Audit Quality and Banks' Assessment of Disclosed Accounting Information

Abstract

The objective of this paper is to investigate whether banks view the information on the off-balance sheet liabilities (specifically, operating leases) disclosed in the notes to the financial statements as more reliable when it is audited by brand name auditors (i.e. a Big 4 audit firm). To the extent that banks assess a higher likelihood that the financial statements could have material misstatements if it is not audited by a Big 4 audit firm, they should charge a higher interest rate on private loans. Our findings suggest that the impact of operating leases on the interest rate is higher if the firm is audited by non-Big 4 audit firms.

Acknowledgements

The authors acknowledge financial support provided by Social Sciences and Humanities Research Council of Canada – Standard Research Grant. We thank Kiridaran Kanagaretnam, Terry Levesque, Theresa Libby, Bruce McConomy, Flora Niu, Ping Zhang and workshop participants at Wilfrid Laurier University for helpful comments and suggestions.

Notes

¹ Note that there were the Big 5 audit firms at the beginning of our sample period which were later reduced to the Big 4 with the demise of Arthur Andersen. However, for simplicity, we refer to these big auditors as the Big 4.

² We use the term disclosure to identify information that is discussed in the notes to the financial statements but not recognised in the body of those statements.

³ See Hirshleifer and Teoh (2003), Barth et al. (2003), Ahmed et al. (2006) and Davis-Friday et al. (1999).

⁴ Khurana and Raman (2004) find their results in a sample of US firms but are unable to replicate those results among firms in Great Britain, Australia or Canada.

⁵ We measure Ab_acc using the Jones model in cross-section for each two-digit SIC code and year combination. We denote the predicted values of the Jones model as normal accruals and the residuals as abnormal accruals (Ab_acc).

⁶ Firms with SIC codes less than 2000 are mainly natural resource firms. Assets leased by those firms could be very different from those of other firms since these firms often lease land to gain access to resources. Firms with SIC codes between 6000 and 6999 include mostly real estate companies and are likely to be lessors rather than lessees. Finally, firms with SIC codes over 7000 are service and utilities firms.

⁷ We initially have 12,896 loans. From that sample, 6362 have operating leases. The remaining observations are lost due to the absence of data for the control variables.

⁸ This is consistent with suggestions made by Petersen (2009) who states: ‘When both a firm and a time effect are present in the data, researchers can address one parametrically (e.g., by including time dummies) and then estimate the standard errors clustered on the other dimension.’

⁹ As it can be noticed in Table 1, there is a large volatility in the interest rates charged by banks. To ensure that extreme observations do not drive the results, we run the tests winsorising the interest rates at 99% and 95% levels. The results are not affected by these additional conditions.

¹⁰ The results can also be interpreted as consistent with banks believing that they can make better estimates of default risk when the disclosed and recognised information are from a client with a higher quality auditor.

¹¹ See Lennox et al. (2012) for a discussion of difficulties associated with addressing the self-selection problem in accounting studies.