Analytical procedures have been part of the audit process for decades, but many auditors fail to understand their objectives or how these procedures should be properly applied for substantive testing, resulting in many audit deficiencies, some of which are serious. Carefully designed analytical procedures can be as effective as substantive tests of details (or, depending on the circumstances, even more effective), but it is far too easy, in the interest of efficiency, to place undue reliance on weak analytical procedures that are too imprecise to have a reasonable chance of detecting a material misstatement.
The primary purpose of this article is to highlight the basics of substantive analytical procedures for use in smaller, less complex audits. Some of these principles may be subtle and therefore not readily apparent in the language of the standards. Although this article focuses on the proper use of analytical procedures as substantive tests in audits, it is not a complete summary of all of the guidance presented in the standards [principally AU-C 520 and PCAOB Auditing Standard (AS) 2305], which is more complicated than many auditors may realize. Auditors placing extensive reliance on substantive analytical procedures should be familiar with the requirements of the applicable standard. More complex analytical techniques that have crept into modern auditing, primarily in larger audit environments, are beyond the scope of this article.
Analytical Procedures as Substantive Tests
The current standards permit, but still do not require, the use of analytical procedures as substantive tests, but auditors commonly use them to achieve audit efficiency in two ways: 1) to corroborate substantive tests of details for the same assertion, thereby enabling a reduction in the scope of the tests of details (for example, by lowering the sample size); or 2) to use in place of a substantive test of details.
The extent to which an auditor can rely on substantive auditing procedures to reduce the risk of material misstatement (or the scope of related substantive tests of details) for any given assertion requires the application of considerable judgment, which may be deemed significant enough to warrant specific documentation (AU-C 320.A12 or AS 1215.12). Such judgments are based on many factors, including but not limited to—
- the nature of the assertion;
- the assessed risks of material misstatement;
- the planned scope of related tests of details;
- the plausibility and predictability of the relationship;
- the auditor’s evaluation of the availability and reliability of data from which an expectation of recorded amounts or ratios may be developed, taking into account the source, comparability, nature, and relevance of such data;
- the perceived precision and reliability of the expectation; and
- the effectiveness of controls over the preparation of the information that are designed to ensure its completeness, accuracy, and validity and the level of confidence therein obtained by the auditor through tests of controls (AU-C 520.05 and .A7; AS 2305.12–.19).
The foregoing notwithstanding, the PCAOB considers it “unlikely” that audit evidence obtained from substantive analytical procedures alone, without a related test of details, will be sufficient to address a significant risk of material mis-statement unless the procedures are supplemented by a test of internal controls determined to be functioning effectively (AS 2305.09). Although the Auditing Standards Board (ASB) standard on analytical procedures appears to permit greater use of auditor judgment as to such matters (AU-C 520.A7), another ASB standard (AU-C 330.A46) more firmly suggests that performing only substantive analytical procedures will not be sufficient to reduce audit risk to an acceptably low level unless supported by evidence obtained from tests of controls.
An auditor must obtain and apply knowledge of the business and the industry in which the client operates to develop reasonably reliable expectations effectively (AU-C 315.A1; AS 2305.03). In addition, the greater the risk of material misstatement for the assertion and the intended reliance to be placed upon the planned substantive analytical procedure, the more predictable relationships used in developing expectations, and the more persuasive the audit evidence obtained from the planned analytical procedure, need to be. When expectations are developed at a more detailed level, it is more likely that the analytical procedure will address more effectively the assessed risk of misstatement to which it is directed.
Relationships are usually more predictable in a stable environment than in an unstable one. In addition, substantive analytical procedures are generally more effective for operating statement accounts than they are for balance sheet accounts because the former represent large volumes of routine, repetitive transactions that tend to be more predictable over time. Relationships involving transactions that are subject to management discretion are often less predictable. Accordingly, it may be advisable to remove such discretionary or unusual transactions from a population before applying an analytical procedure and audit them separately. As expectations become more precise, the range of expected differences becomes narrower and the likelihood increases that significant differences from expectations observed are due to mis-statements (AU-C 520.A10, .A22, and .A23; AS 2305.14 and .17).
Auditors should either test the design and operating effectiveness of controls over financial information used in substantive analytical procedures or perform other procedures to support the completeness and accuracy of the information. The following factors should be considered in assessing the reliability of the data:
- The source of the data (i.e., whether obtained from independent sources outside the entity, sources within the entity, or a variety of sources)
- If the source of the data was internal, whether such source was independent of those who are responsible for the amount being audited, and whether the data was developed under a reliable system with adequate controls
- Whether the data (or the controls over its preparation) was subjected to audit testing in the current or prior year (AUC 520.05b and .A17–.A20; AS 2305.16).
Common Deficiencies Observed
The most common deficiencies in auditors’ use of substantive analytical procedures are threefold and consist of the failure to—
- obtain adequate evidential support for the data used to develop expectations (AU-C 520.05b, .A17, and .A19; AS 2305.16);
- develop reliable and reasonably precise expectations based on plausible and suitably predictable relationships (AU-C 520.05c; AS 2305.13-.14); and
- obtain adequate evidential support for variance explanations provided by management (AU-C 520.07 and A28–A29; AS 2305.21).
In addition, use of more disaggregated data in analytical procedures is generally regarded as more precise and reliable than summarized information (AU-C 520.A22–.A23; AS 2305.19; AR-C 90A.19d and .A146).
A list of several acceptable analytical procedures available (but not required) introduced into the auditing standards in 1978 for use in audits included year-to-year comparisons of financial information. The 1978 standard [Statement on Auditing Standards (SAS) 23.06–.07] noted that the extent to which such procedures should be applied to overall versus disaggregated financial information was solely a matter of judgment, but as in the current standards, it suggested that use of disaggregated data may be more effective, such as in the case of a diversified enterprise. The standard made no distinction in terms of the necessary robustness of the procedures applied based on whether their objective was that of a substantive analytical test, or, if so, its perceived reliability for detecting material misstatements.
Ten years later, that standard was superseded by a more robust standard that more closely matches today’s standards. In the 1988 standard (SAS 56.05, later codified as AU 339.05), the reference to year-to-year comparisons of financial information was subtly modified by adding the words, “giving consideration to known changes,” hinting at the need to develop a quantitative expectation before comparing data. This need was more clearly asserted in the same paragraph, which stated that “analytical procedures involve comparisons of recorded amounts, or ratios developed from recorded amounts, to expectations developed by the auditor … by identifying and using plausible relationships that are reasonably expected to exist based on the auditor’s understanding of the client and of the industry in which the client operates.” Although not set forth as a requirement in the standards, the basis for both establishing an expectation and for judging it to be reliable should be documented when placing significant reliance on it.
It is inherently unreasonable in most environments to expect the current year’s recorded amounts to equal the prior year’s. (If this is believed to be the case, the supporting rationale should be documented.) Nevertheless, even though most auditors today are too young to remember auditing before 1988, they remain stuck to the erroneous notion that a year-to-year comparison with explanations of variances (often unquantified and unsubstantiated, and merely judged by auditors to “look reasonable”) constitutes an effective, substantive analytical procedure upon which they can rely to reduce the risk of material misstatement to an acceptably low level.
Many of those who are smart enough to recognize the need to develop an expectation nevertheless frequently accept management’s expectation without challenge and without seeking supporting evidence. After comparing unsubstantiated expectations to the recorded current period amounts, many auditors will likewise accept management’s variance explanations, often unquantified. Where is the professional skepticism?
The need to examine and evaluate the adequacy of supporting evidence for both the expectations developed by the auditor for use as benchmarks and the explanations obtained from management of any variances from such expectations is the primary factor that distinguishes substantive analytical procedures performed in an audit from those that are generally considered adequate to support a review report.
Nonsubstantive Uses of Analytical Procedures
Modern auditing standards require the use of analytical procedures as risk assessment procedures for purposes of planning an audit and determining areas that require attention in developing audit scope. For example, they are used to identify unusual transactions and events, unexpected ratios and trends, areas with possible risks of management override of controls, and other matters with financial statement and audit planning ramifications (AU-C 315.A14–.A17; AS 2110.46–.48, and 2305.10 and .12). The use of analytical procedures is also required near the end of an audit for retrospectively assessing the adequacy of the audit scope, and possibly identifying a previously unrecognized risk of material misstatement, before forming an overall conclusion on the financial statements (AU-C 520.06 and .A25–.A27; AS 2810.05–.09). It is essential for auditors to recognize that the purpose of these preliminary and late-stage analytical procedures is not merely to explain variances (no matter how reasonable they may appear), but rather to assess the adequacy of the planned or completed auditing in addressing all risks of material misstatement.
Because the objectives of preliminary and late-stage applications of analytical procedures are not to gather audit evidence to support recorded values, the design and reliability of procedures performed for these purposes are not critical to supporting the audit opinion. Therefore, the procedures need not be as robust as when they are used as substantive tests. Accordingly, analytical procedures used for purposes of audit planning and risk assessment or retrospectively assessing the audit scope may be based on data aggregated at a higher level than is recommended for substantive analytical procedures (and for review engagements).