Unintended Effects of Critical Audit Matters: Early Evidence Free

In 2019, the PCAOB implemented the most substantial change to the auditor report in the last 50 years. The PCAOB mandated critical audit matters (CAMs) for large accelerated filers.¹ The implementation of CAMs follows from a renewed interest of regulators to expand information provided in an audit report.

Auditors have long faced the criticism that the largely binary unqualified/qualified opinion is not very useful for assessing the company and the audit (Cohen 1978; Mock et al. 2013). The lack of information content is exacerbated by the observation that there is minimal variation even within the binary categories, as 98 percent of audits receive an unqualified opinion. There have been some efforts to enrich the audit opinion, including adding qualification paragraphs and explanatory language. However, it seems the resultant additional language did not fully address users’ needs, and regulators noted that “expectations gap” (Cohen 1978; International Auditing and Assurance Standards Board (IAASB) 2011; Association of Chartered Certified Accountants (ACCA) 2019) and “information gap” (IAASB 2011) were still present. The “expectations gap” refers to the disconnect between the tasks that stakeholders believe auditors conduct and are responsible for compared to the actual responsibilities of the auditor (IAASB 2011; ACCA 2019). The “information gap” is derivative from the “expectations gap” and refers to the additional information that stakeholders would like to obtain that is currently only known to the auditors in the course of the audit (IAASB 2011; Mock et al. 2013).

In efforts to reduce the “expectations gap” and “information gap,” audit authorities have applied new reforms. In the U.K. and Europe, regulators implemented key audit matters (KAMs). They hoped to “enhance the communicative value of the auditor’s report by providing greater transparency about the audit that was performed” (Financial Reporting Council (FRC) 2016; International Auditing and Assurance Standards Board (IAASB) 2015), thereby helping users understand both the audit and information from the audit. However, academic research has generally not found evidence of an information effect for KAMs (Minutti-Meza 2021), as the capital market does not seem to react to KAMs.

The PCAOB published its own standards for the U.S. market in 2017. The new CAM standards became effective for large accelerated filers with year-ends after June 2019 and all filers after December 2020. The new standard required auditors to disclose CAMs and the audit procedures performed to address these CAMs in the audit report. The formal definition of CAMs is “any matter arising from the audit of financial statements that is communicated to the company’s audit committee, relates to accounts or disclosures that are material to the financial statements and involve especially challenging, subjective or complex auditor judgment” (PCAOB 2017). Similar to international regulators, the PCAOB hoped the new CAMs can “provide investors with audit-specific information that should facilitate their analysis of the financial statements and other related disclosures” (PCAOB 2017).

Focusing on the first year of CAM adoption, Klevak, Livnat, Pei, and Suslava (2023) consider whether the new CAM regulation provides relevant information by examining the aggregate capital market reaction to the newly disclosed CAMs. They find that capital market participants perceive more extensive CAMs and audit procedures as indicators of higher uncertainty in company operations and future cash flows. This sentiment is shared by both general market and more sophisticated market participants such as analysts. Although CAMs are envisioned to report only on specific areas of the audit and audit procedures, early empirical evidence in Klevak et al. (2023) suggests that capital market participants consumed CAM disclosures as if they reflected company risk information. It seems that instead of narrowing the “expectations gap” and “information gap,” the new regulation has further exacerbated them.

Klevak et al. (2023) present an unintended consequence of the new CAM standard. Capital market participants seem to misinterpret the information about the “challenging, subjective, or complex” parts of the audit and the audit procedures performed as information about company uncertainty. Klevak et al. (2023) also provide some evidence of how the market may perceive the extensiveness of CAMs and audit procedures disclosed. In this article, we summarize the findings of Klevak et al. (2023) and discuss the implications for regulators and auditors.

At the time that the new CAM regulations were announced in the U.S., there was no consensus on how these new disclosures would affect investor perceptions of reporting companies. The only empirical evidence came from Europe where CAM-like disclosures called KAMs² have already been implemented. Similar to CAMs, they were envisioned as incremental information in the audit report that would help stakeholders better understand audit process (FRC 2016; IAASB 2015).

Early academic research of KAMs found that capital markets do not respond much to these new audit disclosures, presumably because they were boilerplate or the information had already been mentioned elsewhere. For example, Bédard et al. (2019) find that reporting of Justification of Assessments in France did not affect abnormal returns, trading volume, audit-report lag, audit quality, or audit fees. Similarly, in the U.K., Gutierrez et al. (2018) and Lennox et al. (2023) document no abnormal market reaction to new CAM-like disclosures under ISA 701.

In the U.S., both early academic studies of CAMs and practitioner comments suggest that investors might interpret new audit disclosures as a measure of company risk and uncertainty. Academics show that CAMs correspond with accounting uncertainty (Hollie 2020) and business risk (Christensen, Glover, and Wolfe 2014; Rapley, Robertson, and Smith 2021). A seminal study of U.S. CAMs by Burke, R. Hoitash, U. Hoitash, and Xiao (2023) observe that financial statement footnotes become longer and include more uncertain language for companies disclosing CAMs. In their market reaction tests, Burke et al. (2023) document negative abnormal returns for companies that disclose unexpected CAMs.³ They interpret this negative return as investors’ response to the increased perceived risk.

Institutional investors commented that they use CAMs to “analyse and price risks in our valuation and allocation of capital” (California State Teachers’ Retirement System 2016), to highlight a company’s “risks and challenges” (Hermes Equity Ownership Services Limited 2016), and to “compare [their] primary concerns with difficult issues highlighted by auditors” (Colorado Public Employees’ Retirement Association 2020). Wall Street Journal also shared the risk-defining view of CAMs, as it described CAMs as “giving investors a better view into potential problems that may not have been previously apparent” and “a spotlight on companies’ hairiest internal issues” (Maurer 2019).

As a result of the conflicting perspectives on the effects of CAM and CAM-like disclosures, it is not clear how the U.S. large accelerated filers’ reporting of CAMs may affect perceptions of company uncertainty. If CAMs are used as the PCAOB had envisioned, CAMs should give investors additional clarity about the audit process and the financial statement items (PCAOB 2017). This might result in lower assessment of audit risk and reduced perception of company uncertainty. It is also possible that CAMs might repeat some information that is already known to investors from prior management disclosures. In these situations, Klevak et al. (2023) predict that there should be no discernible additional reaction from the capital markets or analysts and no increased assessments of uncertainty.

On the contrary, if the “expectations gap” is not mitigated, capital market participants may attribute auditors’ CAMs as judgments on companies’ problem areas. Capital market participants may also not fully believe the assurance of audit procedures in the audit. Academic research has shown that the addition of audit procedures to a CAM does not change readers’ reaction to it (Kachelmeier, Rimkus, Schmidt, and Valentine 2020). In these cases, Klevak et al. (2023) predict that there will be increased assessments of company uncertainty in response to the extensiveness of reported CAMs.

Klevak et al. (2023) collect first year of CAM disclosures of large accelerated filers, resulting in a sample of 1,969 reports between July 2019 and May 2020. The study then measures the extensiveness of CAMs for each company using three metrics: (1) the length of CAM disclosures (measured as the number of words), (2) the total number of CAMs in each report, and (3) the total number of audit procedures mentioned in the CAM section. Table 1 shows the distribution of the number of CAMs and number of audit procedures for the 1,969 company annual filings. Most companies report one to three CAMs with one CAM being the most likely choice, and their auditors completed between three and five distinct audit procedures in attempts to address CAMs. Table 1 also lists topics and audit procedures by frequence, with fair value, acquisitions, and revenue recognition being the top-mentioned topics, and tests of controls being the top-mentioned procedure. These most frequent types of CAMs and audit procedures conform to the expectation that CAM are, by definition, complex, more likely to be subject to estimation, and require significant auditor judgment.

Klevak et al. (2023) use regression analysis to investigate the capital market response to the extensiveness of the new CAM disclosures. They find that the standard deviation of stock returns and the change in standard deviation of returns have positive relationships with CAM and audit procedure extensiveness. Table 2 shows the results of these tests. The coefficients in columns (1)–(3) show that CAM measures are positively associated with the standard deviation of stock returns at the filing date. Columns (4)–(6) show the association of CAM disclosures with the changes in the return volatility; all coefficients are positive, and AUDIT_SUM is statistically significant.

In terms of economic significance, the coefficient of 0.006 on the number of CAMs in column (2) means that companies with one CAM (lowest tercile of CAM_COUNT) experience 18 percent⁴ lower standard deviation of returns than companies with three or more CAMs (highest tercile of CAM_COUNT). Similarly, the positive regression coefficient of 0.009 in column (3) means that companies with more than 14 audit procedures (highest quartile of AUDIT_SUM) experience 27 percent higher standard deviation of returns than companies with less than five audit procedures (lowest quartile of AUDIT_SUM).

These results suggest that capital market participants perceive more CAMs and even more audit procedures to address the CAMs as information about enhanced business risk and company uncertainty.⁵ Investors increase their trading in the stock, which increases return volatility.

Table 3 shows the relationship between CAM and audit procedure extensiveness and analyst dispersion. Klevak et al. (2023) test whether the perception of CAMs as an indicator of company uncertainty also persists for more sophisticated market participants such as financial analysts. Positive coefficients suggest that dispersion (columns (1)–(3)) and change in dispersion of analyst forecasts (columns (4)–(6)) is greater for companies with more extensive CAMs and audit procedures. For example, companies with the most extensive CAMs (highest tercile of CAM_COUNT) and audit procedures (highest quartile of AUDIT_SUM) have 22 percent and 56 percent higher analyst dispersion, respectively, than the companies with the least extensive CAMs (highest tercile of CAM_COUNT) and audit procedures (highest quartile of AUDIT_SUM). These results suggest analysts are also more uncertain about the future of companies that have more extensive CAMs and audit procedures, resulting in more disagreement amongst them.

To corroborate the observation that more extensive CAMs and audit procedures appear to be related to increased uncertainty perceptions, Klevak et al. (2023) investigate stock return patterns around the disclosure of CAMs. Table 4 shows negative stock returns in the three-day window around CAM filing date. Specifically, companies with the most extensive CAMs and most extensive audit procedures earn 1.9 and 2.1 percent lower returns, respectively, than the companies with the least number of CAMs and least extensive audit procedures. This further confirms that capital market participants interpret more CAMs to be a sign of higher business risk, and, as a result, lower their valuations of the company.

The purpose of the new CAM standards was to provide more information about the circumstances of the audit and the formation of the audit opinion. In practice, early evidence seems to suggest that capital market participants interpreted CAMs to include information on company risk and uncertainty. Even when auditors described more audit procedures that they performed to mitigate audit risk, market participants took it as a signal of higher business risk.

One of the unexpected findings of Klevak et al. (2023) is that the market is more uncertain when more audit procedures are mentioned in the CAM section. Audit procedures were envisioned as additional language to help market participants better understand the details of the audit, thereby reducing audit risk. However, Klevak et al. (2023) show that various measures of market uncertainty (standard deviation of returns, change in standard deviation of returns, analyst dispersion, change in analysts’ dispersion) seem to be higher for companies with more audit procedures described in the CAM section. This additional uncertainty seems to be an unintended consequence of new CAM disclosures. To mitigate this concern, the PCAOB could provide more guidance on interpreting CAMs, release some new language to clarify that audit procedures lower audit risk, and educate investors about the purpose and value of audit procedures.

Klevak et al.’s (2023) findings suggest even though CAMs were intended to alert investors about challenging accounting items, capital market participants took these issues to represent business risk, and not just audit-specific issues. Auditors should be aware of this mistaken attribution when commenting on CAMs and might consider educating the readers of financial statements about the role of the auditor and the level of assurance provided in the audit. For example, auditors might consider explaining the effects of CAMs to a company’s audit committee when they communicate the results of their work. An audit committee might benefit from learning that when auditors report CAMs, stock values and volatility of the company may be affected,⁶ CAMs may be perceived by the capital markets as information that is unique and not reported in other sections of annual filings,⁷ and CAM information might receive extra attention after its disclosure⁸ as investors might be using information about CAMs to refine their perception of audit risk.⁹