DATE: Aug. 13, 2013
SPEAKER: Lewis H. Ferguson, Board Member
EVENT: PCAOB Open Board Meeting
LOCATION: Washington, DC
The proposed standards represent the culmination of several years' work by the Public Company Accounting Oversight Board and its staff to reconsider the form and content of the standard auditor's report. I support issuance of these standards and want to express my gratitude for the tireless and careful work of the PCAOB staff in putting together the proposal and supporting documents. Preparing this proposal has involved weighing difficult and complex issues and making choices among considerations that are sometimes competing. I agree with the choices represented by the proposed standards but I want to use this statement to set forth some of my own considerations in reaching the decision to support the proposal and some of the questions that remain in my mind.
The current form of the audit report is binary, essentially limited to an opinion that the financial statements being audited do or do not fairly present the financial condition of the audited entity, and has been largely unchanged since the 1940s. Prior to that time, auditor's reports varied widely in form and content, making comparisons difficult. The current form of the report was developed to increase uniformity and, consequently, comparability of audit reports, a goal that the revised and current form of the auditor's report achieved admirably but at the cost of communicating little useful information about what the auditor actually does or learns in the course of an audit. In the decades since 1940, several commissions (including the Cohen and Treadway commissions, referred to in the Board's release) considered and recommended changes to the form of the auditor's report, but only minor changes were ever made. Under these circumstances, it is not surprising that commentators have long spoken of an "expectations gap" between what investors and other users of financial statements expect to learn from auditors and what the standard audit report delivers.
Before undertaking this project, the Board sought the views of its Investor Advisory Group, and Standing Advisory Group, and conducted a public roundtable on the issue. The Board heard views from investors, consumer advocates, issuers, audit committee members, auditors, securities lawyers, regulators and standard setters, among others. The Board also issued a Concept Release on June 21, 2011 seeking comment on a variety of potential changes to the audit report, such as inclusion in the audit report of (1) an auditor's discussion and analysis ("AD&A"), (2) required and expanded emphasis paragraphs in which the auditor would point out certain critical matters in the audit and indicate where such matters were reflected in the financial statements, (3) auditor assurance on other information outside the financial statements and notes and (4) clarification of certain definitional provisions such as the meaning of "reasonable assurance" or the auditor's responsibility for the detection of fraud. Commenters were obviously also free to say that the current audit report did not need to be changed. The Board received 155 comments from interested parties that supported various approaches, including a number that argued that the current auditor's report is adequate and does not need to be changed.
The Release on these standards and its accompanying appendices includes citations to the PCAOB website where the results of these outreach efforts and the comment letters can be seen. The Board has also followed with interest the efforts of other regulators and standard setters, such as the International Auditing and Assurance Standards Board and the U.K.'s Financial Reporting Council, to address the auditor's reporting model.
These outreach efforts revealed several things. There was widespread agreement that the present binary form of the auditor's reporting model is useful but in a limited way. Many financial statement users indicated that they merely glance at the auditor's report to make sure that the opinion is unqualified and then move on to other more useful information when analyzing an entity's financial statements. Many financial statement users also expressed a belief that auditors likely develop a deeper understanding of the enterprises they audit than anyone other than company management and that the insights from such understanding would be useful if shared with the financial statement users. The auditor's views are seen as particularly useful by many financial statement users because the auditors are perceived to be not only deeply informed and expert observers, but also, and perhaps most importantly, independent of company management.
Not surprisingly given these perceptions, many financial statement users, particularly investors, favored an AD&A that would be an analog of the management discussion and analysis ("MD&A") currently required in certain filings under the federal securities laws. Supporters of this approach claimed that it would give them a more detailed view both of the auditor's work and the auditor's impressions and concerns about the entity being audited, such as the quality of the entity's accounting practices and internal controls, as well as areas of particular audit and financial risk, sometimes expressed as giving a window into "what kept the auditor up at night". These commenters argued that an AD&A would be the best vehicle for conveying to users of financial statements the deep understanding of the audited entity that they perceive the auditors to have.
Many issuers, preparers, and auditors expressed concern that an AD&A would usurp management's primary role in financial reporting, could raise difficult quality assurance issues in audit firms,[1] would destroy the uniformity and therefore the comparability of audit reports (a strength of the current model), and could lead to inconsistent disclosures between management and the auditor ("dueling disclosures"), either forcing management to adopt the auditor's views or, if inconsistent disclosures persisted, leading to confusion in the marketplace. These critics contended that the role of the auditor should be confined to attesting to the fair presentation of the financial disclosures, not expressing opinions on other matters such as the business prospects or risks of the company since such matters are beyond the expertise of the auditor. Comments on such matters should be left to management or, if by outsiders, to securities analysts and investment professionals.
Another view expressed is that auditors already have robust and detailed communications with audit committee members and there is no need for auditors to provide more information to other financial statement users. This view implies that audit committees are the best representatives of the interests of shareholders and should be relied on to oversee the auditor and protect the interests of those shareholders. To the extent a problem exists in this area, in this view, it should be dealt with through reforms to audit committee governance and practice, an area over which this Board has no jurisdiction.
Commenters who favored the use of emphasis paragraphs pointed out that this approach could satisfy the goal of giving users of financial statements a better picture of what the areas of the auditor's concern had been (what "kept the auditor up at night"), but that confining such information to the auditor's work in connection with the audit itself would minimize or even eliminate the risk that such information would conflict with management's financial disclosures and the MD&A. Advocates of this approach also pointed out that the auditor's comments would be limited to matters particularly within the auditor's competence and expertise. Skeptics of the approach have pointed out that auditors already have the power to add emphasis paragraphs and that they are only infrequently employed. Another concern expressed was that given the legitimate concern of audit firms' to assure uniformity and achieve consistent quality control, emphasis paragraphs, even if regularly employed, are likely to degenerate into meaningless boilerplate that does not convey useful information to users of financial statements.
Some commenters submitted that the auditor should provide assurance on "other information", matters outside the financial statements and notes such as non-GAAP measures or quantitative disclosures in the MD&A. Suggestions ranged from levels of assurance that would require the auditor to audit such information to statements that the auditor had read the other information and found nothing materially inconsistent with disclosures in the financial statements and notes, and, to the auditor's knowledge, no material misstatements of fact.
Some commenters would limit any changes in the auditor's report to enhancements to certain definitional provisions such as the meaning of "reasonable assurance" and improved descriptions of the auditor's responsibility to detect fraud. Finally, some commenters expressed the view that the auditor's report already achieves its purpose and does not require changes.
Faced with a range of views, as a first step, the Board has had to grapple with the question whether the auditor in fact possesses information that investors would find truly useful and, if so, how that can be conveyed in a way that does not up-end the financial reporting structure set forth in the federal securities laws. In other words, can the "expectations gap" complained of by financial statement users in connection with the audit report be addressed in a way that is useful and not too disruptive of existing practice?
My view on this question rests on three premises: (1) that investors and other financial statement users, in general, desire more rather than less information about those financial statements, so long as the information is accurate and relevant to their potential use of it, and they believe that auditors possess such relevant information; (2) that the efficiency of financial and other markets is directly correlated with the quality and quantity of information available to the markets; and (3) that auditors, in the ordinary course of their audit work, do gain some information that, if made more widely available, would be useful to investors and financial statement users, but that the auditor's knowledge of the entities they audit is both limited and specialized. It is colored by the particular lens through which the auditor views the financial statements, namely, the need to focus on tasks and facts that will enable the auditor to express an opinion on the fair presentation of the financial statements.
Thus, areas that posed particular difficulties in the audit are matters that financial statement users would likely find to be of interest. Such issues faced in the audit could include, among others, management estimates, the valuation of securities that have no readily ascertainable market value, impairments, contingencies such as possible exposures in lawsuits, or government or regulatory investigations, warranty claims and similar items, exposure to loan losses, revenue recognition practices, and going concern issues. If, at the end of the day, the auditor is able to issue a clean opinion, it is implicit that the auditor was ultimately satisfied that the financial statements, including the troublesome items, were fairly stated. Nonetheless, investors would benefit from knowing those areas of the audit on which the auditor concentrated and where further information on them can be found in the financial statements and notes. That is not information that is currently made publicly available by the auditor.
I have not been persuaded that, however much certain financial statement users might like to know the candid views of auditors about the entities they audit, those auditors are particularly qualified, either by training or by the work they do in the audit, to comment on matters beyond the audit, or to even to know what may be most important in the financial statements to financial statement users. Even less are auditors likely to be qualified to comment on business strategies, industry trends, personnel decisions and succession planning, or other matters that are the daily concern of management. I also believe that the quality control and comparability concerns about a free ranging auditor's commentary that have been widely expressed in comments to the Board are serious and well founded. As a consequence, I cannot support an AD&A approach which seems burdened with a host of potential unintended adverse consequences.
In the proposal, the Board has endeavored to craft an approach to the auditor's report that acknowledges the desire of investors for more information from auditors but also recognizes the limits of the auditor's expertise, the risks of "dueling disclosures" and the appropriate role of management as the main purveyor of information about a company's financial condition and business prospects. The Board's approach is designed to thread a path between satisfying what I believe to be the legitimate desire of users for more information from auditors and preserving the well-established allocation of reporting responsibilities between management and auditors. The Board's approach is designed to limit the auditor's reporting obligations to areas of particular auditor competence and expertise, namely the audit itself.
The Board's proposal on the auditor's report does three major things: (1) it makes changes to the auditor's report by requiring the inclusion of a section to discuss certain critical audit matters ("CAMs"); and (2) it requires disclosure in the audit report of the auditor's continuous tenure as auditor of the issuer; and (3) it requires a description of the auditor's responsibility to read and evaluate other information in an annual report and to report on the results of that evaluation. The accompanying standard on "Other Information" requires the auditor to read and evaluate the other information included in annual reports filed with the Securities and Exchange Commission under the Securities Act of 1934, and to determine whether such other information, based on the evidence obtained and conclusions reached during the audit, is materially inconsistent with information in the audited financial statements and notes thereto, or contains any materially misleading statement of fact. The proposed standard would then also require the auditor to include a section in the audit report describing the nature of this review and the conclusions reached as a result.
The Board's proposal requires that the auditor provide additional information in the audit report on those matters that the auditor has considered to be CAMs. CAMs are defined as "Those matters the auditor addressed during the audit of the financial statements that (1) involved the most difficult, subjective, or complex auditor judgments; (2) posed the most difficulty to the auditor in obtaining sufficient appropriate evidence; or (3) posed the most difficulty to the auditor in forming an opinion on the financial statements."
The proposal makes clear that the term "most" is not intended to limit disclosures to a single item in each category. The proposal also makes clear that in the normal case, at least some matters should be disclosed as CAMs and that only in cases involving a company without complex operations will there be no matters listed and discussed as CAMs. The proposal is not prescriptive, but rather sets forth a variety of factors that when present, either singly or together, should lead a reasonable and experienced auditor to conclude that a matter is a "critical audit matter". Factors such as the difficulty of obtaining adequate audit evidence, matters that required outside consultation, and a variety of others are factors are indicators that a matter might be a CAM. From this list, particularly considering those matters reviewed by the EQR, contained in the engagement completion memo and communicated to the Audit Committee, the auditor must choose those items that qualify as CAMs and also must document the reason for that selection. If a matter would appear to a reasonable and experienced auditor to be a CAM, but the auditor nonetheless determines not to include that matter as one of the disclosed CAMs, the standard includes a documentation requirement that the auditor must document the reasons he or she concluded not to include it in the auditor's report as a CAM.
I support the Board's approach because I believe it allows auditors substantial latitude to exercise professional judgment in determining what items to include as CAMs but constrains that judgment by an objective standard, the reasonable and experienced auditor looking at the same evidence. This standard is analogous to the reasonable man standard that is well developed in many areas of law to provide an objective standard against which to measure individual conduct. Such a standard is also flexible in that it can be expected to evolve over time because it is drawn from the state of knowledge, custom and professional practice at any given time, all of which continuously evolve. It provides a standard that audit firms, regulators like the PCAOB, and others can use to measure compliance with the requirements of the standard. Obviously the proposed standard can only work and be effective if auditors in fact are able to identify CAMs in a consistent way using the criteria set forth in the standard and their professional judgment. The Board looks forward to comments on this issue including, if needed, ways to improve the process for identification of CAMs.
Another important question is whether the drive of audit firms, particularly in larger audit firms, to assure uniformity in practice and quality control, will quickly lead language included in the audit report relating to CAMs to be reduced to boilerplate that communicates little or no useful information. The Board's proposal attempts to avoid that problem by tying the disclosures to specific aspects of audits that the auditor found particularly problematic and will refer to the relevant financial statements and disclosures, where applicable. The auditor's disclosures will not involve independent commentary on management's disclosures but rather will point the user to those areas of the audit that were of particular importance or difficulty to the auditor, giving the user a guide to areas for further inquiry. This is intended to provide users with information that they will find helpful in understanding financial statements or particular issuers, but will hopefully avoid the troublesome problems that inconsistent disclosures by auditors and managements would pose. The release asks a number of questions about the Board's approach and I hope that commenters will particularly address whether this approach is likely to provide financial statement users with useful information and whether the Board's efforts to prevent boilerplate disclosures is likely to be successful. Indeed, the Board will also need to learn if there is a possibility that the definition of CAMs in the proposed standard could lead to an over-inclusive provision of CAMs that would not be restricted to matters that are of real significance to investors and financial statement users. Over inclusion, by flooding financial statement users with information, part of which could be of dubious value, would also be undesirable. If the approach is ultimately adopted, experience will quickly answer these questions, but views on whether the proposed approach contains design flaws would be particularly useful.
The release also contains a discussion of the potential costs and benefits of this approach. Perhaps most importantly, it should be noted that the standard does not require the auditor to do any additional audit work. It does require additional work in classifying and disclosing the work that the auditor has already done, but with practice, that should simply become part of the auditor's analysis of the audit results and should not impose significant additional cost burdens. Assuming that the additional disclosures convey useful information to financial statement users, there will be a demonstrable benefit. It is difficult to quantify these costs and benefits in advance and experience will cast better light on both the costs and benefits of the Board's approach. If the disclosure of CAMs does indeed provide useful information to financial statement users, whatever additional costs are in fact incurred would seem to be well worthwhile, but this is another area where comments will be welcome.
The Board's proposal would require the auditor to disclose its continuous tenure as auditor of the company being audited and reported on. The proposal also contains guidance concerning how to deal with situations where either the auditor or the audited entity has changed through merger, disposition, change of name or other occurrence that would indicate some discontinuity. The Board's goal is to give financial statement users additional information concerning how long the auditor has been associated with the company. Such information is available now in many cases because changes of auditor must be disclosed on Forms 8-Ks, which have been available on the SEC's EDGAR filing system for the past 19 years. The problem is that locating that information can be an arduous task for most financial statement users.
Is this tenure information relevant or useful to financial statement users? Opinion varies on that question, but providing it in the audit report should not, in my view, involve additional cost either to the auditor or the issuer, and some financial statement users may find that information to be of interest. As we acknowledge in the release, the PCAOB has not drawn conclusions about how tenure may or may not correlate with audit quality. Nonetheless, as stated earlier, my view is that as a general matter financial statement users are well served by receiving as much information as possible that is even arguably relevant and useful about the relationship between an issuer and its auditor and tenure is one factor in that relationship. Financial statement users can then make their own decisions about whether the information is useful. Accordingly, I support the Board's decision to include this information in the audit report.
As discussed in our proposing release, commenters also supported changes to the auditor's report that describe the auditor's existing responsibility related to information outside the financial statements. The proposed standard requires the auditor's report to discuss the auditor's responsibility for the other information and state whether the auditor identified a material inconsistency, material misstatement of fact, or both, based on his or her procedures.
The Board is aware that in some cases, particularly in the case of larger companies, auditors already perform some procedures over the other information, such as reviewing it for consistency with the audited financial statements or correcting obvious errors. The proposal would require all auditors of public companies to perform specific procedures, which may differ from what firms are currently performing. The objectives and procedures are discussed in detail in the release.
I would like to focus on three issues relating to the proposal today. First, in evaluating whether the other information contains a material inconsistency, a material misstatement of fact, or both, the auditor would refer to the definition of materiality under the federal securities laws. In interpreting those laws, the Supreme Court has held that a fact is material if there is "a substantial likelihood that the . . . fact would have been viewed by the reasonable investor as having significantly altered the 'total mix' of information made available."[2] As the Supreme Court has further explained, determinations of materiality require "delicate assessments of the inferences a 'reasonable shareholder' would draw from a given set of facts and the significance of those inferences to him . . . ." I believe this is an appropriate level of materiality to be used here, as it is well established under the federal securities laws as applicable to corporate reporting. Auditors also should already be familiar with this definition because, among other things, it is used to evaluate whether uncorrected misstatements detected during the audit are material.
Second, the proposal does not indicate that the "other information" procedures are designed to provide a level of assurance (in a technical sense) over the "other information" in the annual report covered by the procedures. We want to be sure that this proposed standard communicates to users of the financial statements and annual reports exactly what the auditor has done, to leave the user of the information with a clear understanding of what reliance to place on the auditor's work.
Third, the Board has made an evaluation of the options for how broadly to have the other information procedures, and reporting requirements, apply. The Board proposes to have the requirements of the proposed other information standard apply only to certain types of annual reports filed under the Securities and Exchange Act of 1934 ("Exchange Act-filed annual reports"). We discuss in the release, particularly in Appendix 6, how this may lead to similar other information being treated differently. The Board at this time is not proposing to have the auditor perform procedures or report on other information contained in 1933 Act registration statements.[3] The Board is aware that in some circumstances a Securities Act registration statement will incorporate by reference a previously-filed annual report and in that case, the "other information" discussion will appear in the auditor's report, and those procedures would have been performed. However, where the Securities Act registration statement includes the information in the annual report in the actual registration statement, the procedures over "other information" (the performance requirement) and the discussion of those procedures (the reporting requirement), would not be required under the proposed standard.
The Board has asked for comments on whether this is the best approach to when to require adherence to the requirements of the other information standard. Also, are there likely potential adverse consequences from the possibilities that the same information could be treated differently depending on whether it is included in an Exchange Act-filed annual report or not? Moreover, the Board continues to assess whether it is possible to propose applying some of the elements of the proposed other information standard regarding the auditor's responsibilities over other information contained in documents filed under the Securities Act. We will be particularly interested in comments regarding whether the Board should expand the scope of the performance requirements.
I believe the proposal strikes a careful balance between meeting the desire of investors for more information from auditors while retaining the primary role of management in the financial reporting process. The identification of those items that the auditor identifies as CAMs should give financial statement users a view of the areas of greatest auditor concern and provide some guideposts to those areas of the financial statements and management disclosures that may merit special scrutiny. At the least they should give investors a basis for further inquiry and investigation. As a result, I support the proposed standard.
I also support inclusion of the auditor's tenure in the audit report because its inclusion is not costly and it will provide information that some financial statement users may find important. Finally, I support the adoption of procedures and standards for the evaluation of other information by the auditor. At the least, the standard will bring clarity and uniformity to an area where today there appears to be wide differences in practice among public companies.
[1] A number of commenters, particularly audit firm representatives, expressed concern that an AD&A would lead to "free writing" by individual audit engagement partners that could vary widely from engagement to engagement, as auditors expressed essentially impressionistic views of the audited entity. The commenters said this could make review both difficult and time consuming from a quality control and consistency viewpoint, could delay opinions, and could seriously diminish comparability among audit reports.
[2] TSC Industries v. Northway, Inc., 426 U.S. 438, 449 (1976). See also, Basic, Inc. v. Levinson, 485 U.S. 224, 231 to 232 (1988).
[3] One of the reasons the Board has exempted most other information contained in filings under the Securities Act of 1933 stems from peculiarities in that Act itself. For example, when a preliminary registration statement is filed under the Securities Act, the auditor's opinion is filed with it. The auditor's report, however, is dated as of its issuance date which may differ from the date of the preliminary registration statement. In the course of becoming effective, as the registration statement is reviewed by the staff of the Securities and Exchange Commission, it may change substantially from the initial registration statement reviewed by the auditor. The audit opinion, including its date, may not, however, change, so requiring the auditor to perform procedures on what may be many changes in the other information (i.e. the registration statement) between the preliminary and final registration statements would raise serious problems under a regime where the audit report included in the registration remains unchanged through the amendment process. In practice, the auditor normally gives the underwriter of the issuer a "comfort letter" that nothing has come to the auditor's attention that is inconsistent with the auditor's report, but that letter is not addressed to the same audience as the audit opinion itself.