Evidence on (the lack of) audit-quality differentiation in the private client segment of the belgian audit market

Abstract

This paper contributes to the empirical audit-quality literature and provides evidence on (the lack of) audit-quality differentiation in the private client segment of the Belgian audit market. Auditor size is used as audit-quality proxy. Prior evidence on audit-quality differentiation between Big Six and non-Big Six auditors in the private client segment of the Belgian audit market is mixed. In this paper we investigate whether these mixed results stem from the inability of the dichotomous Big Six/non-Big Six variable to capture auditor-size differences in a less concentrated audit market. To that end we examine whether alternative continuous measures of audit-firm size (i.e. auditor market share, number of audit-firm clients, number of partners in the audit firm, total assets and operating profit of the audit firm) have a constraining impact on earnings management in a large sample of privately held Belgian companies (n = 1,302). Overall, we do not find evidence that is supportive of quality differentiation in the private client segment of the Belgian audit market.

Acknowledgements

We gratefully acknowledge the comments of the following persons on previous versions of this paper: Jere Francis, Ann Gaeremynck, Parveen Gupta, Axel Haller, Claus Holm, Christopher Humphrey, Niclas Hellman, Robert Knechel, Piet Sercu, Martina Vandebroek, Staf Van Herck, Laurence van Lent, Stefan Sundgren, participants at the 2001 Congress of the European Accounting Association in Athens, Greece, participants at the 2001 Workshop on the Future of Audit Assurance and the Profession in Copenhagen, Denmark, and participants at the 2001 Euroconference on Financial Reporting and Regulatory Practices in Europe in Palermo, Italy. Any remaining imperfections remain the full responsibility of the authors.

Notes

More specifically, DeAngelo (1981) argues that incumbent auditors earn client-specific quasi-rents due to start-up and transaction cost advantages over non-incumbent auditors, because the audit fees after the first-year audit are higher than the avoidable cost of producing audits. The quasi-rents of the auditor's client portfolio act as a collateral bond against ‘cheating’ in order to retain the client in the future, at least as long as the quasi-rents of any particular client are smaller than the quasi-rents of all other clients in the portfolio. The larger the audit firm, the more likely it is that this condition is met. Consequently, audit-firm size serves as a proxy for audit quality.

We continue to use the Big Six/non-Big Six terminology throughout this paper because our empirical data relate to the time period before the merger between Price Waterhouse and Coopers & Lybrand in 1997, and the collapse of Andersen in 2002.

By contrast, DeAngelo (1981) proposes to use the number of clients as a surrogate for audit quality when quasi-rents do not vary across clients, and the proportion of total audit fees depending on retaining any one client in case quasi-rents vary across clients.

Individual financial statements are typically tax driven in Belgium, as taxes are raised on the individual company level, but no separate set of (non-consolidated) financial statements are issued for tax purposes.

Consolidated financial statements are supposed not to be tax driven, as it is required by Belgian law to undo accounting choices that are tax driven in individual financial statements when the statements are consolidated.

Large companies are companies that meet at least two of the following criteria: total assets >3,125,000 euros; turnover >6,250,000 euros, number of employees >50. Companies with more than 100 employees are always classified as a large company irrespective of their total assets or turnover and hence always have to appoint an independent auditor.

In Belgium, auditors are tenured for periods of three years; this requirement is set to promote auditor independence as the opportunity for firms to dismiss their auditor after a non-clean audit opinion is reduced.

During the period 1990–99, 126 disciplinary cases against auditors were initiated relative to an average number of certified auditors of 800 in that period. In about 66% of those cases auditors were indeed sanctioned.

We thank one of the anonymous reviewers for this insight.

Whether a firm actually has tax-loss carry-forwards is not observable. We proxy the existence of tax-loss carry-forwards by whether firms paid taxes in the prior year (that is, no tax-loss carry-forwards) or not (that is, tax-loss carry-forwards).

We use ‘assignments’ to refer to the number of statutory audit jobs.

Results of the Sercu et al. (2002) paper show that tax is only one of the incentives for earnings management in individual financial statements by privately held firms, in addition to incentives provided by other stakeholders such as external financiers. We do not expect an impact of tax incentives on earnings management in consolidated financial statements since Belgian consolidation law requires that any tax-induced accounting choice in individual financial statements be reversed in consolidated financial statements.

The use of tax-loss carry-forwards is unlimited in time in Belgium.

The Belfirst is a CD-ROM issued by Bureau van Dijck and contains the financial statements of all (privately held and publicly held) Belgian and Luxembourg companies that have to submit their financial statements with the National Bank. The financial statements are subject to various arithmetical and logical controls.

The CD-ROM issued by the National Bank is similar to the Belfirst CD-ROM. In contrast to the Belfirst CD-ROM, it also contains information on the number of clients of audit firms over the years 1994–96.

Data based on all firms that submitted financial statements with the Belgian National Bank. Note that the number of mandates per auditor is generally lower for 1996. This is likely due to the fact that not all firms already submitted their financial statements with the National Bank at the date the 1996 CD-ROM was made. The figures, though, give an indication of the ranking of the major audit firms, under the assumption that the firms that submit their financial statements later than the moment at which the CD-ROM was made, are randomly distributed across audit firms.

Our data cover the years 1994–96. The ranking of the largest audit firms has changed since, due to the merger of Price Waterhouse and Coopers & Lybrand in 1997. Our Tables 4 to 8 show that before the merger of Price Waterhouse and Coopers & Lybrand, Price Waterhouse was typically one of the smallest Big Six audit firms in Belgium. Nowadays, PricewaterhouseCoopers is the second largest audit firm in the Belgian audit market. (Before the collapse of Andersen in 2002 and merger into Deloitte & Touche, PricewaterhouseCoopers was the largest audit firm in Belgium.)

We did not find any significant results on the test variables in 10 out of the 12 separate regressions. There was, however, a positively significant result on the test variable (opposite direction than expected) in two positive discretionary accruals sub-samples. A closer look at the data revealed that this is driven by outliers, as the test variable is no longer significant when we estimate the model on the trimmed sample.