The International Standards on Auditing and the Accuracy of Auditors’ Going Concern Disclosures: Evidence for Private Firms in a Low Litigious Environment

ABSTRACT

This paper examines the effect of adopting the International Standard on Auditing (ISA) 570 on going concern (GC) disclosures in the audit report in a European environment characterized by low litigation and reputation risk. We exploit a natural experiment in Belgium, where the global GC standard ISA 570 replaced the local GC standard. We examine whether this shift in standards changed the accuracy of auditors’ GC disclosures. Using a sample of 30,339 stressed private companies with audited financial statements for periods ending on or after 15 December 2012 till 14 December 2016, our findings suggest that the adoption of the ISA 570 improves the accuracy of auditors’ GC disclosures. Specifically, we find that the implementation of the ISA 570 is associated with an 8 percent decrease in false positive misclassifications (GC opinion without subsequent bankruptcy), while the rate of false negative misclassifications (bankruptcy without prior GC opinion) remains unchanged.

Keywords:

• Auditing standards
• Going concern
• Audit quality
• Private firms

Acknowledgments

We thank Andrei Filip (Editor), Ioannis Tsalavoutas (Guest Editor), Lisa Evans (Guest Editor), Fanis Tsoligkas (Guest Editor), Arpine Maghakyan (discussant), Marleen Willekens (discussant), Kris Hardies, Dries Schockaert, and two anonymous reviewers for their support and helpful comments. This paper has also benefited from the comments of the participants of the 2021 EARNet PhD workshop, the 2022 BAFA Audit & Assurance Conference, the 2022 EAA Annual Congress and the ACCO Seminar at the Vrije Universiteit Brussel. This paper has been considered as part of the submissions for the Joint Special Issues on Corporate Disclosures in Accounting in Europe and The British Accounting Review, with the support of the IASB. It was one of the seven papers that were selected for presentation at the online workshop related to these special issues, organised by the Adam Smith Observatory of Corporate Reporting Practices at the University of Glasgow, with the support of the IASB, on 27 June 2022. The presentation is available through the following link: https://www.gla.ac.uk/research/az/adamsmithobservatory/events/headline_886719_en.html

Disclosure Statement

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

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 As of 2023, ISAs have been either fully or partially adopted by 135 countries around the world, including all countries within the European Union (IFAC, 2023). Countries can adopt ISAs in varying forms: (1) ISAs can be required by law or regulation; (2) ISAs can be adopted by the national standard setter; (3) National Standards can be ISAs with any modifications to meet local requirements (Boolaky & Soobaroyen, 2017).

2 Although traditionally referred to as ‘‘errors’’ these misclassifications do not necessarily imply that the audit was sub-standard. In line with Hardies et al. (2018), we therefore use the term ‘‘misclassification’’ rather than ‘‘error’’ as many reasons next to audit failures might explain why such misclassifications occur. Although company failure is an imperfect measure of audit failure, it nevertheless provides a reasonable benchmark for analyzing variation in auditor reporting accuracy (Lennox, 1999).

3 The level of vagueness (as opposed to precision) of auditing standards ‘refers to the variation in possible compliance effort […] reflecting imprecise wording, choices in procedures, intensity of testing, and other factors’ (Simunic et al., 2015, p. 223).

4 Prior to the reform in 2002, the International Auditing and Assurance Standards Board (IAASB) was referred to as the International Auditing Practices Committee (IAPC).

5 The applicable auditing standards in the US issued by the Public Company Accounting Oversight Board (PCAOB) are characterized by a high degree of precision and therefore commonly referred to as being rather rules-based.

6 Or an average of only 0.25 lawsuits per year. Given a population of around 20,000 audited companies in Belgium, this implies an annual audit failure rate of 0,00125% compared to 0,28% in the U.S. (Palmrose, 2000). While lawsuits reach a rate of zero even in a high litigious environment like the U.S. ‘the concern for litigation is not so much the frequency of lawsuits, but the potential magnitudes of lawsuits and the possibility that one or a few large lawsuits have the potential to bankrupt an accounting firm as happened in 1991 with Laventhol Horwarth, the 7th largest US accounting firm at the time’ (Francis, 2004, p. 347) or Arthur Andersen mid-2002.

7 Belgian Law on the organisation of the profession and public supervision of company auditors of December 7, 2016 (implementing the European Directive on statutory audit 2006/43/EC amended by the European Directive 2014/56/EC, repealing all former audit laws).

8 Before (after) December 2015, companies were considered large if they met at least two of the following criteria: (1) turnover (excluding VAT) > 7.3 (9) million euros; (2) total assets > 3.65 (4.5) million euros; and (3) number of employees (yearly average) > 50. These criteria must be considered on a consolidated basis if the company belongs to a group that publishes consolidated statements or if the company is a holding or a listed company.

9 Although reputation risk is often seen in the literature as an incentive for delivering high audit quality, there is little direct evidence of this. For example, Lennox (1999) evidences empirically that litigation risk rather than reputation risk incentivizes auditors. To rule out the confounding effect of litigation, Weber et al. (2008) show in a low litigation setting that clients from auditors subject to a scandal sustain reduced market value and that the auditor experience client loss. The authors conclude that reputation risk does incentivize auditors. However, since this and other studies (e.g., Skinner & Srinivasan, 2012) are based on rare events where reputation is harmed, they still do not provide evidence that reputation risk increases audit quality.

10 Only since January 1, 2017 the responsibility for quality controls, inspections and investigations has shifted away from the professional body to the Belgian Audit Oversight Board (‘College van Toezicht’) and the Sanctions Committee of the Financial Services and Markets Authority (FSMA) (art. 32 Belgian Audit Law).

11 This information is based on the 2012–2016 annual reports of the Belgian Institute of Registered Auditors.

12 The International Standards on Auditing (ISA) including the ISA 570 were adopted by the International Auditing and Assurance Board (IAASB) on December 15, 2008 and became effective for audits of financial statements beginning on or after December 15, 2009. In Belgium, the ISAs only became effective for audits of financial statements ending on or after December 15th, 2012 for Public Interest Entities (PIEs) and December 15th, 2014 for all other entities, following the approval of the Standard on the application of the ISAs in Belgium by the Ministry of Economic Affairs on April 1, 2010. In accordance with the European Directive 2014/56/EU, the Belgian company law defines PIEs as public companies, credit institutions and insurance undertakings (art. 1:13 Belgian Company Law).

13 The Belgian GC audit standard, ‘Audit of a company in financial distress’ (‘Controle van een vennootschap in moeilijkheden’), CA/3.12.1999.

14 Appendix A provides a non-exhaustive listing of paragraphs included in the Belgian local GC audit standard (translated from Dutch to English) compared to the ISA 570 (effective for audits of financial statements beginning on or after December 15, 2009).

15 Despite the similar setting in which our study is conducted, our study differs from the study by Carcello et al. (2009). In particular, our study focuses on investigating the impact of transitioning from a local to a global GC standard, whereas Carcello et al. (2009) examined a change within a local GC standard.

16 This guidance was however neither binding, nor entirely mechanistic in nature: for 37 percent of the companies that violated both financial-juridical criteria the auditor did not disclose GC uncertainties and for 2 percent of the companies that violated neither financial-juridical criteria the auditor did still disclose GC uncertainties.

17 Under the Belgian local GC standard, GC uncertainties were disclosed for 42 (63) percent of the companies that met one (two) of the financial-juridical criteria mentioned in current art. 3:6 §1, 6° Belgian Company Law.

18 Since the expected returns from a viable client are likely larger than those from a client that will cease operations, Matsumura et al. (1997) assume that .$W_S$_. is always greater than .$W_B$_..

19 In Belgium, bankruptcy or insolvency is regulated by Book XX of the Code of Economic Law which stipulates that a company that has permanently ceased to pay and no longer enjoys the confidence of its creditors, is considered to be in a state of bankruptcy. The owner(s), creditors or the Public Prosecution Service can petition the Commercial Court for the company's bankruptcy. If bankruptcy is declared, The Commercial Court appoints an independent trustee (i.e., a lawyer certified in handling bankruptcies), who is responsible for dissolving the company and reimbursing the creditors. This procedure is comparable to a Chapter 7 bankruptcy in the US.

20 Since our analyses are focused on private companies, we exclude 562 company-year observations from 149 individual listed companies.

21 The period 2008–2011 was marked by the global financial crisis and entailed increased auditor conservatism in the form of increased GCOs (Beams & Yan, 2015). To prevent confounding effects, our sample starts in 2012.

22 Besides bankruptcy, a company can also decide voluntarily to cease business. In that case, the dissolution and liquidation is regulated by the Belgian Company Law (Book 2, Title 8) and does not have to be pronounced in Commercial Court.

23 In Belgium, based on audit fees, the next four are BDO, Grant Thornton, RSM and Moore Stephens.

24 The variables SPECFIRM and SPECAP are calculated based on the entire population of Belgian audited companies.

25 Industries are defined by industry sections (A-U) which are based on the standard classification system used in the EU (i.e., NACEBEL codes). Appendix C tabulates a detailed overview of the industries, the corresponding two-digit NACEBEL categories and the number of observations per industry.

26 In terms of company size, the average assets for companies in Panel A is €80,731,045 and for companies in Panel B is €121,388,007.

27 Hardies et al. (2018) report 69 percent type II misclassifications and a prior version of the manuscript reported 28 percent type I misclassifications.

28 Our sample include 348, 42 and 145 (14,320, 3,635 and 11,849) observations from Big 4, 4 second-tier and smaller audit firms, respectively, in the (non-)bankrupt sample. Whilst small audit firms audit approximately 40% of all the audited companies, they represent only 20% of the audit market in Belgium in terms of audit fees.

29 To assess the economic significance of our study, we draw parallels from a Swedish study examining the impact of audit partners’ IQ on type I misclassifications in a comparable low litigious setting. The marginal effect of 0.05 percent, as observed by Kallunki et al. (2019), suggests a relatively smaller effect from an increase in a partner's IQ score – specifically, moving from the lowest to the highest – on type I misclassifications compared to the effect of adopting the ISA 570, which we found to be at 8 percent. The comparison underscores that the adoption of the ISA 570 played an important role in influencing GC reporting accuracy. While extensive archival research has focused on examining the impact of individual auditors on audit quality, a recent study by Bédard et al. (2022) emphasize that the role of the engagement partner is relatively small due to the constraints imposed by audit standards. Our findings align with this notion by highlighting the influence of audit standards in explaining the variability in audit quality.

30 When a company received a GCO in the prior year, the probability of observing the auditor issuing a GCO in the current year increases by 39% (42%) in the (non-)bankrupt sample.

31 This result also demonstrate that the Belgian 2000 GC audit standard was actually very similar to the ISA 570 and that the financial-juridical criteria were not binding for auditors in terms of the issuance of a GCO.

32 Descriptive statistics also suggest auditor differentiation, as evidenced by a decrease in type I misclassification from 34 (31) percent pre-ISA 570 to 18 (23) percent post-ISA 570 for (non-)Big 4 auditors. Conversely, there is an increase (a decrease) in type II misclassifications from 65 (58) percent pre-ISA 570 to 72 (50) percent post-ISA 570 for (non-)Big 4 auditors.

33 In an alternative approach to our split-sample analysis, we examine whether BIG4 moderates the effect of ISA570 on type I misclassifications by incorporating the interaction term BIG4*ISA570 into our model. This analysis reveals significant negative main effects (p = 0.00), and a significant negative interaction effect for BIG4*ISA570 (p = 0.00). This confirms that the impact of ISA570 on type I misclassifications is stronger for Big 4 firms. Further examination through calculation of marginal effects demonstrates that the predicted probability of a type I misclassification is significantly different between Big 4 and non-Big 4 firms prior to (χ2 = 31.73; p = 0.00) and after the implementation of the ISA 570 (χ2 = 121.38; p = 0.00). Prior to (after) the implementation of the ISA 570, Big 4 firms are 3 (6) percent less likely than non-Big 4 firms to commit a type I misclassification.

34 To investigate whether and exclude the possibility that, following the implementation of the ISA 570, non-Big 4 auditors improved their GC reporting accuracy to align with that of Big 4 auditors, we conduct additional analyses. At a univariate level, we find that prior to the implementation of the ISA 570, there is no significant difference in the type II misclassifications rate between Big 4 (65 percent) and non-Big 4 firms (58 percent) (p = 0.211). However, after the ISA 570, our results show a statistically significant difference in type II misclassifications between Big 4 (72 percent) and non-Big 4 firms (50 percent) (p = 0.00). In an alternative approach to our split-sample analysis, we examine whether BIG4 moderates the effect of ISA570 on type II misclassifications by incorporating the interaction term BIG4*ISA570 into our model. This analysis reveals an (in)significant positive effect for BIG4 (ISA570) (p =0.06 (p = 0.65)) and a significant negative interaction effect for BIG4*ISA570 (p = 0.03), suggesting a crossover interaction. This implies that the impact of ISA570 on type II misclassifications differs between Big 4 and non-Big 4 firms. Further examination through calculation of marginal effects demonstrates that, prior to the implementation of the ISA 570, the predicted probability of a type II misclassification is not significantly different between Big 4 and non-Big 4 firms (χ2 = 0.21; p = 0.64). However, after the implementation of the ISA 570, Big 4 firms are 13 percent more likely than non-Big 4 firms to commit a type II misclassification (χ2 = 0.683; p = 0.01). We thank an anonymous reviewer for suggesting these analyses.