PCAOB Inspections and Audit Fees: An Analysis of Inspection Rounds of Small Audit Firms

Abstract

This study investigates the impact on audit pricing of PCAOB inspection outcomes on the quality control systems of small audit firms. Quality control deficiencies (QCDs) provide a strong signal about audit quality, which is expected to lead to price changing. We first document that audit fees in the pre-inspection period were higher for small audit firms without QCDs, implying that quality differences were already partially known and priced in the market. Next, we find evidence that in the post-inspection period audit fees further increase for firms without QCDs, while there is limited evidence for fee increases for firms with remediated QCDs, suggesting changes in audit firm reputation. Interestingly, we do not find support for fee changes for audit firms with disclosed quality control deficiencies. Finally, the magnitude of change in audit fees decreases over time. Collectively, these findings contribute to our understanding of audit quality differences amongst small audit firms and the functioning of the small audit firm market.

Keywords:

• PCAOB inspections
• Audit fees
• Quality control deficiencies
• Small audit firms

Acknowledgements

We thank Juha-Pekka Kallunki (Associate editor) and the anonymous reviewer for their helpful and constructive comments. We also thank NWO, the Netherlands Organisation for Scientific Research, for its financial support for this research project (NWO grant number 406-13-071). We gratefully acknowledge the valuable comments received from Limei Che, Matt Ege, Kris Hardies, Caren Schelleman, and workshop participants at ESSEC Business School, University of Antwerp, the 2017 EARNET symposium, the 2017 UF International Conference on Assurance and Governance and the 2017 International Symposium on Audit Research. Finally, we wish to thank Mona Offermanns and Robert Knechel for their involvement at an early stage of this project.

Notes

1 Other countries include Norway, The Netherlands for Big 4 firms from 2014, and the UK for annually inspected audit firms.

2 Examples of such actions include notifying the SEC, the US Justice Department, and disciplinary proceedings by the PCAOB such as censuring, suspending and barring auditors, or revoking the registration of audit firms, all of which can be accompanied by large financial penalties.

3 Audit firms with remediated quality control deficiencies are not included as a comparison group in Hermanson and Houston (2008). We understand that larger audit firms may be more likely to get an inspection report without quality control deficiencies as they might be better able to manage inspection risk compared to smaller audit firms. PCAOB inspection reports disclose variables reflecting audit firm size including number of partners, number of issuer clients, and number of total professionals to total clients in the inspection year (i.e., every 3rd year for the audit firms in our sample). Since we do not have panel data available, we are unable to control for these audit firm characteristics. As an alternative, we control for audit firm size using the average client total assets of the audit firm and the total revenue of the audit firms in the models to test our hypotheses.

4 We note that there might also be a reputation spillover effect from PCAOB inspection reports to non-audit services, which is beyond the scope of this paper.

5 If audit firms have quality control deficiencies identified during the inspection reports, the PCAOB states in Part II of the inspection report: ‘Any defects in, or criticisms of, the Firm's quality control system are discussed in the non-public portion of this report and will remain non-public unless the Firm fails to address them to the Board's satisfaction within 12 months of the date of this report’. If the audit firm does not have any quality control deficiency identified, the PCAOB states in Part II of the inspection report: ‘The inspection team did not identify anything that it considered to be a quality control defect that warrants discussion in a Board inspection report.’ This distinction in wording allows us to differentiate between firms with and without quality control deficiencies.

6 As Audit Analytics contains neither all inspected audit firms nor the full set of an audit firm's clients, it is not possible to match the inspected audit firms with all their clients. Furthermore, information is incomplete for certain client observations owing to missing data or missing identifiers for matching the different databases. Hence, a number of inspection reports are excluded from the analysis.

7 The number of years before the publication of the first-round inspection varies for client-audit combinations. For example, Moore Stephens Frost PLC had its first inspection report published on 21 January 2005 (this is the time that the PCAOB published its first inspection report on small audit firms). The pre-inspection period includes all clients from 2003 and 2004. Radin Glass & Co LLP had its first inspection report published on 25 January 2007, and thus the pre-inspection period includes all clients from 2003 to 2006.

8 There are two reasons why the number of inspection reports decreases from the first to the second and third inspection round. First, new audit firms entering the market at different times after 2005 did not have their second/third inspections conducted at the end of the sample period. Second, according to Lennox and Pittman (2010), small audit firms that are unable to afford the costs of remaining in the public sector exit the market.

9 For example, for the second inspection sample, we excluded all the observations that have a fiscal year end before the publication of the first inspection and after the publication of the third-round inspection.

10 All variables with dollar values are adjusted for inflation.

11 If an audit firm only has one client or all clients are from the same industry, the auditor is not recognised as an industry specialist.

12 We winsorize our data at the 5th and 95th percentile to make sure that most of the continuous variables remain in the range of three standard deviations from the means.

13 The average ROA is highly negative even after winsorizing at the 5th percentile. This is partly driven by the financial crisis. We control for year fixed effects in our empirical analyses.

14 AS2 and AS5 are not included in the descriptive statistics and the regression analysis for the third-round inspection as only one observation in our third inspection sample is subject to AS2 and all remaining observations are subject to AS5.

15 All VIF values are below 4 (excluding the dummy variables and interaction terms).

16 As we use the natural logarithm of audit fee as our dependent variable, the economic magnitude is calculated as (1-e-0.101)*100 percent.

17 We acknowledge that small audit firms deregistering from the PCAOB in response to inspections might reduce the supply of audit services and therefore may have an effect on audit fees set by other firms. We calculated, based on data on the number of US small audit firms from PCAOB annual reports and on the number of US issuers audited by US small audit firms from Audit Analytics, that the average number of issuer clients per audit firm increases from 2005 to 2009 from 7.24 to 8.43 and then decreases to 6.58 in 2015. These data do not seem to provide support for a more concentrated small audit firm market over the entire sample period. We cannot include these yearly data in our models given that we include year fixed effects. Further, if the audit market is becoming more competitive, audit fees should arguably increase for all audit firms, which is not consistent with our findings that audit firms with disclosed quality control deficiencies decrease audit fees.

18 The descriptives in Table 12 differ from the descriptives in Table 3 as the samples used are different. For example, we include new clients added to the audit firm portfolio after publication of the first round of inspection as our sample in Table 12, Panel A, while in Table 3, Panel B, we included all clients of an audit firm who did not switch auditors for at least one year after the publication of the first round inspection reports. Another reason for the difference is that the sample size for Table 12 is much smaller compared to the sample size for Table 3. Consistent with Hermanson et al. (2007), there is high variability in our data (extreme observations remain even after winsorizing at the 5th and 95th percentile). As a result, the means can vary significantly, especially when the sample size is small.

19 We match without replacement and with a caliper of 0.05.

Additional information

Funding

We thank Juha-Pekka Kallunki (Associate editor) and the anonymous reviewer for their helpful and constructive comments. We also thank NWO, the Netherlands Organisation for Scientific Research, for its financial support for this research project (NWO grant number 406-13-071).