Comments by the Auditing Standards Committee of the Auditing Section of the American Accounting Association on PCAOB Release No. 2015-004, Supplemental Request for Comment: Rules to Require Disclosure of Certain Audit Participants on a New PCAOB Form: Participating Committee Members: Free

The current Audit Standards Committee of the Auditing Section of the American Accounting Association (“the Committee”) shares the perspective expressed in the prior committee's response to PCAOB Release 2013-009 (Anderson, Gaynor, Hackenbrack, Lisic, and Wu 2014). Specifically, we:

• 1. 

  commend the PCAOB (“the Board”) for maintaining the focus on “transparency” rather than “accountability” as originally framed in the 2009 Concept Release (PCAOB 2009b);¹

• 2. 

  believe firm disclosure of the name of the engagement partner will be of limited use to investors, and may be potentially harmful, when making investment decisions sans extraordinary circumstances, both initially and over time;² and

• 3. 

  believe firm disclosure of the names, locations, and extent of participation of other participants has a far greater potential to be investor decision relevant and informative to current and future audit committees than the disclosure of the name of the lead engagement partner. See the Committee's response to Question 7.

We wish to emphasize three points:

• 1. 

  Should the Board choose to disclose the engagement partner on a new PCAOB Form AP, Auditor Reporting of Certain Audit Participants, we believe the Board should require disclosure of the concurring partner as well. See the Committee's response to Question 2.

• 2. 

  Form AP should be developed to ensure the disclosures are captured in a consistent manner over time. See the Committee's response to Question 5.

• 3. 

  We believe the Board should carefully and deliberately consider Professor Kinney's discussion of Knechel, Vanstraelen, and Zerni (2015), and its relevance to the body of literature that the Board has used to support firm disclosure of the engagement partner. Professor Kinney highlights several issues that significantly limit the external validity and the generalizability of the findings obtained in non-U.S. jurisdictions to a U.S. setting (Kinney 2015). See the Committee's response to Question 11, new research.

Comments or suggestions for the Board's consideration follow, organized by the questions posed in the Supplemental Request for Comment.

• Question 1. 

  Would disclosure on Form AP as described in this release achieve the same potential benefits of transparency and an increased sense of accountability as mandatory disclosure in the auditor's report? How do they compare? Would providing the disclosures on Form AP change how investors or other users would use the information?

There is no directly relevant research that we are aware of, in the U.S. or abroad, that examines the effect of the disclosure method for engagement partner names or other audit participants. While prior research finds that the form of disclosure affects readers' use of financial statement disclosures (Frederickson, Hodge, and Pratt 2006; Johnson 1992; Libby and Brown 2013; Yu 2013), we feel it is a stretch to apply findings in that literature to the disclosures considered in Release 2015-004 as reactions to the form of disclosure for previously unreported information is likely fundamentally different from reactions to the form of disclosure for an evolving financial reporting standard.

That said, providing the name of the engagement partner and information pertaining to other participants on PCAOB Form AP rather than in the auditor's report is more consistent with the oversight role the Board. Disclosure on Form AP provides the desired information to interested parties without potentially burdening firms and audit participants with unnecessary legal and regulatory responsibilities regarding issues of consent to use the audit report.

• Question 2. 

  Are there special considerations relating to the Form AP approach that have not been addressed in this supplemental request for comment? If so, what are the considerations? How might the Board address them? What are the costs of Form AP compared to the costs of disclosure in the auditor's report?

If the PCAOB determines that disclosure of the audit engagement partner increases transparency and improves audit quality (see discussion in Appendix 2 of Release 2015-004), then the Board should also consider the consistency of this proposed disclosure with existing SEC rules and auditing standards. Specifically, Section 203 of the Sarbanes-Oxley Act, along with the final rules adopted by the Securities and Exchange Commission ([SEC] 2003), establish independence and rotation requirements not only for the lead engagement partner, but also for concurring partners on audits of SEC registrants. The motivation behind these regulations is that both lead and concurring partners have significant influence over the audit and the quality of the work performed and conclusions reached. Moreover, disclosure of concurring partners (“engagement quality review partners”) would further emphasize the importance of these partners in the audit process as discussed in Auditing Standard No. 7, Engagement Quality Review (PCAOB 2009a).

As a result, requiring disclosure of the name of the concurring partner in Part III of Form AP, along with the lead engagement partner, would seem to be consistent with the rules and standards already in place for these individuals. We also emphasize that isolating the lead engagement partner without also naming other key partners, of which several could exist on the largest engagements, can send the wrong signal to investors about the responsibility for and coordination of the audit.

• Question 4. 

  In addition to the required filing of the Form AP, auditors may decide to voluntarily provide the same disclosures in the auditor's report. Are there any special considerations or unintended consequences regarding voluntary disclosure in the auditor's report? If so, what are those considerations or consequences? How might the Board address them?

Drawing from voluntary disclosure literature (Beyer, Cohen, Lys, and Walther 2010), accounting firms would only choose to voluntarily disclose information in the audit report if disclosure is sufficiently favorable given the additional disclosure costs. Because many of the costs of audit report disclosure noted in the proposal accrue to the auditors and their client firms (e.g., legal liability, need for consents), we do not believe many auditors, if any, will voluntarily disclose additional information in the audit report. Given that quality controls are managed at the firm level, public accounting firms would be expected to adopt a common disclosure strategy for all its audit engagements.

A complicating factor in this decision involves the potential effects on the client firm if investors react to such a disclosure. Moreover, changes in audit outcomes over time for a particular engagement could provide incentives to voluntarily disclose the information in the audit report in one year but not in the next year, which could create uncertainties for investors and other parties. Overall, it seems unlikely that firms will use the voluntary disclosure option, and if they do, they may do so strategically. Therefore, allowing for voluntary disclosure in the audit report may be counter-productive to the Board's aims.

• Question 5. 

  What search criteria and functionality would users want for information filed on Form AP? What additional criteria and functionality beyond what is described in Section IV of this release would be useful? Would third-party vendors provide additional functionality if the Board does not? Are there cost-effective ways to make the disclosure more broadly accessible to investors who may not be familiar with PCAOB forms?

For investors and other parties to efficiently use this data, the names of engagement partners and other participants should be input in a consistent manner. For example, Appendix 1 illustrates that the following information should be disclosed for the audit engagement partner: “Name (that is, first and last name and any middle name[s] and suffix) of the engagement partner on the current period's audit.” Based on this information, we assume the same engagement partner's name could appear as John R. Smith in one year and John Robert Smith, Jr. in another year if a check is not in place to retain a consistent format. This possibility creates a “many-to-many” relationship in any database constructed from this information (i.e., multiple parties could have identical names, and one party could be reported with multiple name variants). The PCAOB should consider adding a question in Form AP asking whether the audit engagement partner had signed a report in the past, and if so, select the individual's name from a pre-existing database list. This method would maintain consistency in disclosures over time. A similar approach could be taken for other participants in the audit.

While this approach should alleviate the many-to-many relation, it retains the possibility of a “one-to-many” relation in that two partners could still have identical names. Another option, albeit a more costly one in terms of administration for the PCAOB, would be to assign partners and other entities a unique identifying number. Unique identifiers reduce the partner-to-name relation to a “one-to-one” relation, which is not only a best practice for database design, but has the added benefit of allowing users to unambiguously identify partners and other participants.

Section IV states that “over time … the PCAOB could allow users to download the search results” (PCAOB 2015, 9). If the primary goal is to increase transparency and allow comparison with audit outcomes, then we encourage the PCAOB to provide download capabilities of Form AP data from the outset. Limiting this functionality to a later date would delay larger scale analyses of these disclosures by investors, academics, and other market participants due to the need for hand collection or collection by third party vendors.

• Question 7. 

  This supplemental request for comment contemplates not requiring disclosure of non-accounting firm participants in the audit as previously proposed. Is it an appropriate approach to not require disclosure of non-accounting firm audit participants? If not, should the Board adopt the requirements as proposed in the 2013 Release or the narrower, more tailored approach described in Section V of this supplemental request, which would not require disclosure of information about non-accounting firm participants controlled by or under common control with the accounting firm issuing the auditor's report, with control as defined in Section V? If the Board were to adopt this narrower, more tailored approach, then is the description of the scope of a potential requirement sufficiently clear? Why or why not? Is the definition of control in Section V appropriate? Why or why not?

The Committee believes the disclosure of non-accounting firm participants, particularly when combined with an indication of the amount of effort they contribute to the audit, will provide useful insight into the audit process. Given non-accounting firm participants are likely to take part in a number of different audit engagements and potentially be used across audit firms, the conclusions that could be drawn regarding reputation would be potentially less misleading than what is inferred from disclosures about a single engagement partner who would be involved in a limited set of engagements over a couple of years or even over a career.

We share the perspective expressed by the prior committee and support the requirements as proposed in the 2013 Release (Anderson et al. 2014). Again, we believe the disclosure of non-accounting firm participants is ultimately more informative than disclosures associated with the lead engagement partner.

• Question 11. 

  Are there additional economic considerations associated with mandated disclosure, either in the auditor's report or on Form AP, that the Board should consider? If so, what are those considerations? The Board is particularly interested in hearing from academics and in receiving any available empirical data commenters can provide.

The Board documents in Appendix 2 of Release 2015-004 that recent research in non-U.S. markets presents mixed evidence on the usefulness of firm disclosure of the engagement partner (e.g., Carcello and Li 2013; Blay, Notbohm, Schelleman, and Valencia 2014; Aobdia, Lin, and Petacchi 2015; Knechel et al. 2015).

The Committee believes the Board should carefully and deliberately consider Professor Kinney's discussion of Knechel et al. (2015), and its relevance to the body of literature the Board has used to support firm disclosure of the engagement partner. Professor Kinney highlights several issues that significantly limit the external validity and the generalizability of the findings obtained in non-U.S. jurisdictions to a U.S. setting (Kinney 2015). Professor's Kinney's comments are necessarily focused on Knechel et al. (2015), but are generally relevant to this body of literature. First, Knechel et al.'s (2015) primary findings are based on private company statutory audits in Sweden, which comprise 99.2 percent of the Swedish audit population and 95 percent of the study's sample (100 percent of the study's going concern sample). Given that Swedish audit partners sign an average of 80.3 audit reports per year, the size and risk of these engagements are much different than publicly traded clients of U.S. audit partners. Second, Sweden did not change its mandatory partner disclosure requirements during the study's sample period, which limits the researchers' ability to test whether mandating disclosure of audit partner identities has a causal effect on auditor behavior, or is associated with reactions from other market participants. These two factors suggest the research findings may be unique to the Swedish audit environment and may not generalize to the U.S. context.

Kinney (2015) also points out that large accounting firms may use an engagement partner assignment strategy such as “best partner-to-riskiest engagements.” If this type of strategy occurs in practice, then the study's findings would have the opposite interpretation since high quality partners would be intentionally assigned to high risk audits. As a result, public disclosure of engagement partners could have two negative consequences:

The Committee believes the Board should be cognizant of these, and other, limitations when using research from non-U.S. jurisdictions to inform the development of U.S. policies.

Rather than reiterating the previous committee's comments (Anderson et al. 2014) or describing research the Board cited in Appendix 2 Release 2015-004, we considered research published or made available since the prior committee's comment. We noted in our response to Question 1 that we are not aware of research that speaks directly to the form of disclosure for entities involved in the audit. However, we identified a number of studies that speak to the potential economic impact of disclosing the identities of those involved with the audit that were not incorporated in Anderson et al. (2014) or referenced in Appendix 2 of Release 2015-004. One caveat—these studies should be considered in light of Kinney (2015).

Several new studies provide empirical results that speak to the usefulness of disclosing engagement partner information. Using market data from China, Wang, Yu, and Zhao (2014) find that an audit partner's past audit failure rate is positively associated with future restatements by the partner's clients, and the association is stronger for engagement partners than reviewing partners.³ They also find that quality control measures at both the firm and engagement level fail to consistently attenuate this association. Using data from China, Cahan and Sun (2015) find that audit partner experience is negatively associated with discretionary accruals, and positively associated with audit fees. Blay et al. (2014) find that newly mandated audit partner signatures in the Netherlands did not change audit quality, as measured by levels of discretionary accruals and clients meeting or beating earnings forecasts. Ittonen, Johnstone, and Myllymäki (2015) examine data from Finland and find that audit partners with greater public-client experience are associated with lower abnormal accruals. They also find that greater public-client specialization is more important when the audit partner has lower overall audit experience. While these studies speak directly to the potential usefulness of identifying engagement partners, their prior audit failure rate, and experience, they may not generalize to the U.S. market due to the differing baseline conditions pointed out by the Board in Release 2015-004 as well as Kinney (2015).

A study by Saito and Takeda (2014) speaks to the issue of identifying other entities involved in the audit. Saito and Takeda (2014) analyzed a specific audit failure by the foreign-affiliate (ChuoAoyama) of a U.S. firm (PricewaterhouseCoopers). They find that the foreign-affiliate's failure damaged the reputation of PwC as well as other Big 4 firms with global networks as measured by stock price premiums. Their finding implies that disclosure of other entities with significant involvement in an audit may be value-relevant for investors. For a more thorough consideration of the disclosure of other entities, see the Committee's response to Question 7.

A critical component of the economic impact of the Board's proposal is the usefulness of the disclosure(s) to audit report and financial statement users. As noted in the prior committee's comment (Anderson et al. 2014), the development of a robust database on audit participants could be beneficial for investors, academics, and other financial statement users. The prior committee's commentary notes that “metrics beyond the name of the engagement partner are needed to make … consequential decisions” (Anderson et al. 2014, C2). Similarly, the Board's request for comment identifies a number of metrics that may be useful, specifically the number of other public company, broker/dealer audits conducted by the engagement partner, years of industry-specific audit experience, tenure as the engagement partner on the audit, the number and nature of restatements the partner is associated with (as the engagement partner), and information regarding any disciplinary procedures.

The academic literature supports the potential usefulness of some of these metrics and their underlying constructs. For example, industry specialization and expertise has repeatedly been found to enhance audit quality (e.g., S. Wright and A. Wright 1997; Taylor 2000; Balsam, Krishnan, and Yang 2003; Krishnan 2003; Payne 2008; Kim, Lee, and Park 2015).⁴ Partner tenure might be informative; much of the academic literature on audit tenure suggests lower audit quality in the initial years of a firm/client relationship (Geiger and Raghunandan 2002; Johnson, Khurana, and Reynolds 2002; J. Myers, L. Myers, and Omer 2003; Carcello and Nagy 2004; Knechel and Vanstraelen 2007; Jackson, Moldrich, and Roebuck 2008; Davis, Soo, and Trompeter 2009). Further, the PCAOB and/or SEC are primary sources for many of these metrics, such as information about restatements and disciplinary actions. To the extent it is feasible, the Board may wish to consider linking its existing, non-confidential data to individual partners in the proposed database.

The Board and individual Board members repeatedly note that certain information in the database will be useful to investors and other interested parties as time passes (PCAOB 2015; Ferguson 2015; Hanson 2015). Board Member Ferguson states:

In other words, it appears the Board expects Form AP data to become more meaningful as audit partners, reporting companies, and other named participants develop a reporting history. These statements recognize an inherent problem with data sources that begin at one point in time, a problem that academics are intimately familiar with—data truncation. The data truncation problem has the potential to limit the usefulness, and thus the economic benefit to users of the Form AP data in the early years of its use.

The Committee suggests that the Board consider the cost-benefit trade-offs associated with steps to alleviate the data truncation problem. For example, the Board could request additional background information pertaining to partners and other audit participants when they are first included in a Form AP filing. This background information should be limited and restricted to metrics that are reasonably available and for which empirical evidence of usefulness exists (such as those noted in the previous subsection, “Additional Form AP Metrics”). As also noted in that subsection, the Board could link its existing, non-confidential information (such as public disciplinary proceedings) with the Form AP data, thereby alleviating part of the burden on filers while addressing the truncation issue. In addition, the Board could consider requesting information pertaining to the incumbent audit firm's previous audits of the registrant for a designated number of years (e.g., three years or five years) in the initial Form AP filing. This “historical” information could be requested only if the current audit firm was the company's main auditor, defined as the signing audit firm and not just listed as a participant firm in prior years. In order to protect audit firms that may need to retrospectively estimate the participation of other audit firms in these earlier periods, the Board could consider adopting “good faith” safe harbor rules for this “historical” audit information that would be included on the initial Form AP filing. The Committee recognizes that requiring background information on partners and audit participant firms, as well as prior audit information in the initial Form AP adoption, will increase the initial costs of gathering data and preparing and filing Form AP. The cost would be a one-time cost for any single partner or named participant. On the other hand, requiring background information increases the potential immediate benefits gained by users of Form AP information. By adopting these suggestions, Form AP would be more useful, more quickly, but initially more costly to prepare.