Non-audit services provided to audit clients, independence of mind and independence in appearance: latest evidence from large UK listed companies

Loan loss provisions in large publicly quoted European banks and auditor independence

“…Overall, we conclude that an impairment of auditor independence is not responsible for underproviding for loan losses by publicly quoted banks in the EU. We attribute the difference between our findings for banks and those of Campa and Donnelly () for nonfinancial firms in the UK to the regulated nature of the banking industry counterbalancing the pressure of unexpected total fees and unexpected NASs fees on auditor independence. In addition, the results we report when estimating normal LLPs and ALLPs (see Table ) provide support for the change from the incurred loss method to the expected loss method for the impairment of loans introduced by the new IFRS 9.…”

“…It is clear from the aforementioned results that fees, particularly unexpected total fees and unexpected NAS fees, and the NASs fee ratio are not associated with provisioning in a manner that would suggest any compromise of auditor independence for large European banks. This contrasts with the rationale of the European Commission for tightening regulation of the audit market and the results of Campa and Donnelly (), who reported that auditor independence is compromised by higher NASs fees for firms outside the financial services sector. The significant results pertaining to fees point to unexpected NASs fees providing spillover effects in strongly regulated countries relative to weakly regulated countries.…”

“…We estimate the level of unexpected fees as the residuals of a model where auditor fees are regressed on a set of firm characteristics as indicated in the model in Equation (2). The model is derived from Fields et al (), Kanagaretnam et al (), and Campa and Donnelly () and reflects the fact that, in accordance with previous research, auditor fees are related to audit complexity ( MTB , GROWTH ), audit risk ( LOSS , RESTAT , ROA , CGSCORE ) (Ashbaugh et al, ; Firth, ), client size ( SIZE ), bank's credit risk ( SNPL , SLCO ), operating risk ( EFFICIENCY ), liquidity risk ( SECURITIES ), and capital risk ( INTANG , CAPRATIO ). It also includes country and year fixed effects.…”

“…We interact REG with the unexpected fees or economic bond ( AUDCLIBOND ) to investigate if and how it moderates the relationship between ALLPs and the economic bond between auditors and their clients. In accordance with Kanagaretnam et al () and Campa and Donnelly (), our model also controls for additional factors that previous research has documented to be associated with earnings management measures, such as firm size, market‐to‐book ratio, firm growth and performance, corporate governance quality, and other control variables that take into account the reversal of accruals over time (past LLPs) and capital management incentives (tier 1 capital ratio and total capital ratio). The model also includes year and country dummy variables.…”

“…It would appear that the misgivings pertaining to reform of the EU audit market for PIEs expressed in Deumes et al () are well founded, at least as far as large banks are concerned. However, it must be admitted that the UK evidence pertaining to less‐regulated nonfinancial companies provided by Campa and Donnelly () would suggest that the EU reforms do have some merit.…”