By Douglas R. Carmichael, Chief Auditor
It is a distinct pleasure to speak to this group that uniquely blends a passionate interest in auditing standards, practice, and education. I have the honor of being the first Chief Auditor of the Public Company Accounting Oversight Board – the primary advisor to the Board on policy and technical issues related to the auditing of public companies, including auditing standards, and the head of the PCAOB’s professional standard-setting division. However, the views I express today are my own and do not necessarily reflect the views of the Board members or other staff members. I am sure that this audience has a general understanding of the PCAOB’s origin, composition, and areas of responsibility, including registering public accounting firms, inspecting the practices of those firms, setting professional standards of auditing, ethics, quality control and independence, and enforcing those standards and applicable securities laws. I do not plan on explaining those things in any detail in my prepared remarks, but I would be pleased to address any questions you have in those areas later. In trying to decide what to talk about today I reviewed the PCAOB’s charge as stated in the Sarbanes-Oxley Act of 2002. The Act requires the PCAOB – … to oversee the audit of public companies that are subject to the securities laws, and related matters, in order to protect the interests of investors and further the public interest in the preparation of informative accurate and independent audit reports for companies the securities of which are sold to, and held by and for, public investors. To carry out this mission, the Act instructs the PCAOB to perform any activities the Board – … determines are necessary or appropriate to promote high professional standards among, and improve the quality of audit services offered by, registered public accounting firms and associated persons, or otherwise to carry out this Act in order to protect investors, or to further the public interest. In short, the mission of the PCAOB is to restore the confidence of investors, and society generally, in the independent auditors of public companies. There is no doubt that repeated revelations of accounting scandals and audit failures have seriously damaged public confidence. The focus on public confidence caused me to recall the work of a widely respected teacher of auditing who observed that if the confidence that society places in the effectiveness of the audit and the opinion of the auditor is lost, then the social usefulness of the audit is destroyed. I am speaking, of course, of Professor Theodore Limperg of the University of Amsterdam and his basic theory of the auditor’s function that has come to be known as the Theory of Inspired Confidence. In a series of essays published over 70 years ago, Professor Limperg set forth a dynamic theory that connects society’s need for reliable financial information to the ability of auditing methods to meet that need. He explained how changes in the needs of society and changes in auditing methods interact to bring about changes in the auditor’s function. Limperg based his theory on the science of business economics and viewed the development of the audit function from an economic perspective. From this perspective the development of a separate function of auditing was a natural product of differentiation in production. As with any other aspect of the production process, when the process could be carried out more efficiently by an autonomous branch than in combination with other processes a separate function developed. He observed, however, that with respect to the practice of public accounting the differentiation was caused by more than efficiency in the production process. The independent auditor acts as a confidential agent of the community, or an agent of confidence for society. For the function of confidential agent, independence of the auditor is essential. “If the community wants to be truly served by the function of the confidential agent, then it cannot be satisfied with an unqualified opinion of the enterprise’s employee accountant. The community asks for an independent opinion of the accounts of the stewardship of the managers, and that can apparently not generally be expected of one who is in the service of the manager and who, organizationally speaking, has to dance to that manager’s tune.” Limperg also pointed out that independence does not relate to the character of the auditor, i.e., the individual’s integrity and objectivity, but rather the logic of the structural arrangement. “The community cannot and may not be content with a relationship which is irrational, economically speaking, and which does not provide the required guarantees for the confidential agent’s effective functioning.” Limperg observed that the confidence inspired by the independent auditor was the essence of the function, its very reason for existing. He described the social responsibility of the independent auditor as follows: “The auditor-confidential agent derives his general function in society from the need for expert and independent examination and the need for an expert and independent opinion based on that examination. The function is rooted in the confidence that society places in the effectiveness of the audit and in the opinion of the [auditor]. This confidence is therefore a condition for the existence of that function; if the confidence is betrayed, the function, too, is destroyed since it becomes useless.” Limperg noted that if the function of the independent auditor is to achieve its objective, then no more confidence may be placed in its fulfillment than is justified by the work carried out, and by the competence of the auditor. Limperg described the essence of the Theory of Inspired Confidence as follows: “The normative core of the Theory of Inspired Confidence is therefore this: the [auditor] is obliged to carry out his work in such a way that he does not betray the expectations which he evokes in the sensible layman and; conversely, the [auditor] may not arouse greater expectations than can be justified by the work done.” This takes the principles-based approach to its logical extreme. At this extreme, there are no definite rules for what procedures an auditor must perform in a particular case, but the general principle that guides the auditor is to perform enough work to meet the expectations the auditor has aroused in society. Thus, the most important factor is society’s needs, and the related factor that interacts with it is the ability of auditing methods to meet society’s needs. However, society’s needs are not fixed and change over time. Also, auditing methods can change and improve over time. In other words, changes in the needs of society and changes in auditing methods combine to result in changes in the auditor’s function. However, the touchstone for the auditor is always to perform the work and obtain the evidence necessary to provide the assurance that society needs and reasonably expects. The enduring principles of Limperg’s theory are especially pertinent at this juncture of the development of the auditing function. We have a particular need in our current environment to try to understand and appreciate the social significance of auditing and the implications for how an audit should be performed. Limperg posited an audit function in the Netherlands that had to be more responsive to the public perception of the assurance provided by an audit than, for example, the British system in which, at least from the early 1900’s on, the audit of financial statements was compulsory as a matter of law. The United States system developed along lines closer to the British system, but without the legal requirement for an audit until passage of the Federal securities laws in the 1930s. The Securities and Exchange Commission, beginning in the 1940s, required that an audit be performed in accordance with generally accepted standards, but permitted the accounting profession to set those standards. Thus, independent auditors made their own interpretation of their role in society. Using Limperg’s framework, auditors, by setting their own standards, were to a degree insulated from the consequences of public dissatisfaction with their interpretation of their own social responsibility. In the late 1970’s, the Cohen Commission identified an expectation gap between what auditors believed to be their role and what user groups believed. This gap was concrete evidence of a lack of social acceptance of the role of the auditor as it had been defined by auditors. The passage of the Sarbanes-Oxley Act of 2002 is an even stronger indication that the social usefulness of the audit as interpreted by practicing auditors was being called into serious question by society. The Act removed from auditors the ability to exclusively interpret their role in society. That task is now in the hands of the PCAOB, and the Board’s charge effectively requires it to be responsive to the public perception of the assurance that society needs and reasonably expects from an audit. The first auditing standard that the Board adopted that, similar to all Board rules, will not become final until approved by the SEC, requires that the auditor’s report refer to the audit as being performed in accordance with the standards of the PCAOB. A reference to performance in accordance with generally accepted auditing standards (or GAAS) is no longer appropriate. The reference to generally accepted standards implied that the legitimacy of the standards depended on their general acceptance by auditors, i.e., a consensus of the accounting profession. That is no longer the case. The audit report reference to the audit being made in accordance with PCAOB standards is much more than a change in the identification of the source of the standards. The PCAOB adopted auditing and other professional standards that had been developed by practicing auditors as they existed at April 16, 2003. When the SEC approved that decision on April 25, 2003, these standards became Federal law. Thus, the standards that determine the work an auditor must do to give an opinion are now a matter of Federal law. However, the performance of the audit has not become a matter of compliance with static legal requirements. The common principle for the practice of auditing that Limperg postulated is now embedded in the mechanism for developing professional standards. The PCAOB’s process must be directed by what is necessary to protect investors and further the public interest. The dynamic theory that connects society’s need for reliable financial information to the ability of auditing methods to meet this need is the essence of the process that the PCAOB must follow. The views of investors and the public will be important input to the process. The PCAOB with the assistance of practicing auditors will need to determine the auditing methods necessary to meet society’s needs. One means the PCAOB will use to accomplish this is the Standing Advisory Group that will soon be appointed. The procedures and evidence necessary to support an audit opinion will have to be grounded on the confidence that society places in the effectiveness of the audit, as interpreted by the PCAOB in accordance with the Act. The way that the PCAOB’s process can be expected to work to develop standards based on the interaction of society’s needs and the ability of auditing methods to meet those needs is demonstrated by another standard recently proposed by the PCAOB – audits of internal control over financial reporting. This standard is related to Section 404 of the Sarbanes-Oxley Act. The SEC adopted rules implementing Section 404(a) that require management to assess and report annually on internal control over financial reporting. The PCAOB has proposed a standard to implement Section 404(b) on the auditor’s attestation of management’s assessment. This proposed standard involves something much more than additional information being covered by the auditor’s report. It is Limperg’s concept of an expansion of the audit function in action. The financial reporting of a public company under the Federal securities laws is a continuous disclosure process. As this continuous disclosure process evolved, the audit function remained largely an annual expression of opinion on the reliability of financial information—annual financial statements—that several months earlier had already impacted securities prices. Auditors did not expand their role to make timely review of quarterly financial statements required until relatively recently. Under the PCAOB’s proposed standard, the audit of a public company is now an audit of the company’s financial reporting process. The annual and quarterly financial statements are the primary outputs of the process, but internal control over financial reporting provides discipline and safeguards over the process that produces those financial reports as well as other timely releases of financial information. The ultimate objective of this new audit of a public company is to meet society’s need for reliable outputs of the financial reporting process. Reliable and credible financial information promotes informed investment and lending decisions. Investors need and want assurance that the public company’s releases of financial information during the period between audited annual financial statements are reliable. The PCAOB’s standard includes many imperatives on the scope of work necessary to give an opinion on internal control over financial reporting. These imperatives are driven by an evaluation of what public expectations require. The auditor has to perform enough work to provide assurance not just that management’s assessment process is adequate, but that the controls over the financial reporting process are effective. The approach to development of this position is the one formulated by Limperg – what work is necessary to provide warranted assurance that the outputs of the financial reporting process are reliable? As Limperg might have put it: The audit procedures required by the PCAOB’s standard are determined by the rational technical demands that arise from the confidence that society will attribute to this new audit function. Limperg’s formulation also has important implications for educators who want to assure that they are teaching the knowledge and skills that will be needed by auditors to perform an audit function that meets society’s needs. For example, as the accounting information in financial statements includes more assets and liabilities stated at fair value, auditors will need the knowledge and skills of a valuation expert. Auditors now meet these needs by using the work of valuation specialists, but that is not sufficient to meet society’s need for reliable fair value financial information. The accounting curriculum must expand to include valuation expertise as part of an auditor’s basic knowledge. I do not want to give the impression by making this recommendation that I favor unrestricted expansion of use of fair value information. In keeping with Limperg’s theory, the auditor should not arouse greater expectations than can be justified by the work performed. When fair value information is based on a model that incorporates management intent rather than being based on objectively verifiable market prices, there is a serious risk of arousing unjustified expectations. This possibility should be of concern to all those involved in the financial reporting process. In particular, accounting standard-setters should carefully consider the verifiability of required information and independent auditors should not accept information that does not meet reasonably expected levels of reliability. Limperg’s formulation also includes society’s expectations being informed by the auditing methods that are understood to be used by auditors. For example, users generally understand that auditors use a sampling approach and that an auditor does not examine every transaction. A reasonable user does not expect that accounting information is precise, but only that it is materially correct. Auditors, and educators of auditors, need to be mindful of the auditing methods that have engendered the confidence of society and not abandon these methods unless they are being replaced by even more effective approaches as part of expanding the function to meet the increased needs of society. For example, the sampling approach that has long been used in auditing depends for its justification on several assumptions. One is that the auditor will understand the company’s business and industry well enough to notice significant or unusual items when they are included in the items tested, scanned, or reviewed. Another is that accounting information is produced by a controlled system that includes overriding disciplines, such as debits being equal to credits and other controls that we take for granted in an integrated accounting system. Auditing standards have long reflected this assumption. The codification section on evidence contains the auditing equation that evidential matter supporting the financial statements consists of two components—the underlying accounting data and all corroborating information available to the auditor. It cautions that “without adequate attention to the propriety and accuracy of the underlying accounting data, an opinion on the financial statements would not be warranted.” (AU 326.16) It also notes that the “internal consistency” of the accounting records itself provides evidence about the fairness of presentation of the financial statements. (AU 326.19) This internal consistency cannot be known to the auditor unless the auditor, among other matters, actually determines that the financial statements agree to or reconcile with the general ledger. Yet in practice today, there are auditors that only audit a trial balance that has not been reconciled to the general ledger. Some auditors would be quick to explain that the trial balance is generated by a computerized general ledger system and that, therefore, actually looking at the general ledger accounts is unnecessary. The reality is that the trial balance is only a piece of paper. It is not the general ledger. The approach to auditing that has evolved over decades and that society has come to place confidence in is based on evidence from the internal consistency of accounting records that is not being obtained by some auditors today. Some would counter that the reviews of the computerized system made by the firm’s computer audit specialists ensures that the trial balance can be relied on as being accurately generated by the computerized general ledger system. The reality is that there is often insufficient discussion between the computer auditors and general auditors for the general auditors to know what assurance is provided by the work being done by the computer specialists. The general auditors do not have enough knowledge to know when computer audit techniques must be used to retrieve data directly from the company’s computerized accounting records. As auditing educators, you can make sure your students are acutely aware of the need to obtain a thorough understanding of both the manual and computerized aspects of the accounting system. Another previously neglected area in both auditing education and practice is expertise in the ways that management and employee frauds are perpetrated and concealed. The PCAOB in its inspections of auditing practice in 2004 will be very interested in how well SAS 99 on the auditor’s responsibility for fraud detection is implemented. The link here to Limperg’s theory is clear. The detection of material fraud is a reasonable expectation of users of audited financial statements. Society needs and expects assurance that financial information has not been materially misstated because of fraud. Unless an independent audit can provide this assurance, it has little if any usefulness to society. New required auditing procedures will be of little use unless the auditors implementing them understand the possibilities of fraudulent misstatement of financial statements and how perpetrators attempt to conceal fraud. You can make sure your students have that knowledge. Limperg’s theory also sheds light on the issue of whether there should be two sets of auditing standards. It is difficult to imagine circumstances that would justify a divergence between levels of assurance because different standards were applicable to audits of public versus private companies. The only possibility that Limperg’s theory would countenance would be if the perception of users of private company financial statements was that auditors of such financial statements inspired a distinctly different level of confidence in the effectiveness of the audit. This could occur if auditors of private companies, by design or direction, intended to create this difference in public perception, but it is difficult to see how that would be in the public interest. In closing, I would like to acknowledge that my remarks about Limperg’s theory are based on a booklet published by the Limperg Institute’s Scientific Board that had his essays translated into English. This publication issued in 1985 is a convenient source that I recommend to you. My first contact with Limperg’s ideas, however, goes back to the early 1970’s when, as the first auditing researcher at the AICPA, I wrestled with giving wider exposure to Limperg’s enduring theory of auditing. At that time, members of the firm he founded, Moret & Limperg, assisted me in translating his essays. I assure you that the translation made available in the mid-1980’s is more elegant than the translation I worked with. One key difference relates to the Dutch phrase that Limperg used to describe his theory—“leer van het gewekte vertrouwen” of which “Theory of Inspired Confidence” is referred to in the essay as “a rather clumsy sounding English translation of the Dutch.” In my earlier translation, the phrase we used was the Theory of Created Confidence. I think it is a more apt phrase because it captures the notion that the confidence the public has in the auditor’s opinion is one that has to be created. It is one that must be earned by those who practice and teach auditing and also by those who aspire to set standards for auditors. A great many things in my own career and in the world of auditing prevented me from writing or speaking about Limperg’s theory until now, but I sincerely appreciate the opportunity to share those thoughts with you today.