We estimate the models above using two way fixed

This preview shows page 10 - 12 out of 25 pages.

We estimate the models above using two-way fixed-effects panel regressions. To test the association between the variables of interest and our audit quality proxies and audit fees, the White cross-section robust standard errors were calculated. Previous studies indicate that the firm characteristics outlined above are useful predictors of the quality of audited earnings and audit fees (see, e.g., Becker et al. 1998 ; Francis et al. 1999 ; Bartov, Gul, and Tsui 2001 ; Frankel et al. 2002 ). In particular, LEVERAGE, LAGLOSS, and ROA are proxies for the financial condition of the firm. According to H. DeAngelo, L. DeAngelo, and Skinner (1994) , troubled firms may have strong incentives to use income-increasing accruals. Moreover, accruals models may overestimate the accruals for firms performing poorly (see, e.g., Dechow, Sloan, and Sweeney 1995 ). Similarly, audit clients with more leverage or lower performance are expected to be seen as more risky by the auditors and, consequently, the audit fee is expected to be higher (see, e.g., Craswell and Francis 1999 ). Furthermore, we expect abnormal accruals and audit fees to be positively associated with INVREC because firms with high proportions of inventory and receivables are less transparent, more difficult to audit, and may have greater opportunities for earnings management (e.g., Francis, Nanda, and Olsson 2008 ; Carcello, Hermanson, Neal, and Riley 2002 ). We also control for whether the firms have foreign operations ( FOREIGN ), as this is likely to increase the complexity of the firm and thus impact on accruals, audit fees, and the choice of auditors ( Francis et al. 2009 ). The one-year sales growth ( SALES_GR ) and the price-to-book ratio ( PB ) are included to control for firm growth, since growth firms have been found to have more incentive to try to match their earnings benchmarks ( Menon and Williams 2004 ; Francis and Yu 2009 ). Finally, previous studies (e.g., Simunic 1984 ; Francis and Yu 2009 ) indicate that LOGASSETS, AC, 8 and BIG4 are negatively (positively) associated with earnings 8 According to the ‘‘ comply or explain ’’ rule of the NASDAQ OMX Nordic exchange, establishing an audit committee is voluntary for the sample firms. 32 Ittonen and Trønnes Auditing: A Journal of Practice & Theory August 2015
management (audit fees). Larger firms may have stronger governance structures, lower information asymmetries, and are generally subject to greater monitoring by auditors and financial analysts (see, e.g., Meek, Rao, and Skousen 2007 ). Moreover, Becker et al. (1998) argue that firm size may also be a surrogate for numerous omitted variables. RESULTS REGARDING AUDIT QUALITY AND AUDIT FEES Table 3, Panel A presents descriptive statistics for the variables used in Equation (2). We winsorize all continuous variables at the 1st and 99th percentile. 9 The first three rows represent our dependent variables, followed by our three variables of interest. Total accruals are negative in the sample period with mean (median) value of ± 0.045 ( ± 0.040), and the absolute discretionary accruals have a mean (median) value of 0.650 (0.040). The mean (median) value of audit fees is 169,000 ( 142,000).

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture