When Do Investors Value Key Audit Matters?

Abstract

While prior studies have considered the effect of the enhanced auditor’s report (EAR) on a range of auditor, management and investors’ judgments, we consider whether the inclusion of key audit matters (KAMs) in the EAR affects investor perceptions of the value of the audit as well as the credibility of the auditor. These are important issues linked to the rationale for the introduction of the EAR. As prior studies have demonstrated that audit firm size affects investor perceptions across a range of measures, we predict that the link between KAMs (absent versus present) and investor perceptions around value and credibility is affected by audit firm size. We find that the inclusion of KAMs improves perceived value and credibility only when a Non-Big 4 firm conducts the audit. When a Big 4 firm conducts the audit, perceived value and credibility are high whether KAMs are included or not. We also find that perceived credibility mediates the joint effect of KAMs and audit firm size on the perceived value of the audit. In additional analysis, we find that the inclusion of KAMs draws investors’ attention to new and expanded messages, taking their attention away from messages considered core to the audit report.

Keywords:

• Key audit matters
• Audit firm size
• Perceived value
• Perceived credibility

JEL classifications:

• M42

Acknowledgements

We thank William Messier (the editor), two anonymous reviewers, participants at the 2019 AFAANZ Annual Conference, 2019 EAA Annual Congress, 2018 ANCAAR Audit Research Forum, and the Australian Auditing and Assurance Standards Board for their helpful comments and suggestions.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

ORCID

Robyn Moroney http://orcid.org/0000-0003-4434-617X

Soon-Yeow Phang http://orcid.org/0000-0002-5647-1122

Xinning Xiao http://orcid.org/0000-0003-2701-9767

Notes

1 According to paragraph 16 of ISA 701 Communicating Key Audit Matters in the Independent Auditor’s Report, the auditor can determine, depending on the facts and circumstances of the entity and the audit, that there are no KAMs to communicate or that the only KAMs communicated are those matters giving rise to a modified opinion in accordance to ISA 705 Modifications to the Opinion in the Independent Auditor’s Report or material uncertainty related to the going concerns in accordance to the ISA 570 Going Concern.

2 There are circumstances where KAMs may not enhance the perceived value of an audit. This will occur when the information provided in KAMs is already known or expected (Bédard, Gonthier-Besacier, & Schatt, 2019), when KAMs are boilerplate such that they do not provide new or useful information (Brasel et al., 2016) and when technical language is used, hindering investor understanding of KAMs (Bédard et al., 2019; Reid et al., 2019). Given the ongoing debate on the consequences of KAMs, it remains an empirical question whether KAMs affect the perceived value of the audit, when no other information is provided and the language used is not boilerplate or technical.

3 The source credibility framework developed by Hovland and his colleagues has been widely used in persuasion research (see Pornpitakpan, 2004, for a review).

4 We conducted the experiment across two weeks of a 12-week course. None of the researchers taught the course. A further sixty-six students participated in the experiment where the audit report was in the old format. Results for this group are included in a footnote in the results section of the paper.

5 This is similar to the age of participants in other studies (Cheng, Green, & Ko, 2015; Davidson & Stevens, 2013).

6 This is slightly lower than the 25 percent reported by Elliott, Grant, and Rennekamp (2017) and the 27 percent reported by Dilla, Janvrin, and Jeffrey (2013). Seventy-four percent of participants come from families who invest. In 2014, around 45 percent of investors sought advice from personal networks (friends, family and colleagues) to make their investment decision, and in 2017, 37 percent of Australian adults (6.9 million people) invest on the stock exchange (ASX, 2017).

7 This is comparable to Harding and He (2016) whose participants had completed an average of 4.77 finance courses and Tan, Wang, and Zhou (2014) whose participants had completed an average of 2.34 (2.87) accounting (finance) courses.

8 Participant gender (p > 0.7), age (p > 0.1), major (p > 0.2), work experience (p > 0.1), and investing experience (p > 0.5) are all not significant when included as covariates in our main analysis.

9 The researchers provided the parameters for the information to be included in the task. The researchers requested that the financial statement extracts and KAMs be generic. The board member is a partner at a Big 4 firm. We base the materials on a real company’s financial statements and audit report.

10 Participants in the KAMs present manipulation received an audit report with either one or three KAMs. We varied the number as members of the AUASB were interested in whether number of KAMs affects participant responses. We collapsed the two number of KAM conditions to conduct our 2×2 analysis. We explain the results for number of KAMs in a footnote at the end of the Results section.

11 This information was included in the case materials.

12 While our findings are statistically significant, the difference between the means across the four cells is relatively small. We use an adaptation of Cohen’s d (Cohen, 1988) for unequal samples, which allows us to compare the real difference between the mean in cell 3 with the mean across cells 1, 2 and 4. We find that the effect size for value is small (d = 0.393). Our results suggest that the mean difference in perceived value of the auditor is of modest practical significance.

13 We again calculate Cohen’s d (Cohen, 1988), comparing the mean in cell 3 with the mean across cells 1, 2 and 4. We find that the effect size for auditor credibility is medium (d = 0.593). Our results suggest that the mean difference in perceived credibility of the auditor is of moderate practical significance.

14 Prior research uses a similar approach to conduct contrast-coded path analysis (e.g. Brown, 2014; Clor-Proell & Maines, 2014).

15 The Comparative Fit Index (CFI) is 1, which is above the generally accepted minimum value of 0.95 (Byrne, 2013) and the Root Mean Square Error of Approximation (RMSEA) is 0, below the 0.05 rule of thumb indicating a good fit (MacCallum, Browne, & Sugawara, 1996).

16 An additional 66 participants received an audit report in the old format. We re-ran our analysis for those who received an audit report in the old format and those who received an audit report in the new format (no KAMs). There were no significant results for audit report format. The KAMs present treatment included 66 participants who read an audit report with one (1) KAM and 67 participants who read an audit report with three (3) KAMs. We re-ran our analysis to determine whether the number of KAMs affects perceived value and perceived auditor credibility and it did not.

17 Audit report format (old versus new) and number of KAMs (one versus three) do not impact any of these variables.

18 Audit report format (old versus new) and number of KAMs (one versus three) do not impact any of these variables.

19 Audit report format (old versus new) did not impact any of these variables. Number of KAMs impacted perceived riskiness only, with participants indicating that the investment was riskier when fewer KAMs were reported (t194 = 1.96, p = 0.05, two-tailed, untabulated).

20 Participants could not go back and re-read the case including the audit report when answering our questions.

21 Two coders, who were blind to the conditions, performed the coding. Among the two, one was independent from this study and was blind to the research question and hypotheses. The coders’ initial agreement rate was 90%. Inter-rater reliability of the coding is sufficiently high (Cohen’s Kappa = 0.89, p < 0.01). .

22 They identified the following items as being core: the auditor’s opinion, truth and fairness of the financial statements, compliance with accounting standards, and auditor independence. They identified the following items as being new or expanded: explanation of accounting treatments, the auditor’s responsibility, KAMs, audit assertions and procedures, comprehensiveness of the audit, and faithful representation.

23 It is unclear why firm size has an apparent effect on participant listing of core items from the audit report. This observation is interesting but unexplainable based on the data collected.

24 When we remove KAMs from the list of new/expanded messages, the difference is only significant when the audit is conducted by a Non-Big 4 firm (t194 = 2.81, p < 0.01, two-tailed, untabulated).