm-pairs that have stable teamwork experience above the sample median in each year are classified in the more stable teamwork experience subsample, and the rest of the firm-pairs are classified in the less stable teamwork experience subsample We decompose the firm-pairs with a common signing auditor (SamePartner ¼ 1) into two subsamples: firm-pairs with a common audit team (SameTeam ¼ 1) and firm-pairs with only one common signing auditor (SamePartner_DiffTeam ¼ 1) Earnings comparability is measured by À1 Tacc_Diff in Panels A and B, À1 AbnAcc_Diff in Panels C and D, and EComp_Cov in Panels E and F For brevity, control variables are not tabulated t-values are calculated based on robust standard errors clustered at the client firm-pair level See Appendix A for variable definitions The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 133 Given the above results, a natural question that arises is whether auditors’ teamwork experience subsumes their individual work experience in implementing earnings comparability To answer this question, we repeat the analyses in Table after classifying firm-pairs with SamePartner_DiffTeam ¼ into three groups: firm-pairs with SamePartner_Senior ¼ 1, firm-pairs with SamePartner_Junior ¼ 1, and the remaining firm-pairs (SamePartner_Others ¼ 1) Untabulated results indicate that in the more stable teamwork experience subsample, earnings are more comparable for firm-pairs audited by the same audit team than for firm-pairs audited by the same senior signing auditor, but different audit teams This difference disappears in the less stable teamwork experience subsample Furthermore, we find that senior signing auditors’ greater influence over client firms’ closeness of (total and abnormal) accruals relative to junior signing auditors’ influence persists only in the less stable teamwork experience subsample Overall, these results suggest that whereas the senior signing auditors always exhibit a unique audit style, a stable audit team reduces the difference in the individual auditor style effect on earnings comparability between senior and junior signing auditors VI ADDITIONAL ANALYSES Self-Selection Our empirical analyses are susceptible to a potential self-selection problem because client firms are not randomly assigned to individual auditors; rather, they select into a specific auditor’s client portfolio based on firm characteristics, auditor features, private information, or other unobservable factors If the decisions to hire certain individual auditors are correlated across firmpairs, then our OLS regressions that ignore self-selection will yield biased results In this section, we perform further analyses to alleviate the self-selection concern Removing Firm-Pairs with Similar Interests in Hiring Individual Auditor Specialists One possible alternative explanation for our results is that two client firms with similar firm characteristics have a similar demand for auditors with industry expertise Thus, earnings comparability between the two client firms may not reflect the impact of individual auditor style; instead, it may merely capture the effect of some unobservable firm characteristics that are common between them To ensure that our results are not affected by this type of selection bias, we remove from the sample firm-pairs that hire the same individual auditor who is an industry specialist.23 We reestimate Model (3) and find qualitatively the same results (untabulated) Mandatory Audit Partner Rotation In 2003, the China Securities Regulatory Commission and the Ministry of Finance jointly issued a rule that requires individual auditors who sign the audit report of a listed company to be rotated off after five years To the extent that mandatory partner rotations are relatively independent of the client firm and auditor characteristics, our analyses of the relation between individual auditor style and earnings comparability in this context should be less subject to self-selection bias In each year, we identify firm-pairs with a change in at least one signing auditor due to scheduled mandatory rotation (i.e., the signing auditor has reached the maximum allowable length of tenure) (Lennox et al 2014) Then, we estimate the following regression:   À1 Tacc Diff ijt or À AbnAcc Diff ijt ẳ d1 Change SamePartnerjt ỵ d2 Change Diff erentPartnerjt ỵ d3 PostChange SamePartnerjt ỵ d4 PostChange Diff erentPartnerjt ỵ Controls þ eijt ð4Þ where Change_SamePartner (Change_DifferentPartner) is an indicator variable that equals for firm-pairs that switch from having different signing auditors (the same signing auditor) to having the same signing auditor (different signing auditors), and otherwise PostChange_SamePartner (PostChange_DifferentPartner) is an indicator variable that equals in the period after 23 We calculate each individual auditor’s market share in a given industry-year as the proportion of total sales audited by that auditor to total sales of all firms in that industry-year We define industry specialists as the top three individual auditors who have the largest market shares In sensitivity tests, we redefine industry specialists as individual auditors whose market shares are in the top quintile of the individual market share distribution We also recalculate market share as the proportion of total assets audited by each individual auditor to total assets of all firms in a given industry-year Our results are robust to these alternative proxies for industry expertise The Accounting Review Volume 95, Number 3, 2020 134 Chen, Chen, Chin, and Lobo firm-pairs experience a change from having different signing auditors (the same signing auditor) to having the same signing auditor (different signing auditors), and otherwise For this test, we exclude firm-pairs that switch from having one common signing auditor to having another common signing auditor after the mandatory rotation, and firm-pairs that have different signing auditors both before and after the mandatory rotation We expect d3 to be positive if switching to having the same individual auditor style increases earnings comparability for firm-pairs By contrast, d4 should be negative if dropping the same individual auditor style reduces earnings comparability.24 We identify 244 and 349 firm-pairs that experience a Change_SamePartner and a Change_DifferentPartner, respectively, and report the results of estimating Model (4) in Table 7, Panels A and B We compare earnings comparability over one year, two years, and three years before and after the mandatory rotation Our results are qualitatively the same for both comparability measures For brevity, we only tabulate the results based on À1 AbnAcc_Diff We find a significantly positive d3 and its magnitude does not vary much as we expand the time horizon d4 is significantly negative across all time horizons Our test easily rejects the null that d3 equals d4 These results suggest that the direction of causality is likely to run from individual auditor style to client firms’ earnings comparability.25 Mandatory IFRS Adoption We use the mandatory adoption of IFRS in China as another exogenous shock to draw causal inferences China’s Ministry of Finance amended the Accounting Standards for Business Enterprises, which became effective in 2007 The amended standards are substantially in line with IFRS Prior to the convergence toward IFRS, Chinese GAAP was similar to U.S GAAP in that it provided more specific guidelines for firms to follow when preparing financial statements With less implementation or interpretive guidance from standard setters and regulators in the post-IFRS era, external audit is likely to play a more important role in ensuring financial reporting credibility A persistent audit style ensures consistent application of IFRS, which helps achieve earnings comparability Under IFRSbased reporting, there are more instances of judgments and more instances when the auditor’s judgments differ from management’s judgments Such differences must be negotiated as part of the audit process Inherent personal characteristics significantly impact the auditor-client negotiation process (Brown and Wright 2008) If signing auditors exhibit unique audit styles, then IFRS reporting provides them with more room to exercise professional judgment and control over how to implement and interpret accounting standards Accordingly, we expect a larger effect of individual auditor style on audit outcome after the adoption of IFRS We partition the sample into a pre-IFRS subsample (2003 to 2006) and a post-IFRS subsample (2008 to 2011) and use the same number of years in the two subsamples to ensure a balanced comparison.26 We reestimate Model (3) separately for the two subsamples and report the results in Table 7, Panels C and D The results are qualitatively the same across the three comparability measures, so we only tabulate the results based on À1 AbnAcc_Diff for brevity We find consistent evidence that audit office style and individual auditor style become more effective in the post-IFRS period To the extent that the mandatory adoption of IFRS is not related to client firm-level and individual auditor-level factors that shape auditor hiring decisions, these results provide a strong basis for inferring the direction of causality Comparability of Accounting Systems If similar economic fundamentals drive client firms’ decisions to hire a common auditor, then using closeness of accruals and earnings covariance to infer the individual auditor style effect may exacerbate the concern about self-selection bias We address this concern by performing a sensitivity test using De Franco et al.’s (2011) approach to measure comparability of the accounting system The rationale of this measure is that two firms with comparable earnings should have a similar measurement (mapping) function, such that for a given set of economic events, the two firms report similar earnings It is reasonable to argue that individual auditors can exert influence over the client firms’ accounting systems, but it is unlikely that two client firms will choose the same individual auditor simply because of their similar accounting systems 24 25 26 We not conduct these analyses using the earnings co-movement measure because it requires eight consecutive quarters to estimate earnings covariation, and signing auditor changes may occur during the eight-quarter window We note that the choice of the succeeding signing auditors after the mandatory rotation may still lie with the client and the audit firm Therefore, firmpairs that experience a Change_SamePartner may be subject to self-selection However, for firm-pairs that experience a Change_DifferentPartner, the choice of a succeeding (different) signing auditor is unlikely to be driven by any economic comparability that shapes the decision to hire a common signing auditor before the mandatory rotation In other words, we believe that the self-selection bias due to economic comparability, if any, in the prerotation period is of less concern in the post-rotation period for this subsample Because 2007 is the first year of IFRS adoption, we drop this transition year from our analyses Our results for the post-IFRS period are qualitatively unchanged if we include all data from 2008 to 2013 in our sample The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 135 TABLE Relation between Individual Auditor Style and Earnings Comparability: Addressing Endogeneity Panel A: Mandatory Signing Auditor Rotation—Switching to Having the Same (Different) Signing Auditors (Dependent Variable ¼ À1 AbnAcc_Diff ) [t1, tỵ1] Variable Change_SamePartner (d1) Change_DifferentPartner (d2) PostChange_SamePartner (d3) PostChange_DifferentPartner (d4 ) [t2, tỵ2] [t3, tỵ3] Coeff t-stat Coeff t-stat Coeff t-stat À0.047 À0.040 0.006** À0.007*** À1.22 À1.04 2.30 À2.84 À0.004 0.004 0.007*** À0.006*** À0.11 0.10 2.84 À2.83 À0.010 À0.003 0.006** À0.007*** À0.29 À0.09 2.56 À3.12 Controls Industry and Year FE R2 Number of Firm-Pairs Yes Yes 0.76 1,777 Yes Yes 0.77 2,302 Yes Yes 0.76 2,604 Panel B: Test of Coefficient Differences in Panel A [t1, tỵ1] d3 d4 ẳ [t2, tỵ2] [t3, tỵ3] Value t-stat Value t-stat Value t-stat 0.013*** 3.28 0.013*** 3.69 0.013*** 3.67 Panel C: Mandatory IFRS Adoption (Dependent Variable ¼ À1 AbnAcc_Diff ) Pre-IFRS (2003–2006) Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner (b3) Post-IFRS (2008–2011) Coeff t-stat Coeff t-stat 0.014 À0.001 0.002** 0.010*** 0.56 À0.56 2.12 5.27 0.032 0.001 0.006*** 0.018*** 1.19 1.03 5.46 7.80 Controls Industry and Year FE R2 Number of Firm-Pairs Yes Yes 0.76 63,711 Yes Yes 0.69 90,805 Pre-IFRS (2003–2006) Post-IFRS (2008–2011) Panel D: Test of Coefficient Differences in Panel C b2 b3 b1 b2 b3 À b1 ¼ À b2 ¼ (Post-IFRS) À b1 (Pre-IFRS) ¼ (Post-IFRS) À b2 (Pre-IFRS) ¼ (Post-IFRS) À b3 (Pre-IFRS) ¼ Value t-stat Value t-stat 0.003 0.008*** 1.27 3.90 0.005*** 0.012*** 0.002 0.004** 0.008*** 2.74 5.42 0.93 2.09 2.78 (continued on next page) To measure the accounting system of firm i, we estimate the following firm-specific regression using firm i’s eight previous quarters of earnings and stock returns in each year: Earningsit ẳ ỵ bi Returnit ỵ eit 5ị where Return is the raw stock return during quarter t The estimated coefficients, a^i and b^i , are firm i’s accounting system that The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 136 TABLE (continued) Panel E: Using the De Franco et al (2011) Measure of Earnings Comparability (Dependent Variable ¼ CompAcct) Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner (b3) Controls Industry and Year FE R2 Number of Firm-Pairs Coeff t-stat 0.055*** À0.000 0.001*** 0.002*** 10.98 À0.16 2.69 3.63 Yes Yes 0.354 216,592 Panel F: Test of Coefficient Differences in Panel E b2 À b1 ¼ b3 À b2 ¼ Value t-stat 0.001** 0.001*** 2.00 2.76 Panel G: Mean Values of the Comparability Measures for the Propensity Score Matched Samples À1 Tacc_Diff À1 AbnAcc_Diff EComp_Cov Firm-pairs with the same signing auditor (4) Firm-pairs with different audit firms (1) (4) À (1) À0.097 À0.113 0.016*** À0.080 À0.092 0.012*** 0.322 0.261 0.061*** Firm-pairs with the same signing auditor (4) Firm-pairs with the same audit firm, but different audit offices (2) (4) À (2) À0.094 À0.108 0.014*** À0.079 À0.087 0.008** 0.326 0.249 0.077*** Firm-pairs with the same signing auditor (4) Firm-pairs with the same audit office, but different signing auditors (3) (4) À (3) À0.096 À0.108 0.012*** À0.079 À0.088 0.009*** 0.321 0.258 0.063*** Group ***, ** Indicate significance levels at percent and percent, respectively (two-tailed) Panels A and B report regression analyses for firm-pairs that switch from having the same signing auditor to different signing auditors, and firm-pairs that switch from having different signing auditors to the same signing auditor due to the mandatory rotation requirements We examine the difference in earnings comparability one year (Column 1), two years (Column 2), and three years (Column 3) before and after the mandatory rotation For brevity, we only report the results when we measure comparability by À1 AbnAcc_Diff Panels C and D examine the effect of having a common signing auditor on earnings comparability for client firm-pairs in the pre-IFRS period, and client firm-pairs in the post-IFRS period The pre-IFRS period covers 2003 to 2006, and the post-IFRS period covers 2008 to 2011 For brevity, we only report the results when we measure comparability by À1 AbnAcc_Diff Panels E and F examine the relation between client firms’ accounting system comparability and individual auditor style We measure comparability of accounting systems between client firm-pairs using CompAcct, consistent with De Franco et al (2011) For brevity, control variables are not tabulated Panel G reports the mean values of the comparability measures for the propensity score matched samples We separately match firm-pairs audited by the same individual auditor (group (4)) to firm-pairs audited by different audit firms (group (1)), to firm-pairs audited by the same audit firm, but different audit offices (group (2)), and to firm-pairs audited by the same audit office, but different individual auditors (group (3)) We form matched pairs by identifying the pairings that result in observations with the smallest propensity score differences for all control variables used in Model (3) t-tests are used to test the differences between the means In Panels A–F, t-values are calculated based on robust standard errors clustered at the client firm-pair level See Appendix A for variable definitions maps its economic events into its earnings Similarly, we obtain estimates of firm j’s accounting system through a^j and b^j We use firm i’s and firm j’s respective accounting systems to predict their earnings, E(Earnings)iit and E(Earnings)ijt, assuming that they have the same Returnit Our measure of accounting systems comparability (CompAcct) is the negative value of the average absolute difference between E(Earnings)iit and E(Earnings)ijt over the previous eight quarters In Table 7, Panels E and F, we reestimate Model (3) using CompAcct as the dependent variable We continue to find a significant individual auditor style effect, and it is larger than the effects of audit office style and audit firm style on comparability This test provides some comfort that our main results in Table not merely reflect a spurious correlation between firm-pairs’ economic comparability and their auditor choices The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 137 Propensity Score Matching An implicit assumption of Model (3) is that there is an identical functional relation between the control variables and earnings comparability for firm-pairs in each group If the true relation either differs across firm-pairs subject to different types of audit style or is inconsistent with the functional form imposed by our research design, then the regression results will produce biased parameter estimates We use the propensity score matching technique to alleviate the concern about functional form misspecification (Shipman et al 2017) Specifically, we separately match firm-pairs audited by the same individual auditor to firm-pairs audited by different audit firms, to firm-pairs audited by the same audit firm, but different audit offices, and to firm-pairs audited by the same audit office, but different individual auditors We form matched pairs by identifying the pairings that result in observations with the smallest propensity score differences for all control variables used in Model (3) Table 7, Panel G reports the mean values of our comparability measures for the matched samples We continue to find that firm-pairs audited by the same individual auditor report the most comparable earnings among all firm-pairs in the various matched samples.27 Earnings Comparability or Earnings Uniformity? An alternative possibility for our findings is that common audit style leads to uniformity among client firms’ reporting Uniformity makes unlike things look alike, which impairs comparability and informativeness In this section, we perform analyses to differentiate the two opposite implications of individual auditor style for client firms’ earnings informativeness Specifically, we address this question from financial analysts’ perspective Our motivation stems from De Franco et al.’s (2011) result that earnings comparability lowers the cost of acquiring information and increases the overall information about the firm to analysts They find that analyst forecast accuracy increases and dispersion in earnings forecasts decreases in earnings comparability Accordingly, we predict that if two client firms subject to the same individual auditor style have more comparable earnings, then they will have greater analyst forecast accuracy and lower forecast dispersion By contrast, if common audit style impairs earnings informativeness by forcing uniform GAAP applications, then we should detect no difference in analyst forecast accuracy or forecast dispersion or even differences in the opposite direction We estimate the following regression model: À Á FError Avgijt or Dispersion Avgijt ẳ b0 ỵ b1 SameFirm Diff Off ice Diff Partnerijt þ b2 SameOff ice Diff Partnerijt þ b3 SamePartnerijt þ Controls ỵ eijt 6ị where FError_Avgijt is the average absolute value of the earnings forecast errors in year t, divided by stock price at the end of year tÀ1, for firm i and firm j in the pair; and Dispersion_Avgijt is the average value of the standard deviation across all analysts’ earnings forecasts, divided by stock price at the end of year tÀ1, for firm i and firm j in the pair.28 Controls can be broadly classified into two groups The first group of control variables includes all the control variables in Model (3) The second group is based on prior literature on the determinants of analyst forecast accuracy and dispersion (Behn, Choi, and Kang 2008) They include both the difference in and the minimum level of analyst following, forecast horizon, earnings volatility, return on equity, and EPS for the firm-pair Table 8, Panels A and B report the results of estimating Model (6) and testing coefficient differences, respectively We find a significantly negative relation between FError_Avg and SamePartner (b3 ¼ À0.001, t ¼ À2.59) in the first column, and a similar negative relation between Dispersion_Avg and SamePartner (b3 ¼À0.001, t ¼À2.62) in the second column However, we fail to detect a significant effect of audit firm style or audit office style on the two analyst forecast properties, as both b1 and b2 are insignificant Overall, our results are consistent with having a common signing auditor for client firm-pairs enhancing, as opposed to impairing, earnings informativeness VII CONCLUSIONS We examine whether individual auditors influence client firms’ earnings comparability After distinguishing between unique audit styles at different levels, we find that the effect of individual auditor style dominates the effects of audit office style and audit firm style in the implementation of earnings comparability We further show that the effect of individual auditor style on earnings comparability is more prominent for large audit firms than for small audit firms, consistent with the notion that the signing auditors of large audit firms are more confident of their personal work styles We also find that both signing auditors on 27 28 In untabulated analyses, we also estimate OLS regressions of our comparability measures on SamePartner and the control variables in Model (3) for each matched sample We continue to find a significantly positive coefficient on SamePartner in each matched sample regardless of how we measure comparability We include only client firms with at least three analyst forecasts in a given year in the sample for this test The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 138 TABLE Relation between Individual Auditor Style and Client Firm-Pairs’ Average Analyst Forecast Errors and Dispersion Panel A: The Effect of Having a Common Signing Auditor on the Average Analyst Forecast Error and Dispersion for Client Firm-Pairs FError_Avg Variable Intercept SameFirm_DiffOffice_DiffPartner (b1) SameOffice_DiffPartner (b2) SamePartner (b3) Tenure_Diff Tenure_Min IndustrySpec Size_Diff Size_Min Lev_Diff Lev_Min MB_Diff MB_Min LossProb_Diff LossProb_Min Std_Sales_Diff Std_Sales_Min Std_CFO_Diff Std_CFO_Min Std_SalesGr_Diff Std_SalesGr_Min Numest_Diff Numest_Min AvgHorizon_Diff AvgHorizon_Min Dispersion_Diff Dispersion_Min FError_Diff FError_Min StdROE_Diff StdROE_Min EPS_Diff EPS_Min Dispersion_Avg Coeff t-stat Coeff t-stat À0.038*** À0.000 À0.000 À0.001** À0.000 0.000 À0.000 À0.001*** À0.000 0.004*** 0.005*** À0.000 À0.000** 0.002** 0.011*** 0.004* 0.008 0.000 0.004 À0.000** 0.000 À0.001*** À0.001*** 0.007*** 0.008*** 0.274*** 0.701*** À5.78 À0.26 À1.13 À2.59 À0.27 0.54 À0.09 À3.65 À1.54 3.87 4.80 À1.20 À2.39 1.97 5.89 1.74 0.95 0.11 0.43 À1.98 0.88 À5.67 À4.73 9.04 11.78 16.25 21.93 À0.015*** À0.000 À0.000 À0.001*** À0.000 0.000 À0.001 0.000 0.000 0.003*** 0.004*** 0.000* À0.000 0.005*** 0.008*** 0.004** 0.025*** À0.004** À0.018*** À0.000 À0.000 À0.000* 0.001* 0.001* 0.001*** À2.80 À0.05 À0.46 À2.62 À1.36 0.81 À1.47 0.82 1.10 3.13 3.81 1.70 À1.62 3.82 4.73 2.56 5.14 À2.23 À2.71 À1.16 À0.91 À1.76 1.86 1.75 2.75 0.035 0.437*** À0.016 0.009 0.047*** 0.000 0.90 20.37 À0.31 0.04 8.87 0.00 0.052 0.583* 0.068*** À0.067*** Industry and Year FE R2 Number of Firm-Pairs 1.27 1.91 6.38 À5.24 Yes 0.575 44,591 Yes 0.330 44,591 Panel B: Test of Coefficient Differences in Panel A FError_Avg b2 À b1 ¼ b3 À b2 ¼ Dispersion_Avg Coeff t-stat Coeff t-stat À0.000 À0.001** À0.44 À2.04 À0.000 À0.001** À0.22 À2.29 ***, **, * Indicate significance levels at percent, percent, and 10 percent, respectively (two-tailed) This table examines the effect of having a common signing auditor on the average analyst forecast error and dispersion for client firm-pairs The sample of this test consists of all firm-pairs with sufficient analyst earnings forecast data t-values are calculated based on robust standard errors clustered at the client firm-pair level See Appendix A for variable definitions The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 139 the engagement team can affect client firms’ earnings comparability, but the individual effect is more prominent for the senior signing auditor This result suggests that more experienced auditors are likely to have a more mature and confident audit style that ensures more consistent and accurate application of GAAP across different audit engagements Finally, we find that stable teamwork experience enhances the effect of individual auditor style on client firms’ earnings comparability, consistent with individual auditors strengthening their personal work styles through social learning within the engagement team Two caveats of our study are in order First, we conduct our analyses in the Chinese setting, where some institutional structures differ from those in the U.S Therefore, our findings may not be readily generalizable, and the external validity of our findings should be assessed before applying them to other institutional settings Second, despite our best efforts to address the self-selection bias, we recognize that this problem may be inherent in our study and cannot be resolved completely The literature does not provide much guidance on how to model partner-client selection choices However, with partner-level data now available in the U.S., we expect that researchers will be able to more accurately pinpoint client- and partner-specific characteristics that drive individual auditor hiring decisions REFERENCES Abdolmohammadi, M., and A Wright 1987 An examination of the effects of experience and task complexity on audit judgments The Accounting Review 62 (1): 1–13 Amer, T., K Hackenbrack, and M Nelson 1994 Between-auditor differences in the interpretation of probability phrases Auditing: A Journal of Practice & Theory 13 (1): 126–136 Bae, G S., S U K Choi, and J H W A Rho 2016 Audit hours and unit audit price of industry specialist auditors: Evidence from Korea Contemporary Accounting Research 33 (1): 314–340 https://doi.org/10.1111/1911-3846.12145 Balsam, S., J Krishnan, and J S Yang 2003 Auditor industry specialization and earnings quality Auditing: A Journal of Practice & Theory 22 (2): 71–97 https://doi.org/10.2308/aud.2003.22.2.71 Barth, M E., W R Landsman, M Lang, and C Williams 2012 Are IFRS-based and US GAAP-based accounting amounts comparable? Journal of Accounting and Economics 54 (1): 68–93 https://doi.org/10.1016/j.jacceco.2012.03.001 Becker, C L., M L DeFond, J Jiambalvo, and K R Subramanyam 1998 The effect of audit quality on earnings management Contemporary Accounting Research 15 (1): 1–24 https://doi.org/10.1111/j.1911-3846.1998.tb00547.x Behn, B K., J Choi, and T Kang 2008 Audit quality and properties of analyst earnings forecasts The Accounting Review 83 (2): 327– 349 https://doi.org/10.2308/accr.2008.83.2.327 Bonner, S E., and B L Lewis 1990 Determinants of auditor expertise Journal of Accounting Research 28 (Supplement): 1–20 https:// doi.org/10.2307/2491243 Brown, H L., and A M Wright 2008 Negotiation research in auditing Accounting Horizons 22 (1): 91–109 https://doi.org/10.2308/ acch.2008.22.1.91 Carey, P., and R Simnett 2006 Audit partner tenure and audit quality The Accounting Review 81 (3): 653–676 https://doi.org/10.2308/ accr.2006.81.3.653 Cascino, S., and J Gassen 2015 What drives the comparability effect of mandatory IFRS adoption? Review of Accounting Studies 20 (1): 242–282 https://doi.org/10.1007/s11142-014-9296-5 Chen, C., C Lin, and Y Lin 2008 Audit partner tenure, audit firm tenure, and discretionary accruals: Does long audit tenure impair earnings quality? Contemporary Accounting Research 25 (2): 415–445 https://doi.org/10.1506/car.25.2.5 Chen, C., D W Collins, T D Kravet, and R Mergenthaler 2018 Financial statement comparability and the efficiency of acquisition decisions Contemporary Accounting Research 35 (1): 164–202 https://doi.org/10.1111/1911-3846.12380 Cheng, Y S., Y P Liu, and C Y Chien 2009 The association between auditor quality and human capital Managerial Auditing Journal 24 (6): 523–541 https://doi.org/10.1108/02686900910966512 Chi, H., and C Chin 2011 Firm versus partner measures of auditor industry expertise and effects on auditor quality Auditing: A Journal of Practice & Theory 30 (2): 201–229 https://doi.org/10.2308/ajpt-50004 Chi, W., H Huang, Y Liao, and H Xie 2009 Mandatory audit partner rotation, audit quality, and market perception: Evidence from Taiwan Contemporary Accounting Research 26 (2): 359–391 https://doi.org/10.1506/car.26.2.2 De Franco, G., S P Kothari, and R Verdi 2011 The benefits of financial statement comparability Journal of Accounting Research 49 (4): 895–931 https://doi.org/10.1111/j.1475-679X.2011.00415.x Eshleman, J D., and P Guo 2014 Do Big auditors provide higher audit quality after controlling for the endogenous choice of auditor? Auditing: A Journal of Practice & Theory 33 (4): 197–219 https://doi.org/10.2308/ajpt-50792 Financial Accounting Standards Board (FASB) 2010 Conceptual Framework for Financial Reporting Statement of Financial Accounting Concepts No Norwalk, CT: FASB Francis, J R., and M D Yu 2009 Big office size and audit quality The Accounting Review 84 (5): 1521–1552 https://doi.org/10.2308/ accr.2009.84.5.1521 Francis, J R., M L Pinnuck, and O Watanabe 2014 Auditor style and financial statement comparability The Accounting Review 89 (2): 605–633 https://doi.org/10.2308/accr-50642 The Accounting Review Volume 95, Number 3, 2020 140 Chen, Chen, Chin, and Lobo Francis, J R., K Reichelt, and D Wang 2005 The pricing of national and city-specific reputations for industry expertise in the U.S audit market The Accounting Review 80 (1): 113–136 https://doi.org/10.2308/accr.2005.80.1.113 Gul, F A., D Wu, and Z Yang 2013 Do individual auditors affect audit quality? Evidence from archival data The Accounting Review 88 (6): 1993–2023 https://doi.org/10.2308/accr-50536 Harris, S B 2016 Auditor independence and PCAOB’s investor-protection Available at: https://corpgov.law.harvard.edu/2016/07/12/ auditor-independence-and-pcaobs-investor-protection/ Jones, J J 1991 Earnings management during import relief investigations Journal of Accounting Research 29 (2): 193–228 https://doi org/10.2307/2491047 Kaplan, S E., E F O’Donnell, and B M Arel 2008 The influence of auditor experience on the persuasiveness of information provided by management The Accounting Review 27 (1): 67–83 https://doi.org/10.2308/aud.2008.27.1.67 Kawada, B 2014 Audit offices and the comparability and quality of clients’ earnings Working paper, San Diego State University Kennedy, J., and M E Peecher 1997 Judging auditors’ technical knowledge Journal of Accounting Research 35 (2): 279–293 https:// doi.org/10.2307/2491366 Kim, J B., L Li, L Y Lu, and Y Yu 2016 Financial statement comparability and expected crash risk Journal of Accounting and Economics 61 (2/3): 294–312 https://doi.org/10.1016/j.jacceco.2015.12.003 Kim, S., P Kraft, and S G Ryan 2013 Financial statement comparability and credit risk Review of Accounting Studies 18 (3): 783–823 https://doi.org/10.1007/s11142-013-9233-z Knechel, W R., A Vanstraelen, and M Zerni 2015 Does the identity of engagement partners matter? An analysis of audit partner reporting decisions Contemporary Accounting Research 32 (4): 1443–1478 https://doi.org/10.1111/1911-3846.12113 Kothari, S., A Leone, and C Wasley 2005 Performance matched discretionary accrual measures Journal of Accounting and Economics 39 (1): 163–197 https://doi.org/10.1016/j.jacceco.2004.11.002 Lennox, C S., X Wu, and T Zhang 2014 Does mandatory rotation of audit partners improve audit quality? The Accounting Review 89 (5): 1775–1803 https://doi.org/10.2308/accr-50800 Li, L., B Qi, G Tian, and G Zhang 2017 The contagion effect of low-quality audits at the level of individual auditors The Accounting Review 92 (1): 137–163 https://doi.org/10.2308/accr-51407 McKnight, C A., and W F Wright 2011 Characteristics of relatively high-performance auditors Auditing: A Journal of Practice & Theory 30 (1): 191–206 https://doi.org/10.2308/aud.2011.30.1.191 Messier, W F., V Owhoso, and C Rakovski 2008 Can audit partners predict subordinates’ ability to detect errors? Journal of Accounting Research 46 (5): 1241–1264 Ministry of Finance of the People’s Republic of China (MOF) 1995a Audit Report Independent Auditing Standards No (In Chinese.) Beijing, China: MOF Ministry of Finance of the People’s Republic of China (MOF) 1995b Annual Report Audit Practice Guidelines of Certified Public Accountants in China No (In Chinese.) Beijing, China: MOF Nelson, M 2009 A model and literature review of professional skepticism in auditing Auditing: A Journal of Practice & Theory 28 (2): 1–34 https://doi.org/10.2308/aud.2009.28.2.1 Nelson, M., and H Tan 2005 Judgment and decision making research in auditing: A task, person, and interpersonal interaction perspective Auditing: A Journal of Practice & Theory 24 (s-1): 41–71 https://doi.org/10.2308/aud.2005.24.s-1.41 Pincus, K V 1990 Auditor individual differences and fairness of presentation judgments Auditing: A Journal of Practice & Theory (3): 150–166 Shelton, S W 1999 The effect of experience on the use of irrelevant evidence in auditor judgment The Accounting Review 74 (2): 217– 224 https://doi.org/10.2308/accr.1999.74.2.217 Shipman, J E., Q T Swanquist, and R L Whited 2017 Propensity score matching in accounting research The Accounting Review 92 (1): 213–244 https://doi.org/10.2308/accr-51449 Su, L., and D Wu 2016 Is audit behavior contagious? Teamwork experience and audit quality by individual auditors Working paper, Lingnan University and The Chinese University of Hong Kong Wallman, S 1996 The future of accounting, Part III: Reliability and auditor independence Accounting Horizons 10: 76–97 Yip, R W Y., and D Young 2012 Does mandatory IFRS adoption improve information comparability? The Accounting Review 87 (5): 1767–1789 https://doi.org/10.2308/accr-50192 Zerni, M 2012 Audit partner specialization and audit fees: Some evidence from Sweden Contemporary Accounting Research 29 (1): 312–340 https://doi.org/10.1111/j.1911-3846.2011.01098.x The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 141 APPENDIX A Variable Definitions Variable Main Variables Tacc_Diff AbnAcc_Diff EComp_Cov SameFirm_DiffOffice_DiffPartner SameOffice_DiffPartner SamePartner SamePartner_Senior SamePartner_Junior SamePartner_Others SamePartner_DiffTeam SameTeam Control Variables Tacc_Min AbnAcc_Min Size_Diff Size_Min Lev_Diff Lev_Min MB_Diff MB_Min CFO_Diff CFO_Min LossProb_Diff LossProb_Min Std_Sales_Diff Std_Sales_Min Std_CFO_Diff Std_CFO_Min Definition The absolute value of the difference in total accruals between two client firms in a pair Total accruals are the difference between income before extraordinary items and cash flows from operations adjusted for cash flows from extraordinary items, scaled by lagged total assets The absolute value of the difference in abnormal accruals between two client firms in a pair Abnormal accruals are the residuals calculated from the cross-sectional Jones (1991) model after controlling for contemporaneous ROA Earnings co-movement between two client firms in a pair It is the R2 from regressing one client firm’s earnings on the other client firm’s earnings in the same quarter over eight consecutive quarters An indicator variable that equals if both client firms in a pair are audited by the same audit firm, but different audit offices and different individual auditors, and otherwise An indicator variable that equals if both client firms in a pair are audited by the same audit office, but different individual auditors, and otherwise An indicator variable that equals if the two client firms in a pair have a common signing auditor, and otherwise We not consider different roles of signing auditors (review versus engagement) in this measure An indicator variable that equals if the two client firms in a pair have a common review auditor, but different engagement auditors, and otherwise An indicator variable that equals if the two client firms in a pair have a common engagement auditor, but different review auditors, and otherwise An indicator variable that equals if SamePartner ¼ 1, but SamePartner_Senior ¼ and SamePartner_Junior ¼ 0, and otherwise An indicator variable that equals if the two client firms in a pair have only one common signing auditor, and otherwise An indicator variable that equals if the two client firms in a pair have two common signing auditors, and otherwise Minimum value of total accruals between two client firms in a pair Minimum value of abnormal accruals between two client firms in a pair The absolute value of the difference in natural logarithm of total assets between two client firms in a pair Minimum value of natural logarithm of total assets between two client firms in a pair The absolute value of the difference in debt-to-assets ratio between two client firms in a pair Minimum value of debt-to-assets ratio between two client firms in a pair The absolute value of the difference in market-to-book ratio between two client firms in a pair Market-to-book ratio is market value of equity divided by book value of equity Minimum value of market-to-book ratio between two client firms in a pair The absolute value of the difference in cash flows from operations scaled by lagged total assets between two client firms in a pair Minimum value of cash flows from operations scaled by lagged total assets between two client firms in a pair The absolute value of the difference in loss probability between two client firms in a pair Loss probability is the proportion of loss quarters (i.e., income before extraordinary items is negative) in the past eight quarters Minimum value of loss probability between two client firms in a pair The absolute value of the difference in standard deviation of quarterly sales over the past eight quarters between two client firms in a pair Minimum value of standard deviation of quarterly sales over the past eight quarters between two client firms in a pair The absolute value of the difference in standard deviation of quarterly operating cash flows over the past eight quarters between two client firms in a pair Minimum value of standard deviation of quarterly operating cash flows over the past eight quarters between two client firms in a pair (continued on next page) The Accounting Review Volume 95, Number 3, 2020 Chen, Chen, Chin, and Lobo 142 APPENDIX A (continued) Variable Std_SalesGr_Diff Std_SalesGr_Min CFOComp_Cov RetComp_Cov Tenure_Diff Tenure_Min IndustrySpec Variables in Additional Tests 985University Graduate_Degree Female Education_Cohort CurrentJob_Exp PriorTop10_Exp Rank Change_DifferentPartner PostChange_DifferentPartner Change_SamePartner PostChange_SamePartner CompAcct FError_Avg Dispersion_Avg Definition The absolute value of the difference in standard deviation of quarterly sales growth over the past eight quarters between two client firms in a pair Sales growth is the change in total sales in year t divided by total sales in year tÀ1 Minimum value of standard deviation of quarterly sales growth over the past eight quarters between two client firms in a pair Operating cash flows co-movement between two client firms in a pair It is the R2 from regressing one client firm’s operating cash flows on the other client firm’s operating cash flows in the same quarter over eight consecutive quarters Stock return co-movement between two client firms in a pair It is the R2 from regressing one client firm’s monthly stock return on the other client firm’s monthly stock return over eight consecutive quarters The absolute value of the difference in signing auditor tenure between two client firms in a pair For each client firm, we calculate its signing auditor tenure as the natural logarithm of the average number of years that its two signing auditors (review auditor and engagement auditor) each have worked on the engagement Minimum value of signing auditor tenure between two client firms in a pair An indicator variable that equals if the two client firms in a pair have at least one signing auditor who is an industry specialist, and otherwise We calculate each individual auditor’s market share in each industry-year as the proportion of total sales audited by that auditor to total sales of all firms in that industry-year We define the top three individual auditors who have the largest market share as industry specialists An indicator variable that equals if the common signing auditor graduated from a Project 985 university in China (i.e., a top Chinese university), and otherwise An indicator variable that equals if the common signing auditor has an advanced education degree, and otherwise An indicator variable that equals if the common signing auditor is a female, and otherwise An indicator variable that equals if the common signing auditor was born in or after 1971, and otherwise An indicator variable that equals if the number of years that the common signing auditor has worked for the current audit firm employer is above the sample median, and otherwise An indicator variable that equals if the common signing auditor has prior work experience in a Top Ten audit firm, and otherwise An indicator variable that equals if the common signing auditor is a partner, and otherwise An indicator variable that equals for firm-pairs in the subsample of mandatory rotation from having the same signing auditor to having different signing auditors, and otherwise An indicator variable that equals in the years after the mandatory rotation from having the same signing auditor to having different signing auditors for the two client firms in a pair, and otherwise An indicator variable that equals for firm-pairs in the subsample of mandatory rotation from having different signing auditors to having the same signing auditor, and otherwise An indicator variable that equals in the years after the mandatory rotation from having different signing auditors to having the same signing auditor for the two client firms in a pair, and otherwise À1 multiplied by the mean absolute value of the difference between the predicted earnings of firm i and firm j, given their respective accounting functions and firm i’s economic events (quarterly return), over the past eight quarters Average value of analyst earnings forecast errors for two client firms in a pair Analyst earnings forecast error is the absolute value of the difference between actual earnings before extraordinary items and the mean of all analysts’ last forecasts over the period starting three days after last year’s earnings announcement and ending three days before the current year’s earnings announcement, scaled by stock price at the beginning of the year Average dispersion in analyst earnings forecasts for two client firms in a pair Dispersion in analyst earnings forecasts is the standard deviation of all analysts’ last forecasts over the period starting three days after the last year’s earnings announcement and ending three days before the current year’s earnings announcement, scaled by stock price at the beginning of the year (continued on next page) The Accounting Review Volume 95, Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 143 APPENDIX A (continued) Variable Numest_Diff Numest_Min AvgHorizon_Diff AvgHorizon_Min FError_Diff FError_Min Dispersion_Diff Dispersion_Min StdROE_Diff StdROE_Min EPS_Diff EPS_Min The Accounting Review Volume 95, Number 3, 2020 Definition The absolute value of the difference in natural logarithm of the number of analysts following each firm in a client firm-pair Minimum value of natural logarithm of the number of analysts following each firm for a client firm-pair The absolute value of the difference in forecast horizon between two client firms in a pair Forecast horizon is calculated as the natural logarithm of the average number of calendar days between analyst earnings forecast release dates and actual earnings announcement date Minimum value of forecast horizon between two client firms in a pair The absolute value of the difference in analyst earnings forecast errors between two client firms in a pair Minimum value of analyst earnings forecast errors between two client firms in a pair The absolute value of the difference in dispersion across analyst earnings forecasts between two client firms in a pair Minimum value of dispersion across analyst earnings forecasts between two client firms in a pair The absolute value of the difference in standard deviation of return on equity over the last five years between two client firms in a pair Minimum value of standard deviation of return on equity over the last five years between two client firms in a pair The absolute value of the difference in earnings per share scaled by stock price at the beginning of the year between two client firms in a pair Minimum value of earnings per share scaled by stock price at the beginning of the year between two client firms in a pair Copyright of Accounting Review 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 ... Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 127 TABLE Relation between Individual Auditor Style and Earnings Comparability (H1) Panel A: Regression... 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 141 APPENDIX A Variable Definitions Variable Main Variables Tacc_Diff AbnAcc_Diff EComp_Cov SameFirm_DiffOffice_DiffPartner... Number 3, 2020 Do Firms That Have a Common Signing Auditor Exhibit Higher Earnings Comparability? 139 the engagement team can affect client firms? ?? earnings comparability, but the individual effect